surrogate approach to ssess social resilience in … majeed... · figure 1.1. total reported...

SURROGATE APPROACH TO ASSESS

SOCIAL RESILIENCE IN DISASTER

MANAGEMENT

Abdul Majeed Aslam Saja

Submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

School of Built Environment

Science and Engineering Faculty

Queensland University of Technology

2020

Surrogate Approach to Assess Social Resilience in Disaster Management i

Keywords

Community Resilience; Disaster Resilience; Disaster Management; Disaster Risk

Reduction; Resilience Assessment; Resilience Measurement; Social Resilience;

Surrogate; Surrogate Framework.

Surrogate Approach to Assess Social Resilience in Disaster Management ii

Abstract

The resilience concept has been studied and applied across many disciplines such

as ecology, psychology, and engineering, in the past decades. While the adoption of

the resilience concept in disaster risk management goes back many years, it gained

wider attention among disaster management stakeholders with the adoption of the

Sendai Framework for Disaster Risk Reduction in 2015. Yet, there is no consensus

among different stakeholders as to how it can be measured effectively within the

practical limits of time and human or financial resources.

The importance of social resilience characteristics such as the role of social

support and networks has been highlighted in managing past disasters. New definitions

have evolved, which increasingly recognise the importance of both, social systems and

processes in the preparedness for a disaster. Understanding social resilience can assist

in the formulation of effective disaster management policies to help communities

better prepare for, respond to, and recover from disasters. However, the measurement

of social resilience indicators is not always practical or effective, due to conceptual

and methodological constraints. For example, existing measures of social resilience

indicators do not account for the multi-faceted and dynamic nature of indicators. Some

resilience frameworks also utilise direct measures of resilience, which is resource and

time intensive, and difficult to replicate over time. Hence, new approaches are needed

to measure resilience which optimises resources and within a shorter period of time,

while ensuring the adequacy and reliability of the resilience measurement.

The use of the surrogate approach can help to overcome these limitations by

identifying key facets of resilience indicators. Surrogates are used when a target

indicator is too complex and/or not feasible to measure directly. A surrogate is a

measure to effectively represent a target indicator that is intended to be measured,

based on a well-established relationship between the target indicator and the surrogate.

Further, a sound methodological approach to guide the use of surrogates has not been

investigated to-date in a disaster context, which makes this study innovative and

theoretically significant. To address this research gap, a surrogate development

framework was produced from the review of existent surrogate approach literature in

environmental science to assess social resilience in a disaster context. The surrogate

Surrogate Approach to Assess Social Resilience in Disaster Management iii

development framework was developed in three steps across three interrelated phases

as part of a mixed methods research strategy.

In the first step, five key social resilience indicators, which are very dynamic and

complex for direct measurement, were selected from the ‘5S’ social resilience

framework based on surrogate decision criteria identified in this study. The ‘5S’ is an

adaptive and inclusive framework that was developed in this study based on a critical

review of existing social resilience frameworks, by structuring social resilience in

terms of key dimensions, characteristics, and indicators to guide in selecting social

resilience indicators for surrogate approach. In the second step, key themes of the five

selected social resilience indicators were explored through a case study approach,

utilising interviews. Six potential surrogates were identified for each social resilience

indicator, their surrogacy relationship was established, and surrogate measurement

protocols were proposed. In the third and final step, the potential surrogates were

evaluated against five surrogate evaluation criteria by a wide range of disaster

management experts through an online survey. The survey results were then analysed

using a Multi-Criteria Decision Making technique to rank potential surrogates. By

synthesising the findings from the online survey, first ranked surrogates for social

resilience indicators were selected as priority surrogates for assessing social resilience

indicators.

The findings from this research make an important contribution to advance

resilience assessment in disaster management by applying an innovative approach to

conceptualise, identify, and select surrogates to assess social resilience indicators. The

integrated (revised) surrogate development framework tested in this research through

a robust multi-phase mixed method sequential research strategy sets a new way

forward for resilience assessment in disaster management research. Further, first

ranked surrogates and their measurement protocols for operationalisation, devised

from the research findings have wide practical applicability, largely in any urban

context. They will also guide policy makers and practitioners, particularly at the local

and sub-national levels, to overcome the existing challenges in resilience assessment

in a disaster context. Future research is needed to apply the integrated surrogate

development framework in different contexts, and operationalise the first ranked

surrogates using the proposed measurement protocols to further advance the social

resilience assessment research using surrogate approach.

Surrogate Approach to Assess Social Resilience in Disaster Management iv

Table of Contents

Keywords .................................................................................................................................. i

Abstract .................................................................................................................................... ii

List of Figures ......................................................................................................................... vi

List of Tables ........................................................................................................................ viii

List of Appendices .................................................................................................................. ix

Glossary of Terms .....................................................................................................................x

List of Abbreviations .............................................................................................................. xi

Statement of Original Authorship .......................................................................................... xii

Acknowledgements ............................................................................................................... xiii

List of publications ............................................................................................................... xiv

Chapter 1: Introduction ............................................................................................ 1

Background .....................................................................................................................1

Social resilience to disasters ...........................................................................................3

The potential use of surrogate approach in social resilience assessment ........................4

Research problem ...........................................................................................................5

Research Question, Aim, Objectives and Outputs ..........................................................6

Research significance and contributions.........................................................................7

Research scope ...............................................................................................................8

Thesis outline ..................................................................................................................9

Chapter 2: Literature Review ................................................................................. 12

Social resilience definitions in disaster management ...................................................13

A critical review of social resilience assessment frameworks in disaster

management ..................................................................................................................16

A ‘5S’ inclusive and adaptive social resilience framework in disaster management ...28

Summary and Implications ...........................................................................................32

Chapter 3: Surrogate approach to assess social resilience to disasters ............... 34

An overview of surrogate approach ..............................................................................35

A surrogate development framework for assessing social resilience to disasters.........47

Summary .......................................................................................................................61

Chapter 4: Research Method .................................................................................. 62

Philosophical position of research ................................................................................64

Research methods and strategies ..................................................................................67

Social resilience indicators selected for developing surrogates ....................................74

Research Ethical considerations ...................................................................................79

Surrogate Approach to Assess Social Resilience in Disaster Management v

Phase I – Qualitative case study research for identifying potential surrogates

(Interviews) ...................................................................................................................81

Phase II – Quantitative survey to evaluate and rank potential surrogates (Online

survey) ..........................................................................................................................97

Summary .....................................................................................................................114

Chapter 5: Interview analysis-Identification of potential surrogates ............... 115

Surrogate measures to assess social mobility and access to transport in a disaster

context (Indicator #1) .................................................................................................117

Surrogate measures to assess social trust in a disaster context (Indicator #2) ............133

Surrogate measures to assess learnings from the past disasters as social competence

(Indicator #3) ..............................................................................................................151

Surrogate measures to assess involvement of people with specific needs as social

equity in a disaster context (Indicator #4) ..................................................................168

Surrogate measures to assess cultural/behavioural norms as social beliefs in a disaster

context (Indicator #5) .................................................................................................185

Summary of potential surrogates and selection of set of surrogates for evaluation in

phase II .......................................................................................................................203

Chapter 6: Survey analysis - Evaluation and ranking of potential surrogates 206

6.1. PROMETHEE Ranking and analysis of potential surrogates with equal criteria

weights ........................................................................................................................208

Calculation of surrogate evaluation criteria weight ....................................................220

Overall PROMETHEE ranking with weighted criteria ..............................................220

First ranked social resilience surrogate in disaster management ................................225

Summary of findings from the survey research ..........................................................227

Chapter 7: Synthesis of key findings .................................................................... 228

Key findings in relation to research objective one (RO1) ..........................................229

Key findings in relation to research objective two (RO2) ..........................................232

Key findings in relation to research objective three (RO3) ........................................240

Revisions to the conceptual surrogate development framework ................................247

Chapter 8: Conclusions and recommendations for future research ................. 252

Achievement of research objectives one to three .......................................................252

Addressing the key research question and aim ...........................................................255

Study contributions to knowledge and practice ..........................................................256

Study limitations .........................................................................................................258

Key recommendations for future research ..................................................................259

Bibliography ........................................................................................................... 260

Appendices .............................................................................................................. 277

Surrogate Approach to Assess Social Resilience in Disaster Management vi

List of Figures

Figure 1.1. Total reported economic losses per year, with major events highlighted 1998-2017 1

Figure 1.2. Thesis structure with key sections 11

Figure 2.1. Chapter 2 and key sections in the thesis structure 12

Figure 2.2. PRISMA Flow diagram used for selection of social resilience frameworks 17

Figure 2.3. Summary of the literature review highlighting knowledge gaps in social resilience 26

Figure 2.4. Three layer structure of the ‘5S’ social resilience framework 28

Figure 2.5. An inclusive and adaptive ’5S’model social resilience framework 29


Figure 3.2. Graphical representation of surrogate approach models 37

Figure 3.3. Conceptual analogous surrogate model 38

Figure 3.4. Adaptive Surrogacy Framework 40

Figure 3.5. Vulnerability indicator development framework 42

Figure 3.6. Surrogate development framework to assess social resilience in a disaster context 48

Figure 3.7. Surrogate approach decision flowchart (Key step A) 50

Figure 3.8. A framework to identify potential surrogates (Key step B) 51

Figure 3.9. Steps to select optimum surrogates (Key step C) 56

Figure 3.10. Surrogate evaluation criteria pentagon 58


Figure 4.2. Sequential design integrating qualitative and quantitative research methodologies 69

Figure 4.3. Overall research framework with key steps and processes 72

Figure 4.4. Expanded research framework in three phases 73

Figure 4.5. Extract from the overall research process for the literature review and research design 74

Figure 4.6. Extract from the overall research process for phase I 81

Figure 4.7. Case study locations in the Sri Lanka map 86

Figure 4.8. Administrative and authority governance structures in Sri Lanka 87

Figure 4.9. Sample selection process diagram 89

Figure 4.10. Sequential exploratory design (qualitative to quantitative) 93

Figure 4.11. Levels of interpretation in thematic analysis 94

Figure 4.12. Flow chart depicting the step by step process of case study data analysis 95

Figure 4.13. Sample Leximancer concept/theme maps from interview data 96

Figure 4.14. Cross-case synthesis diagram 97

Figure 4.15. Extract of overall research process for phase II 97

Figure 4.16. Survey design elements 99

Figure 4.17. Nine stages of MCDA process 104

Figure 4.18. Preference functions used in PROMETHEE 108

Figure 4.19. Multi-expert multi-criteria group decision support system flowchart 109

Figure 4.20. Sample PROMETHEE inputs, ranking, and GAIA plot 111


Figure 5.2. Leximancer concept/theme maps from interview data of case study #1 117

Surrogate Approach to Assess Social Resilience in Disaster Management vii




Figure 5.6. Summary of synthesis for first social resilience indicator – social mobility 132





Figure 5.11. Summary of synthesis for first social resilience indicator – social trust 150





Figure 5.16. Summary of synthesis for first social resilience indicator – social competency 167





Figure 5.21. Summary of synthesis for first social resilience indicator – social equity 184





Figure 5.26. Summary of synthesis for first social resilience indicator – social beliefs 201


Figure 6.2. GAIA representation of surrogates to measure indicator#1 210





Figure 6.7. Comparison of net flow values (ranking) of surrogates for indicator #1 222





Figure 6.12. First ranked surrogates for priority assessment of social resilience indicators 226


Figure 7.2. Key step A and related outcomes to achieve RO1 229

Figure 7.3. Key step B and related outcomes to achieve RO2 232

Figure 7.4. Key step C and related outcomes to achieve RO3 241

Figure 7.5. An integrated (revised) surrogate development framework for resilience assessment 248

Figure 8.1. Overall schematic showing achievement of research objectives in each key step 253

Surrogate Approach to Assess Social Resilience in Disaster Management viii

List of Tables

Table 2.1. Social resilience definitions in the disaster management literature 14

Table 2.2. Social resilience frameworks evaluated in this review 19

Table 2.3. Comparison of methods to assess resilience 24

Table 3.1. Some examples of surrogate approach in other disciplines 36

Table 3.2. Comparison of key steps in indicator development frameworks 44

Table 4.1. Alternative combinations of knowledge claims, strategies of inquiry, and methods 66

Table 4.2. Three designs of mixed method research 68

Table 4.3. Details of social resilience indicators selected for developing surrogates 77

Table 4.4. Samples selected for interviews in all study locations 90

Table 4.5. Interview participant profiles 90

Table 4.6. Samples selected for interviews in one study location (DS division) 91

Table 4.7. Key characteristics and profile of survey respondents 101

Table 4.8. Five points scale continuum for each criteria 102

Table 4.9. Fundamental scale of absolute numbers and the scale used in the survey 103

Table 4.10. Comparison of Multi-Criteria Decision Analysis (MCDA) methods 106

Table 5.1. Surrogates (higher-order themes) mapping for indicator #1 . 128

Table 5.2. Surrogates (higher-order themes) mapping for indicator #2 146




Table 5.6. Cross-case tabulation for potential surrogates 203

Table 5.7. Potential surrogates selected for evaluation in online survey 205

Table 6.1. Overall PROMETHEE rankings for five social resilience indicators 208

Table 6.2. Criteria weight obtained from survey responses using AHP 220

Table 6.3. Overall PROMETHEE rankings with weighted criteria 221

Surrogate Approach to Assess Social Resilience in Disaster Management ix

List of Appendices

Appendix A Existing social resilience measures identified in social

resilience frameworks analysed in this study

Appendix B QUT Human Research Ethics Committee Approval

Appendix C Letter of recruitment for interview participants

Appendix D Participant interview consent form

Appendix E Participant information sheet for interviews

Appendix F Interview guide

Appendix G Participant recruitment email for the survey

Appendix H Participant information for the survey

Appendix I Survey questionnaire in key survey

Appendix J Abstracts of published and under review manuscripts

Surrogate Approach to Assess Social Resilience in Disaster Management x

Glossary of Terms

Disaster

“Serious disruption of the functioning of a community or a society

at any scale due to hazardous events interacting with conditions of

exposure, vulnerability and capacity, leading to one or more of the

following: human, material, economic and environmental losses

and impacts” (UN, 2016, p. 13).

Disaster

Resilience

“The ability of a community exposed to hazards to resist, absorb,

accommodate to and recover from the effects of a hazard in a timely

and efficient manner, including through the preservation and

restoration of its essential basic structures and functions”

(UNISDR, 2009, p. 24).

DS division

DS stands for ‘Divisional Secretariat’ division. Several DS

divisions make a district, which is the sub-national level

administration of the state.

GN division

GN stands for ‘Grama Niladhari’ division. The smallest

administrative geography in the Sri Lanka state administrative

system.

Higher-

order theme

A set of closely related themes combined to create a more useful

theme for further description of the context.

Leximancer

Leximancer is a data mining tool to analyse the content of textual

transcripts, which can display concepts and themes frequently

occurring in the transcripts (Leximancer, 2017).

Resilience

“The measure of a system’s or part of a system’s capacity to absorb

and recover from the occurrence of a hazardous event”

(Timmerman, 1981, p. 21).

Social

Resilience

The ability of social entities and social mechanisms to effectively

anticipate, mitigate and cope with disasters and implement

recovery activities that minimize social disruptions and reduce the

impact of future disasters (Bruneau et al., 2003; Kwok et al., 2016;

Rockström, 2003).

Surrogate

A surrogate is defined as an indicator effectively representing

another indicator that is intended to be measured (Miguntanna et

al., 2010; Rodrigues & Brooks, 2007)

Urban

“A smallest administrative division which has a minimum

population of 750 persons, a population density greater than 500

persons per square kilometre, firewood dependence of less than 95

% households, and well-water dependence of less than 95%

households” (Weeraratne, 2016, p. 4).

Surrogate Approach to Assess Social Resilience in Disaster Management xi

List of Abbreviations

AHP Analytic Hierarchy Process

CAT Capacities Coping, Adaptive, and Transformative Capacities

CBO Community Based Organisation

CR Consistency Ratio

CRED Centre for Research on the Epidemiology of Disasters

DMC Disaster Management Centre

DRR Disaster Risk Reduction

DS Divisional Secretariat (Divisional Administrative Division)

GAIA Geometrical Analysis for Interactive decision Aid

GDSS Group Decision Support System

GN Grama Niladari (Village Administrative Division in Sri Lanka

HREC Human Research Ethics Committee

KM Kalmunai (A DS division in Sri Lanka)

KMM Kalmunai Muslim (A DS divisional office in Sri Lanka)

KMT Kalmunai Tamil (A DS divisional office in Sri Lanka)

MCDA/MCDM Multi-Criteria Decision Analysis/Making

NGO Non-Government Organisation

OREI Office of Research Ethics and Integrity

PROMETHEE Preference Ranking Organization METHod for Enrichment

Evaluation

PwSN People with Specific Needs

RO Research Objective

RQ Research Question

SFDRR Sendai Framework for Disaster Risk Reduction

SM Sainthamaruthu (A DS division in Sri Lanka)

UNISDR United Nation – International Strategy for Disaster Reduction

Surrogate Approach to Assess Social Resilience in Disaster Management xii

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution. To the best

of my knowledge and belief, the thesis contains no material previously published or

written by another person except where due reference is made.

Signature:

Date: 16.03.2020

QUT Verified Signature

Surrogate Approach to Assess Social Resilience in Disaster Management xiii

Acknowledgements

First and most of all, I wish to express my heartfelt appreciation to my principal

supervisor, Dr. Melissa Teo, and associate supervisor, Prof. Ashantha Goonetilleke,

for their continuous guidance and insights throughout the period of my PhD studies.

Without their support, my PhD journey and thesis would not have been possible. My

sincere thanks are also extended to Prof. K.W.G. Rekha Nianthi and Dr. Jagath

Gunathilake from University of Peradeniya, and Dr. A.M. Ziyath from Zedz

Consultants for their thoughtful insights and comments on the manuscripts produced

in this study.

I am very grateful to Queensland University of Technology (QUT) and

University Grants Commission (UGC) of Sri Lanka for providing the necessary

financial support during the period of this study. I also take this opportunity to

acknowledge the time and productive discussions by all disaster management experts

who participated in the interviews and survey, which were carried out as part of this

study.

My sincere thanks also go to my academic colleagues at the South Eastern

University of Sri Lanka and my research colleagues at QUT for sharing their optimism

during the struggles of this study.

Thanks to Dr Christina Houen of Perfect Words Editing for editing this thesis

according to the guidelines of the University and the Institute of Professional Editors

(IPEd).

Last, but not the least, I owe an enormous debt of gratitude to my family

members, particularly to my parents, my wife, and my three beloved children, and our

friends, for their love and joy during the difficult times of this study.

I dedicate my PhD thesis to my parents, Majeed and Hanoon, my wife,

Sherfin, and my three beloved children, Nadtha, Nuzayh and Ulfa.

Surrogate Approach to Assess Social Resilience in Disaster Management xiv

List of publications

This research yielded a number of journal papers that were published. An abstract of

each paper in the published or submitted format is provided in Appendix J.

Journal papers published:

(1) Saja, A.M.A., Goonetilleke, A., Teo, M., & Ziyath, A. M. (2019). A critical review

of social resilience assessment frameworks in disaster management. International

Journal of Disaster Risk Reduction, 101096.

Journal Impact Factor – 2.568 (2018) SJR – Q1 (2018)

Citations – 9 [Google scholar] in March 2020

(2) Saja, A.M.A., Teo, M., Goonetilleke, A., & Ziyath, A. M. (2018). An inclusive and

adaptive framework for measuring social resilience to disasters. International

journal of disaster risk reduction, 28, 862-873.


Citations – 24 [Google scholar] in March 2020

(3) Saja, A. M. A., Teo, M., Goonetilleke, A., Ziyath, A. M., & Nianthi, K. W. G. R.

(2020). Surrogate Measures to Assess Mobility of People as a Resilience Indicator

in Disaster Management: An Exploratory Study in Southeastern Sri Lanka.

International Journal of Disaster Risk Science. doi:10.1007/s13753-020-00251-4


(4) Saja, A.M.A., Teo, M., Goonetilleke, A., Ziyath, A. and Gunatilake, J. (2020),

"Selection of surrogates to assess social resilience in disaster management using

multi-criteria decision analysis", International Journal of Disaster Resilience in the

Built Environment, Vol. ahead-of-print No. ahead-of-print.

https://doi.org/10.1108/IJDRBE-07-2019-0045.

Journal Cite Score – 1.130 (2018) SJR – Q1 (2018)

Chapter 1: Introduction 1

Chapter 1: Introduction

This chapter outlines the background (Section 1.1), social resilience to disasters

(Section 1.2), potential use of surrogate approach in social resilience assessment

(Section 1.3), and research problem (Section 1.4). Section 1.5 details research

question, aim, objectives, and outputs. Section 1.6 details the research significance and

contribution and Section 1.7 describes the research scope. Finally, Section 1.7 includes

an outline of this thesis, which is explained using a thesis structure as shown in Figure

1.2.

BACKGROUND

Over the past decades, disasters have continued to devastate many communities.

Significant investments in critical infrastructures and livelihoods have been

increasingly destroyed by disasters. During the last 20 years, over 1.3 million people

have died, 4.4 billion people were injured, or became homeless and needed emergency

assistance due to natural disasters (CRED, 2018). In 2018 alone, there were 281

disaster events recorded globally, with 10,733 deaths, and over 60 million people

affected (CRED, 2019a). The average annual total reported economic losses due to

Figure 1.1. Total reported economic losses per year, with major events

highlighted 1998-2017 (Source: (CRED, 2018), p. 13)


disasters worldwide was 50-100 billion US$ (CRED, 2018), as shown in Figure 1.1.

During 1998-2017, major disasters such as storms killed around 233,000 people and

earthquakes killed more than 747,234 people, highlighting the severity of the disaster

impact, globally (CRED, 2018). Asian countries are the most disaster-affected regions

in the world. For example, in 2018 alone, Asian countries accounted for 45% of

disaster events, 80% of the total number of persons killed, and 76% of the persons

affected, around the world (CRED, 2019b).

Communities who have experienced a major disaster are irrevocably changed in

a number of significant ways, including changes to community demographic structure,

resulting in different forms of resilience to the next disaster. The concept of ‘Build

Back Better’ was coined after the 2004 Indian Ocean Tsunami as a strategy “to

improve living and environmental conditions including through integrating disaster

risk reduction into development measures”, aimed at building more resilient nations

and communities to future disasters (UNISDR, 2015a, p. 8). In line with the strategy

of ‘Build Back Better', new thinking in community resilience concepts, such as

‘bouncing forward from disasters’ has become imperative for effective disaster risk

reduction (Manyena, 2016, p. 41). Two global frameworks, the Sendai Framework for

Disaster Risk Reduction and the 2030 Sustainable Development Goals, adopted in

2015 by United Nations, advocate for greater emphasis in investing in building

resilience and effective risk governance.

In the last decade, a number of frameworks and tools have been proposed for

assessing the resilience of communities to disasters. They have varied in their

approach, emphasis, and scope in determining key indicators to measure resilience

characteristics (Cutter, 2016). While a comprehensive measurement of resilience is

needed, it requires wide-ranging indicators, and their use is often resource and time

intensive (Ziyath et al., 2013). Hence, a practical and robust approach is needed to

measure resilience of communities (Kulig et al., 2013), one that can be adapted and

utilized in different contexts (Sharifi, 2016), to support future resilience building

initiatives.

The resilience of communities is a multi-faceted concept and its characteristics

can be categorized into five key domains: social, economic, environmental,

institutional, and infrastructure (Sharifi, 2016). Within a larger framework for

resilience of communities to disasters, social dimensional resilience characteristics are


described as ‘social resilience’. Social resilience is viewed as a key component in

resilience by many researchers (Kwok et al., 2016), due to the enormous impact of

disasters on people and societies. Social resilience is described as the “ability” or

“capacity” of people, social units (such as communities, social organizations), and

social systems (ranging from families to wider society) to cope, withstand and recover

from a disaster (Bruneau et al., 2003; Maguire & Hagan, 2007).

The importance of social resilience characteristics, such as the role of social

networks in responding to and recovering effectively from disasters, has been proven

in many disaster situations (Aldrich & Meyer, 2015). However, collecting data to

assess social resilience has been challenging in practice due to the very dynamic nature

of social systems and mechanisms. For example, social resilience indicators such as

social trust and community inclusiveness are challenging to assess directly and

accurately in the field (Kwok et al., 2016). Further, the key challenge in using direct

measures is about time and resource constraints for data collection, and replicability

of the method in order to regularly update the resilience status. In contrast, many

existing frameworks have used publicly available census data. The key drawback in

using census data is the limitation of their timeliness and poor ability to adequately

depict the resilience status of a target community. These limitations require innovative

approaches to resolve the difficulty of assessing social resilience indicators.

SOCIAL RESILIENCE TO DISASTERS

Disasters are defined by the United Nations as a

“serious disruption of the functioning of a community or a society at any scale

due to hazardous events interacting with conditions of exposure, vulnerability

and capacity, leading to one or more of the following: human, material,

economic and environmental losses and impacts.” (UN, 2016, p. 13)

Rapid urbanization and poor development planning have increased community

exposure to disasters, generating new risks or exacerbating existing ones, and have

resulted in a sharp increase in disaster related losses (UNISDR, 2015b). A key reason

why existing hazards often evolve into disasters is the failure of communities to

manage risk effectively (Birnbaum et al., 2016). A global emphasis on building

resilience to disasters among communities has been increasing as a result of large

number of devastating disaster events in the last two decades (Cox & Hamlen, 2015).


Communities need to proactively mitigate risks and build resilience to reduce

damage caused by disasters, and to recover more rapidly from disasters (Birnbaum et

al., 2016). The resilience capabilities of communities, however, and their speed and

extent of recovery from disasters, often differ significantly (Burton, 2015). Levels of

social resilience are often dependent on a number of complex local and context-

specific factors, such as socio-economic status, extent of external support and aid

provision, and past experience of disasters. The complexity of the social resilience

phenomena necessitates that policy makers, practitioners and researchers understand

the unique characteristics of resilient communities to help them better prepare for and

recover from disasters.

In this study, social resilience is defined as the ability of social entities and social

processes to effectively anticipate, mitigate, and cope with disasters, and implement

recovery activities that minimize social disruptions and reduce the impact of future

disasters (Bruneau et al., 2003; Kwok et al., 2016; Rockström, 2003). This definition

is adapted to consider the abilities of both social entities and social processes, along

the different phases of a disaster, i.e. ex-ante, disturbance, and ex-post. Resilience in

general, and more specifically, social resilience in a disaster context, may involve a

transformation to another state of social systems, rather than conservation of the

functionalities of existing social systems (Alexander, 2013): i.e., the system does not

necessarily need to return to its pre-disaster state. Hence, social resilience needs to be

conceptualised as a proactive ability of social entities and mechanisms, as opposed to

defining it only as the reactive capability of responding to a crisis (Matyas & Pelling,

2015), because the impact of disasters can be largely mitigated by enhancing resilience

before disasters occur (Birnbaum et al., 2016).

THE POTENTIAL USE OF SURROGATE APPROACH IN SOCIAL

RESILIENCE ASSESSMENT

The complexity and dynamic nature of social systems and processes often

challenge the assessment of their resilience in a disaster context. An innovative

application of a surrogate approach can help to overcome challenges in assessing

resilience in disaster management. This has been tested in other disciplines such as

environmental science and clinical medicine to address similar challenges in the

assessment of complex concepts. A surrogate is defined as an alternative measure to

determine the target indicator (Miguntanna et al., 2010; Rodrigues & Brooks, 2007).


The adoption of the surrogate approach in disaster management can capture key facets

of a resilience indicator to be measured, so that the challenges in measuring complex

and abstract resilience indicators can be addressed.

The surrogate approach has not been investigated in disaster management

literature to-date. Surrogates are useful for several reasons. For example, the time

required to measure the target system can be much shorter with a surrogate, whilst

surrogates are easier to measure than the target system, and the sample size or the

extent of the measurement entity becomes much less for a surrogate than for the target

system (Baker, 2005). In the surrogate approach, a methodical step is applied to

conceptualise, identify, and evaluate potential surrogates in order to finally select the

most robust set of surrogates for application.

The surrogate approach can be applied to assess target indicators that are often

difficult, or not feasible through direct measures in resilience assessment. Once the

surrogate is selected for a particular context through a rigorous and methodological

process of identification and evaluation, the selected surrogates can be used to monitor

and update the resilience status regularly. The application of selected surrogates at the

community and sub-national levels will play a vital role in devising the priority

resilience building activities.

RESEARCH PROBLEM

The concept of resilience has gained more prominence in disaster management

in recent years, due to ever increasing losses and damage caused to communities by

disasters. However, the complex and dynamic nature of social resilience characteristics

due to varied conceptualisation approaches in different contexts has further

complicated the assessment of social resilience to disasters (Burton, 2015). The key

challenge in social resilience is to translate abstract and complex concepts to enable

an adequate assessment that can inform effective resilience investment decisions.

Consequently, the need for a resource- and time- effective approach to assessing social

resilience to disasters has been advocated by many researchers.

Existing social resilience assessments have used direct measures, such as

household surveys and/or data obtained from census information sources available in

public databases (Saja et al., 2019). For example, direct methods such as household

questionnaire surveys have been used to measure indicators such as levels of risk


awareness and preparedness, past disaster experience, and trust in authorities (Tapsell,

2007). However, effective use of direct measures for social resilience assessments is

minimal (Beccari, 2016); the feasibility and replicability of such methods are

questionable due to time and resource limitations to meet the need for continuous

updating and monitoring of resilience (Ziyath et al., 2013).

Indirect measures are helpful when it is complex or not feasible to measure the

intended indicator directly (Becker et al., 2015). However, the existing social

resilience frameworks typically omit more dynamic and process oriented, but

important, resilience indicators for adequate measure of resilience. The selection of

variables from census reports has an important limitation that the required objective

variables may not be available or translatable to the indicator being measured. This is

also problematic in many countries, where census is not undertaken frequently. Hence,

there is an urgent need to conceptualise an innovative approach to enable effective

assessment of social resilience that can objectively capture key facets of indicators to

be assessed and can be easily and regularly updated. This remains a research gap that

needs to be addressed. The surrogate approach can help to address the challenges in

identifying key facets of process-oriented social resilience indicators which are

difficult to measure. Hence, this research proposes to test the surrogate approach by

developing a framework for conceptualising, identifying, and evaluating key facets of

social resilience indicators, which can act as potential surrogates to measure the target

indicator.

RESEARCH QUESTION, AIM, OBJECTIVES AND OUTPUTS

The key research question (RQ) generated from the literature review is:

How can key social resilience indicators in a disaster context

be measured using surrogate approach?

Research Aim: To investigate the above key RQ, the main aim of this research was

to develop an approach for conceptualising, identifying, and evaluating surrogates to

assess social resilience indicators in a disaster context.

Three research objectives (RO) were developed to allow detailed investigation of the

key research question. The corresponding research objectives and respective outputs

are provided below.


Research Objective 1 (RO 1): To identify key social resilience indicators that

require surrogate approach by developing an inclusive and adaptive social

resilience framework in a disaster context

Output 1: An inclusive and adaptive social resilience framework and surrogate

decision criteria to identify social resilience indicators to apply surrogate approach

Research Objective 2 (RO 2): To identify potential surrogates to measure key

social resilience indicators in disaster management

Output 2: A framework to guide the identification of potential surrogates to measure

social resilience in disaster management

Research Objective 3 (RO 3): To evaluate and select optimum surrogates for

application by ranking the potential surrogates against surrogate evaluation

criteria

Output 3: A framework to evaluate and rank potential surrogates to select the optimum

surrogate(s) for application

RESEARCH SIGNIFICANCE AND CONTRIBUTIONS

The introduction of a surrogate approach in disaster management research is

particularly significant; it will meet the challenge of devising a method for adequate

assessment of social resilience that can help to formulate resilience building programs

and policies by local disaster management stakeholders. Building resilient

communities is crucial to preparing, mitigating and recovering from potential

disruptions caused by disasters (Cox & Hamlen, 2015). Effective disaster preparedness

and risk reduction strategies can be planned and implemented by assessing existing

social resilience conditions at the local level. However, an adequate assessment of

social resilience has remained a challenge, due to several practical and methodological

difficulties. Hence, developing surrogates and translating them to real world

application to assess resilience indicators will help in effective resilience investment

decision making (Lindenmayer et al., 2015b).

The integrated surrogate development framework proposed in this research will

facilitate the robust assessment of social resilience indicators that are not feasible to

assess through existing direct methods and census-based measures. The use of the

surrogate approach as a novel method tested in this research is a key contribution to


the resilience assessment theory in disaster management, which has not been

developed to-date in the disaster management sector.

The surrogate model and a method to operationalise the set of robust surrogates

identified from the outcomes of this research will assist policy makers and practitioners

to assess social resilience at local and sub-national levels. An adequate assessment of

dynamic and process oriented social resilience indicators in a disaster context has been

an important knowledge gap in disaster management. This research has contributed to

addressing this challenge by using a surrogate approach. This approach will also guide

future research for assessing other dynamic and process oriented resilience indicators

in disaster management that were found to be very difficult to assess through

conventional methods.

RESEARCH SCOPE

This study focuses on developing a new approach to assess social resilience indicators

using surrogates in disaster management. The scope of this research study is as

follows.

(1) Urban context: Data collection to identify potential surrogates in the first phase of

this research was done in an urban context, using Sri Lanka as a case study. The focus

was on urban context to make the findings widely applicable, as 54% of the world

population reportedly live in urban areas, and 16% live in large cities around the world

(WorldBank, 2014). The study findings and outcomes are generic and can be

applicable in any urban context with proper contextualisation of socio-economic and

disaster characteristics.

(2) Expert consultation: This study employed sequential exploratory mixed method

research design, which used both, qualitative and quantitative data collection methods.

The initial consultation with experts through interviews in the case study (Phase I –

qualitative research) was done with disaster management practitioners working at the

local and sub-national levels in Sri Lanka. Experts at the national and international

level were reached through an online survey (Phase II – quantitative research) for

evaluating the potential surrogates. This method facilitated to reach a wide audience

of practitioners, policy makers, and researchers working in disaster management in

different countries, which helped to increase the validity of the findings.


THESIS OUTLINE

This thesis consists of eight key chapters, and a graphical representation

depicting the key content covered in each chapter is shown in Figure 1.2:

Chapter 1 – ‘Introduction’ explains briefly the background tor this thesis, social

resilience to disasters, the potential use of surrogate approach in social resilience

assessment, the research problem, research question, research aim, objectives and

outputs, research signification and contribution, and finally the scope of this study.

Chapter 2 – ‘Literature Review’ is structured into four sections based on the

critical review of literature relevant to this research study. These sections include:

social resilience definitions in disaster management (Section 2.1), a critical review of

social resilience assessment frameworks in disaster management (Section 2.2), a ‘5S’

inclusive and adaptive social resilience framework in disaster management (Section

2.3), and finally, a summary and implications for this research (Section 2.4).

Chapter 3 – ‘Surrogate approach to assess social resilience to disasters’ has three

sections. Section 3.1 provides an overview of the surrogate approach, Section 3.2

details the framework developed to identify and assess the use of social resilience

surrogates in the disaster context, and finally there is a summary of the chapter (Section

3.3).

Chapter 4 - ‘Research method’ is divided into eight sections. This chapter aims

to provide an overall summary of the research methods applied in this research. Section

4.1 gives an overview of philosophical assumptions of the research, Section 4.2

provides an overall research method, and Section 4.3 details the selection of social

resilience indicators for developing a surrogate approach. Section 4.4 details ethical

research considerations. Section 4.5 details the qualitative method applied in phase I

to identify surrogates, and Section 4.6 details the quantitative method applied in phase

II to evaluate and select surrogates; the final Section 4.7 provides a summary.

The next two chapters (Chapter 5 and Chapter 6) report the findings of this study

in two phases, respectively. Chapter 5 discusses the main findings of phase I interviews

across four case studies for five social resilience indicators – social mobility, social

trust, social competence, social equity, and social beliefs which are covered

respectively in Sections 5.1 to 5.5. The final Section 5.6 reports the overall findings


from phase I that include a cross-case synthesis of potential surrogates and selection

of surrogates for evaluation in phase II of this study.

Chapter 6 details the findings from phase II of this study and has five sections.

Section 6.1 presents the evaluation and ranking of potential surrogates using equal

criteria weights. Section 6.2 presents the calculation of criteria weights from the survey

data, while Section 6.3 presents the PROMETHEE ranking obtained with experts’

consolidated criteria weights. Section 6.4 presents the list of first ranked surrogates

selected based on the ranking of potential surrogates obtained in the online survey.

Finally, Section 6.5 provides a summary of phase II findings.

Chapter 7 – ‘Synthesis of key findings’ is divided into four sections, which

synthesises the key findings in Chapters 4, 5 and 6. Section 7.1 synthesises the findings

in relation to RO1, Section 7.2 synthesises the findings in relation to RO2, and Section

7.3 synthesises the findings in relation to RO3. Section 7.4 presents an integrated

(revised) surrogate development framework for future application in disaster

management by revising the conceptual surrogate development framework tested in

this study.

The final chapter (Chapter 8), ‘Conclusions and recommendations for future

research’, has five sections. These include: Section 8.1 - Achievement of three study

objectives’, Section 8.2 – Addressing the key research question, Section 8.3 – Study

contributions to knowledge and practice, Section 8.4 – Study limitations, and Section

8.5 – Key recommendations for future research.


Figure 1.2. Thesis structure with key sections

Chapter 2: Literature Review 12

Chapter 2: Literature Review

This chapter consists of four sections of the literature review, including a

summary section (Section 2.4) at the end, as shown in Figure 2.1. The first three key

sections (Sections 2.1 to 2.3) discuss the following topics in social resilience literature:

- Section 2.1 focuses on social resilience definitions in disaster management.

- Section 2.2 provides a critical review and identifies key knowledge gaps of social

resilience assessment frameworks developed in a disaster context.

- In Section 2.3, an inclusive and adaptive ‘5S’ social resilience framework in

disaster management is presented.

- Finally, the summary and implications of the literature review are provided (in

Section 2.4); the contents of Chapter 2 are summarised highlighting key problems

and knowledge gaps in social resilience assessment in a disaster context; from this

a key research question studied in this research is formulated.

Figure 2.1. Chapter 2 and key sections in the thesis structure


SOCIAL RESILIENCE DEFINITIONS IN DISASTER MANAGEMENT

This section details the different ways that social resilience is defined in disaster

management literature, and the definition adapted in this study.

Resilience means the ability of an object or entity to return to its original shape

after an adverse event. The concept of resilience has been studied and applied across

many diverse disciplines, including ecology, biology, social-ecological systems, social

science, and psychology (Ainuddin & Routray, 2012; McMillen et al., 2016; Norris et

al., 2008; Quinlan et al., 2015). For example, the ‘bounce-back’ analogy from

engineering resilience, the ‘resistance’ concept in social vulnerability, and the

‘robustness’ concept in social-ecological systems theory, have contributed to diverse

interpretations of resilience in the current literature (Matyas & Pelling, 2015).

Specifically, social resilience has been broadly studied in natural resource

management, social change and development, and disaster management (Keck &

Sakdapolrak, 2013).

Due to the complexity of defining resilience, new thinking in resilience

recognises the complex relationships between the built, natural, and social

environments and their influences on the understanding of resilience to disasters

(Norris et al., 2008). Quinlan et al. (2015) suggested, however, that “while multiple

conceptions of resilience can be problematic in terms of common indicators and

comparable metrics, they can also extend the concept to a broader spectrum of contexts

and drive exploration for better approaches to implementation” (p. 679).

The use of the resilience concept has developed in the ecological sciences since

the 1970s. Although the history of resilience concept can be traced back many decades,

the term ‘resilience’ was brought to prominence in ecology by C.S. Holling,

(Alexander, 2013). Holling (1973) defined it as a measure of the ability of an

ecological system to sustain disturbances and still persist. In one of the first definitions

of resilience in a disaster context (Klein et al., 2003), Timmerman (1981) suggests that

resilience is “the measure of a system’s or part of a system’s capacity to absorb and

recover from the occurrence of a hazardous event” (p. 21).

The study by Adger (2000) on ecosystems is widely acknowledged as the first

study to define social resilience (Keck & Sakdapolrak, 2013). In the context of socio-

ecological systems, in this case a mangrove conversion, Adger (2000) defined social


resilience as “the ability of communities to withstand external shocks to their social

infrastructure” (p. 361). Adger (2000) further highlighted the need to consider

contextual social attributes in defining social resilience because of varying differences

in community institutions and resource priorities.

Social resilience in a disaster context is defined in many different ways, as shown

in Table 2.1: a) abilities of social entities or largely social systems that include families,

the wider community, social groups, organisations, resources, and structures (Khalili

et al., 2015; Maguire & Hagan, 2007); b) abilities of social mechanisms such as

understanding and managing emerging risks as well as self-organisation and

transformation capacities (Kimhi & Shamai, 2004; Shaw et al., 2014); c) abilities of

both, social entities and mechanisms (both outcome and process oriented features)

(Kwok et al., 2016); and d) community’s coping, adaptive, and transformative

capacities to withstand and recover from disasters (Parsons et al., 2016). Additionally,

the diversity and dynamics of the context, in which the concept of resilience is

operationalised, brings more complexity to the definition (Alexander, 2013).

Table 2.1. Social resilience definitions in the disaster management literature

Type of

definitions

Disaster

context Social resilience definition Ref.

a)Ability of

social entities

All

disasters

“The capacity of social groups and

communities to recover from, or respond

positively to, crises.”

“The capacity of a social entity (e.g., a

group or community) to bounce back or

respond positively to adversity.”

(Maguire

& Hagan,

2007, p.

17)

Community

seismic

resilience

“Ability of social units (e.g., organisations,

communities) to mitigate hazards, contain

the effects of disasters when they occur, and

carry out recovery activities in ways that

minimise social disruption and mitigate the

effects of future earthquakes.”

(Bruneau

et al.,

2003)

(p.4)

All

disasters

“Internal ability of the social system to

counteract events described as the failure of

expectation toward its environment during

disasters, crises and emergencies.”

(Lorenz,

2013)

(p.12)

All

disasters

“The ability of a community to withstand

external social shock toward enhancing

social capacity to resist disaster losses

during disaster and regenerate after

disaster.”

(Khalili et

al., 2015,

p. 249)


Type of

definitions

Disaster

context Social resilience definition Ref.

Socio-

Ecological

Systems

“Ability of groups or communities to cope

with external stresses and disturbances as a

result of social, political and environmental

change.”

(Adger,

2000, p.

347)

All

disasters

“The capacity of a social system (e.g., an

organisation, city, or society) to proactively

adapt to and recover from disturbances that

are perceived within the system to fall

outside the range of normal and expected

disturbances.”

(Boin et

al., 2010,

p. 9)

b) Ability of

social

mechanisms/

processes

Drought

Mitigation

“Social coping mechanisms that are used to

cope with extreme unmanageable shocks.”

(Rockströ

m, 2003,

p. 871)

Coastal

flooding

“Social resilience = risk perception X self-

perception X accepting change X self-

organisation.”

(Shaw et

al., 2014)

(p. 202)

All

disasters

Individuals’ sense of the ability of their

own community to deal successfully with

the emerging threat.

(Kimhi &

Shamai,

2004)

(p. 442)

c) Ability of

social entities

& mechanisms

All

disasters

“The ability of a community's social

environment to effectively anticipate, cope

with, and recover from disasters, which

depends on the presence and robustness of

other community features, resources, and

processes”

(Kwok et

al., 2016)

(p. 205)

d) Coping,

adaptive, and

transformative

capacities

All

disasters

“The capacity of communities to prepare

for absorb and recover from natural hazard

events, and the capacities of communities to

learn, adapt and transform towards

resilience”

(Parsons

et al.,

2016)

(p. 3)

In this study, social resilience is defined as the ability of social entities and

processes to effectively anticipate, mitigate, and cope with disasters, and implement

recovery activities that minimise social disruptions and reduce the impact of future

disasters (Bruneau et al., 2003; Kwok et al., 2016; Rockström, 2003). This definition

is adapted to consider the abilities of both social entities and social processes, along

the different phases of a disaster, i.e. ex-ante, disturbance, and ex-post.


A CRITICAL REVIEW OF SOCIAL RESILIENCE ASSESSMENT

FRAMEWORKS IN DISASTER MANAGEMENT

A systematic literature review was conducted to critically analyse current social

resilience assessment frameworks in the disaster context. The key findings and

knowledge gaps are presented in this section.

2.2.1 Social resilience assessment frameworks in disaster context

Many social resilience assessment frameworks have been developed in a disaster

context as a result of different ways of defining and conceptualising social resilience

(Aldunce et al., 2015; Cutter, 2016; Djalante & Thomalla, 2011; Norris et al., 2008).

The understanding of different concepts used to frame social resilience in a disaster

context is critical to further advancing these frameworks to comprehensively measure

social resilience characteristics. Table 2.2 summarises 31 social resilience frameworks

that were identified through a comprehensive literature search and critically analysed.

A systematic literature review was conducted to critically analyse current social

resilience frameworks in disaster context. In selecting literature for inclusion in the

review, social resilience frameworks developed within the last decade (2005-2015) in

the disaster management sector were identified. The period between 2005 and 2015

coincided with the period of the implementation of the Hyogo Framework for Action

(HFA) which was the global guiding framework for disaster risk reduction efforts since

2005. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses –

PRISMA (Moher D, 2009) method was used to select social resilience frameworks for

a detailed review as shown in Figure 2.2. PRISMA is an established method for

guiding systematic review of academic literature, and is based on four steps:

Identification, Screening, Eligibility, and Inclusion.

Key steps of the PRISMA method is explained below:

Step 1 - Identification stage: The keywords, “Social” OR “Community” AND

“Resilience” were used in the search to identify all potential peer-reviewed research

articles on social resilience. The initial literature search was done using the Scopus

database, which is a comprehensive research database for peer-reviewed literature, for

the period between 2005 and 2015 in Title, Abstract, and Keywords (Scopus, 2017).

This resulted in the identification of 12,121 research articles.


Step 2 - Screening stage: In this stage, search limiters on relevant disaster related

disciplines, such as social science, environmental science, and multidisciplinary

studies were used. Journals that are not relevant to disaster research and in languages

other than English were excluded. Consequently, 1,194 relevant research articles were

selected.

Step 3 – Eligibility check: In this step, the titles of the selected 1,194 research

articles were screened to narrow down the search for the most relevant articles on

social resilience to disasters. The title and abstracts of articles that do not relate to

social/community resilience in any natural disaster context were excluded. Finally, 172

research articles were selected for detailed review of abstracts.

Step 4 – Inclusion step: Based on the review of the abstracts of the 172 articles,

16 social resilience frameworks were selected. Using the snowball technique, 15

Figure 2.2. PRISMA Flow diagram used for selection of social resilience frameworks

for analysis


additional articles were included. Accordingly, a total of 31 different social resilience

frameworks were selected for detailed analysis.

In Table 2.2, the 31 frameworks are categorised by the researcher (year), hazard

focus, and country of development (geographic coverage) along with the method

adopted for developing the framework. Only social resilience characteristics were

included from the multi-dimensional disaster resilience frameworks. However, other

resilience characteristics that are closely related to social resilience, such as social-

culture and human capital, were also considered in the evaluation.

The 31 social resilience frameworks can be categorised into two main

approaches: distinctive single dimensional frameworks, and those embedded within

multi-dimensional community resilience frameworks. While the embedded social

resilience framework is part of a broader multi-dimensional framework, a distinctive

framework is a stand-alone social resilience framework.

Distinctive single dimensional frameworks: These frameworks consider only

one dimension of community resilience - social resilience or economic resilience

characteristics. They cover single-dimensional resilience characteristics in more detail.

For example, Kwok et al. (2016) identified 66 unique social resilience characteristics

and finally proposed 18 key social resilience characteristics. In another study, Khalili

et al. (2015) identified 18 social resilience characteristics. Both frameworks have

captured some similar characteristics/indicators, such as leadership and social trust.


Table 2.2. Social resilience frameworks evaluated in this review

(1) Distinctive social resilience frameworks

# Framework Country Hazard Method adapted Ref.

1 Social Resilience for New Zealand New Zealand Multi hazard Workshop (Expert Opinion from Researchers,

Practitioners, and policy makers)

(Kwok et al., 2016)

2 Social-Ecological Flood Resilience Index Southern Cape of South

Africa

Flood Secondary data from public databases (Census

data)

(Kotzee & Reyers, 2016)

3 Temporal Social Resilience Framework Australia Flood Literature review Case study in 2 locations (Khalili et al., 2015)

4 Social System Resilience Not specific Multi hazard Literature review (Lorenz, 2013)

5 Social Resilience Framework Australia Common to all

disasters

Literature review (Maguire & Hagan, 2007)

(2) Social resilience frameworks embedded within the multi-dimensional community resilience frameworks

# Framework Country Hazard Method adapted Ref.

6 Australian Natural Disaster Resilience

Framework Index (ANDRI)

Australia Multi hazard Literature review (Parsons et al., 2016)

7 Community Disaster Resilience

Index (CDRI)

Korea Multi hazard Index development with secondary data (Yoon et al., 2016)

8 Community Disaster Resilience Framework

for Iran

Iran Multi hazard Expert opinion through Focus Group Discussions (Ostadtaghizadeh et al.,

2016)

9 Community Resilience Score Card Australia Multi hazard Score card approach (Arbon et al., 2016)

10 Social Resilience Index (SRI) (Component

Community Resilience Index)

Pakistan Flood Questionnaire and secondary data (Qasim et al., 2016a)

11 Resilience Assessment

in Slums

Kenya Slums Multi hazard Workshop and case study approach (Woolf et al., 2016)

12 Conjoint Community Resiliency

Assessment Measure (CCRAM)

Israel-Gaza borders Security threats Web-based survey (Longitudinal study) (Leykin et al., 2016)

13

Community Resilience to Disasters in Saudi

Arabia

(CRDSA)

Saudi Arabia Multi hazard Delphi based expert opinion/AHP (Alshehri et al., 2015a)

14 Composite Community Resilience Index Baluchistan, Pakistan Earthquake Literature review, Composite index (Ainuddin et al., 2015)


15 Disaster Resilience Indicators Mississippi Gulf Coast

in the United States

Hurricane Case study, secondary data for index

development

(Burton, 2015)

16 Rural Resilience Index (RRI) USA Multi hazard Literature review, Pilot study (Cox & Hamlen, 2015)

17 Baseline Resilience Indicators for Communities (BRIC)

US Counties Multi hazard Public and freely accessible data sources, and created dimensional scores

(Cutter et al., 2014), (Cutter et al., 2010)

18 Climate Disaster Resilience Index (CDRI) Chennai, India Multi hazard Literature review, CDRI scores (Joerin et al., 2014)

19 Social Dimension Score (Resilience Score) Indonesia Multi hazard In-depth Interviews, Focus Group Discussions (Kusumastuti et al., 2014)

20 Baseline Resilience Indicators

for Communities (BRIC)

Sarasota County,

Florida, USA

Multi hazard Interviews, plan review, focus group, and spatial

analysis

(Frazier et al., 2013)

21 Community Resilience Framework Los Angeles, USA Health focus Online survey (Chandra et al., 2013)

22 Communities Advancing Resilience Toolkit

(CART)

USA Multi hazard A toolkit consists of assessment and data

collection tools

(Pfefferbaum et al., 2013)

23 Disaster Resilience Index Japan Coastal hazards Analytic Hierarchy Process (AHP) (Orencio & Fujii, 2013)

24 Communities Advancing Resilience Toolkit

(CART)

USA Multi hazard Web-based survey using expanded version of the

CART

(Norris et al., 2008),

(Sherrieb et al., 2012)

25 Community Disaster Resilience Indicators

(CDRI)

US Gulf Coast Region Multi hazard Results from implementing a project (Peacock et al., 2010)

26 Community Resilience

Index (CRI)

Mississippi counties Multi hazard Index creation (Archival and population data) (Sherrieb et al., 2010)

27 Coastal Community Resilience Index

(CCRI)

USA Multi hazard Community Self-Assessment Tool (Index) (Sempier et al., 2010)

28 PEOPLES Resilience Framework USA Multi hazard Literature Review (Renschler et al., 2010)

29 Disaster Resilience of Place (DROP) USA Multi hazard Conceptual model based on literature review (Cutter et al., 2008)

30 Building resilience in rural communities

toolkit

Australia Multi hazard Research/community consultation (Hegney et al., 2008)

31 Capital based Community Disaster

Resilience

USA Multi hazard Index development with secondary data (Mayunga, 2007)


However, there are many differences as well. For example, Khalili et al. (2015) used

social innovation and learnings from previous disasters, which were not prioritised in

the framework developed by Kwok et al. (2016). This leads to a lack of consistency in

fundamental characteristics of social resilience used in contextual frameworks.

Embedded social resilience frameworks: These frameworks view community

resilience as multi-dimensional, using variations of resilience characteristics in the

following five dimensions - social, economic, infrastructure, institutional, and

environment. More than two-thirds of the embedded social resilience frameworks have

not considered key dimensions such as community aspirations, goals and efficacy,

social institutions, social safety measures, equity and diversity/inclusiveness, local

culture/beliefs/faiths, community processes, and education.

There is a tendency to prioritise easily measurable resilience characteristics such

as social demography and social networks, and many process-related resilience

characteristics have received relatively lower priority due to difficulty in measurement.

The broader the framework with multi-dimensions, the lesser is the focus on each

dimension, and vice-versa. The scale of exclusion of important social resilience

characteristics is evident in many embedded frameworks, since it needs to include

characteristics in all dimensions. Hence, they mostly leave out key social resilience

characteristics that may require extensive assessment, which limits adequate

assessment of social resilience.

2.2.2 Social resilience characteristics and indicators

A typical social resilience framework has two layers: characteristics and

indicators. Most of the social resilience frameworks have adapted the

characteristics/indicators-based approaches to measure social resilience, because those

frameworks can be easily operationalised. In general, a set of indicators is needed to

characterise social resilience as well as for the practical application of social resilience

frameworks to different types of disasters and for varying contexts. A resilience

indicator is a description of observable or measurable information that is used to

identify the state or function of social entities (Jülich, 2017; UnitedWay, 1996).

A set of indicators will collectively measure a resilience characteristic. However,

at times, the distinction between characteristics and indicators is not rigid (Twigg,

2009). The resilience indicators can be process indicators or outcome indicators. The

outcome indicator shows how well certain activities accomplish their proposed results,


and the process indicator is a measure of how well the activities are implemented

(Doorn, 2017). Both process and outcome indicators are important for measuring

social resilience, because process indicators help to understand the community and

sustainability of the community programs, whereas outcome indicators reflect the

achievements of community capacity building and empowerment (Kafle, 2012).

A lack of agreement and consistency in framing the social resilience

characteristics adds a further degree of complexity to the useful transformation and

utilisation of the concept in multiple contexts. For example, around one-third of the

frameworks only have two layers and do not identify indicators. Two-thirds of the

frameworks analysed had a third layer as indicators, with the exception of a few

frameworks that have indicators as the second layer in one social dimension. Some

frameworks have termed indicators as variables. There is also lack of clarity about

characteristics and indicators in some frameworks, and characteristics are sometimes

labelled as indicators. Therefore, a key knowledge gap in social resilience research is

a systematic structuring of a social resilience framework in logical layers, which can

guide the practitioners to operationalise it uniformly across different communities.

Future development of social resilience frameworks needs to consider key process-

oriented social resilience characteristics that are critical for adequate measurement of

social resilience, since they are largely neglected in the existing frameworks.

2.2.3 Challenges in measuring social resilience indicators

Social resilience to disasters can be measured for many different purposes. These

include: (a) measuring the relative level of resilience between two geographic

locations; (b) measuring changes in resilience status (trend) over the continuous span

of time within a single phase of the disaster; (c) measuring resilience in multiple stages

of a community at discrete time intervals, such as the changes between different

disaster phases (before, during, and after a disaster); (d) measuring hazard-specific

resilience of a community; and (e) combination of above scenarios. However,

“resilience has proven difficult to measure, and an alternative to estimating resilience

directly is to monitor characteristics of systems that are related to the resilience of the

system and are measurable” (Bennett et al., 2005, p. 946).

The key conceptual challenge in measuring resilience is that resilience is not

only an outcome, but also a process oriented phenomenon (Cox & Hamlen, 2015). The

outcome oriented resilience characteristics are static conditions with a single


measurable target, whereas process oriented characteristics describe dynamic

properties of resilience (Cutter, 2016). Within most existent literature, social resilience

is perceived as a static characteristic (Cutter et al., 2008) for measurement rather than

a process related indicator (Sharifi, 2016), mainly because process indicators are not

easily measurable. For example, the number of civic organisations is an outcome

indicator to measure civic engagement in social networks. However, the existence of

many civic organizations cannot alone enhance social resilience. It is important to

understand how the type of activities undertaken by civic organisation can increase

social resilience (Cutter, 2016), which are process oriented indicators. Likewise,

adaptive capacity cannot be measured using “number of years of schooling” as an

indicator alone, because people’s adaptive capacity cannot be simply measured by how

long they attend school (Levine, 2014). Therefore, a failure to measure both, outcome

and process features of resilience results in inadequate and inaccurate assessment of

social resilience. However, it is often a challenging task to operationalise resilience

frameworks due to the multifaceted dimensions of resilience (Cutter et al., 2008) and

multiple abilities (Levine, 2014).

There are methodological challenges in measuring resilience too, which include:

the adequacy of indicators and how to measure them; and conceptual differences

further to the arbitrary definition of indicators (Levine, 2014). In practice, it has been

a challenge to collect data to inform the social resilience characteristics due to time

and resource constraints as well as the dynamic nature of characteristics that have

multiple inter-relationships, resulting in a complex network model (Ziyath et al.,

2013). These challenges are further compounded, as there is currently limited guidance

on what characteristics to measure, which indicators to use, and for what purpose and

context (Cutter, 2016). As a result, many social resilience frameworks commonly

capture easily accessible static indicators, while leaving out the dynamic and complex

key characteristics. Since it is difficult to directly measure many of the social resilience

characteristics, proxy measures are proposed/used which mostly are obtained from

secondary and census data information sources available in public databases.

Resilience frameworks that were operationalised to-date have used direct

methods such as household surveys and interviews, and/or data obtained from census

reports from public databases to assess resilience. Table 2.3 below highlights key

differences of two widely applied methods – using direct household surveys and


census measures to assess resilience. The merits in direct household measures are

demerits in using census data measures, and vice-versa, as listed in Table 2.3. For

example, direct measures such as household surveys are time and resource intensive

process, while resilience measures using census data can be faster. However, both

these methods have limitations in the replication of the process to regularly update the

resilience status. This is because census data is not frequently updated and frequent

replication of direct measures are also very difficult in most contexts due to limited

resources.

Table 2.3. Comparison of methods to assess resilience

Resilience measure using

household surveys

Resilience measure

using census data

Data sources Households Census data and

statistics

Method of

accessing

data

Web/Household survey, 1-1 or focus

group interviews, consultation, self-

assessment

Census reports and data

from public databases

Merits

More realistic at the time of

measurement or for a period of time

Minimal use of

resources and faster

Data reflects the perception of social

resilience in the community

Easy to develop indices

and statistical models

Can capture process based measures

and qualitative analysis is possible

Reproduced when

census is updated

Demerits

Time consuming, finance and human

resource intensive process

Lack of accuracy

compared to surveys

due to not capturing

key social dynamics

and processes

Replication and regular updates are

difficult in resource-limited contexts

Results can be easily

reproduced, replicated,

and improved only

when the new census

becomes available

Limited set of easily measurable

indicators are chosen

Mostly capture

outcome measures

Existing indicators to capture social resilience are inadequate, since resilience

characteristics such as social capital, social dynamics and interactions are hard to

measure in practice (Cutter, 2016). One common approach in social resilience

measurement is to use publicly available and easily accessible data sets such as census

data as proxy measures to formulate indicators (Cutter, 2016). The effectiveness of

this approach is highly dependent on the accuracy of the data set, and the ability to

methodically and accurately characterise social resilience indicators within a broader


framework (Levine, 2014; Ziyath et al., 2013). Therefore, another key knowledge gap

in social resilience measurement is the need for novel ways to capture key social

resilience characteristics with more conceptual clarity of their measurement.

Social resilience is also too complex to reduce it to a numeric value (Woolf et

al., 2016) and capturing dynamic social resilience characteristics using a static

measurement is not easily achievable (Schipper & Langston, 2015) for effective

decision making. Inadequate information about measurement tools and techniques and

lack of clear guidelines for measuring proposed resilience indicators in many

frameworks limit the operationalisation of the framework in different contexts

(Serfilippi & Ramnath, 2018). In the absence of robust guidelines accompanying the

set of indicators that are proposed to measure resilience, the possibility exists for the

misinterpretation of the indicator data, which may lead to inaccurate outcomes. Doorn

(2017) highlighted that the use of non-aggregated indicators can be useful for

contributing to detailed analysis of resilience than aggregated indices, for the latter are

mainly helpful in the evaluation of intervention effectiveness and drawing attention to

an issue. Resilience measurements also need to move beyond existing index based

quantifications to produce results that are oriented towards resilient development

outcomes and strategies to enhance resilience at community level (Peters et al., 2016).

Hence, new methods are needed for detailed analysis of resilience indicators that can

capture important facets of key social resilience characteristics, such as social capital,

social mechanisms, and social dynamics (Cutter, 2016), which are multi-faceted,

process oriented and abstract concepts.

2.2.4 Key knowledge gaps and recommendations for further research

The following three key knowledge gaps for future research on social resilience

in a disaster context are highlighted as shown in Figure 2.3, based on the critical

analysis of existing social resilience assessment frameworks.

Firstly, many challenges inherent in adequately measuring social resilience, as

highlighted in Section 2.2.3, necessitate a well-structured, rigorous, and adaptable

framework in the context of disaster management. Developing an integrated and

adaptable framework for consistent operationalisation remains an important research

gap in social resilience research. A comprehensive social resilience framework which

considers multiple dimensions of the social resilience concept is a pre-requisite for any

detailed assessment of resilience. However, the existing social resilience assessment


frameworks have largely been developed without consensus on the key characteristics

of social resilience and how to robustly measure them. The key question which remains

unanswered is the extent to which existing frameworks can be adapted for different

contexts, since many of them have been developed within a limited context, scale and

scope. Therefore, there is a need to explore the development of a generic framework

Figure 2.3. Summary of the literature review highlighting knowledge gaps in

social resilience assessment


that could guide the selection of social resilience indicators methodically to capture

multiple social resilience dimensions, and which can be adapted to different contexts.

Secondly, most social resilience frameworks have focused on outcome oriented

social resilience indicators, since they can easily produce social resilience indices.

These indices however, have limited value for interpreting the actual resilience status

of a community. In contrast, process-oriented indicators are based on dynamic

properties which are essential to comprehend social resilience adequately, have been

largely neglected. This is mainly because they are difficult to measure due to their

dynamic nature. Hence, new methods to capture key process-oriented social resilience

indicators that are critical for adequate measurement of social resilience are needed,

which is another key knowledge gap to be addressed in social resilience assessment

research. An adequate measurement of social resilience requires both, outcome and

process oriented social resilience characteristics that can capture social dynamics and

perceptions. Hence, future research needs to find innovative methods such as multi-

stakeholder engagement tools to capture the dynamic nature of social resilience.

Thirdly, conceptual frameworks should be translated into practice to help

decision-makers to operationalise social resilience measures (Abenayake et al., 2016).

Conceptual and measurement complexity and diversity have led to limited agreement

on a standard approach to operationalising the resilience concept in disaster

management (Cutter, 2016; Ostadtaghizadeh et al., 2015; Sharifi, 2016). A key

research gap still exists in relation to the assessment of the reliability and consistency

of resilience measures, when frameworks are operationalised in real-world

applications. A novel approach is needed to address the limitations in using existing

methods. Many existing social resilience frameworks developed in a disaster context

have applied selected measures from publicly available census data, since many of the

social resilience indicators require direct measures, which are time and resource

intensive. However, the use of census data from public databases to measure resilience

indicators does not depict resilience adequately or accurately since most often census

data is outdated and does not provide enough information about the dynamic nature of

resilience indicators. Further, the absence of a methodical process for identifying and

selecting the best measures from census data is limited. Hence, the use of surrogates

in lieu of actual parameters needs to be explored to address the challenges in measuring

resilience.


A ‘5S’ INCLUSIVE AND ADAPTIVE SOCIAL RESILIENCE

FRAMEWORK IN DISASTER MANAGEMENT

This section reports how a ‘5S’ inclusive and adaptive social resilience

framework was developed in this study by critically analysing existent social resilience

frameworks.

A resilience framework should be able to guide the selection of appropriate

social resilience characteristics and their adequate operationalisation based on the

context of its application (Parsons et al., 2016). In this study, an inclusive framework

was developed from the critical analysis of existent frameworks to systematically

select required key indicators from different social resilience dimensions. A set of key

process-oriented social resilience indicators that are difficult or not feasible to measure

through existing methods needs to be methodically selected to apply new methods of

measurement such as the surrogate approach. In this study, the ‘5S’ framework was

used for the purpose of selecting those indicators to develop surrogates that can cover

multiple key social resilience dimensions for adequate measurement of social

resilience.

2.3.1 ‘5S’ social resilience framework structure and components

In the ‘5S’ social resilience framework, key social resilience characteristics were

clustered in sub-dimensions to create a well-structured framework based on a three-

layer structure: key dimensions, resilience characteristics, and indicators, as shown in

Figure 2.4. The social resilience framework consists of five social dimensions (denoted

as ‘5S’ framework): social structure, social capital, social

mechanisms/competence/values, social equity and diversity, and social

beliefs/culture/faith (Figure 2.5). These were based on the most commonly used

categories of social resilience as identified in the research literature. By incorporating

key characteristics and indicators in five key dimensions, a ‘5S’ framework was

developed to measure social resilience, as shown in Figure 2.5.

Social resilience to disasters

Key

dimensions

Resilience

Characteristics

Process and

Outcome

Indicators

Figure 2.4. Three-layer structure of the ‘5S’ social resilience

framework


Figure 2.5. An inclusive and adaptive ’5S’ model social resilience

framework


The ‘5S’ framework is flexible and can be adapted to any geographical, hazard,

or community context by shifting the priority of the characteristics and indicators

according to the context. The proposed ‘5S’ model is inclusive in the sense that a

comprehensive set of key social resilience characteristics and indicators is

incorporated within a single structure. A brief explanation of each dimension and its

composition of social resilience characteristics is provided below.

Social structure:

Social structure is defined broadly to include a wide-range of social

characteristics, including network and relationships (Nadel, 2013). However, some

specific definitions of social structure are confined to define population distribution

and composition in a geographic space. The definition of social structure as the

distribution of population that includes parameters such as gender, ethnicity,

education, and income in multiple social layers (Blau, 1977) was adopted for this

study. The attributes of social structure are important to understand and differentiate

specific population and demographic parameters such as household income and age

distribution (Renschler et al., 2010) (See Figure 2.5 for key characteristics and

indicators). There are three characteristics and nine indicators in social structure. The

characteristics include: social demography, household structure, and mobility of

people and families.

Social capital:

Social capital is a dominant, highly influential, and widely studied aspect in

determining social resilience to disasters (Aldrich & Meyer, 2014a). Initial focus of

social capital on relationships in social structures and networks by Hanifian (1916),

Bourdieu (1986), and Coleman (1990) were advanced by Putnam (1993), to features

of social organisations such as networks, norms of reciprocity, and trust, that facilitate

action and cooperation for mutual benefit (Aldrich & Meyer, 2014a; Sanyal &

Routray, 2016). Later, social capital was classified into three types: bonding, bridging,

and linking capital (Aldrich & Meyer, 2014a). It can also be categorised as structural

and cognitive social capital (Sanyal & Routray, 2016). Social capital in this study

includes: social ties within community groups, mostly associated with family

relationships; and place of attachment (bonding); networking abilities made up of

economic and other ties that are external to the community (bridging); and the

interaction between social groups and community networks with the governing


authorities, state organisations, and non-state local institutions (linking) (Adger,

2003). Under the social capital dimension, there are three resilience characteristics and

nine indicators. Three characteristics are: social cohesion, social support, and social

networks (Figure 2.5).

Social mechanisms/competencies/values:

Social mechanisms include the process of developing community goals, plans,

priorities, and engagement of the community in the resilience building process. The

process oriented resilience characteristics also include community competence,

collective attitude, and shared values towards coping and adapting to disasters. Under

the social mechanism/competencies/values dimension, there are five social resilience

characteristics and 15 indicators. The resilience characteristics include: community

engagement, community goals/efficacy, community shared values and attributes,

community processes, and community competence (Figure 2.5) (Cutter et al., 2010;

Khalili et al., 2015; Paton et al., 2001).

Social equity and diversity:

Eliminating excessive burden on marginalised communities due to inequitable

distribution of critical resources, and increasing equity and social justice, are core

principles of social resilience (Plough et al., 2013). When a disaster strikes, people

who do not have access to equitable resources, such as families living below the

poverty line, may be affected significantly differently than other people within the

same community (Fothergill & Peek, 2004; Lovell & Le Masson, 2014). Social

resilience also depends on the diversity of resources, because communities that rely

on a limited range of resources often struggle to cope (Norris et al., 2008), and less

socially-diverse communities encounter greater difficulties to recover from

disturbances (Ahern, 2011). There are three characteristics and nine indicators in

social equity and diversity. The three characteristics include: fair access to basic needs

and services, community inclusiveness and equality, and diverse skill sets and

workforce (Figure 2.5).

Social beliefs/culture/faith:

Not only human resources and physical assets, but local culture and social

beliefs can play a critical role in determining social resilience to disasters (Kwok et

al., 2016). Social beliefs need to be positively capitalised in communities that are


oriented with their own local culture and faith systems (Kwok et al., 2016;

Ostadtaghizadeh et al., 2016). Social resilience frameworks developed in some

communities are strongly grounded in culture and faith. For example, Ostadtaghizadeh

et al. (2016) considered culture as a separate dimension of resilience. There are two

characteristics and four indicators in social beliefs/culture/faith. The characteristics

include: local cultural beliefs/norms and religious beliefs/norms (Figure 2.5).

SUMMARY AND IMPLICATIONS

The complex nature of the resilience concept, due to varied approaches has

further complicated the measurement of social resilience to disasters. Consequently,

the necessity of a consistent approach to measure social resilience to disasters has been

advocated by many researchers. The indicator approach has been broadly used in

recent research for measuring social resilience to disasters. However, social resilience

frameworks commonly capture easily accessible indicators while leaving out complex,

but key, resilience characteristics. Most of the proposed social resilience frameworks

have largely focused on outcome features of resilience rather than process related

indicators, mainly because process indicators are not easily measurable. A failure to

measure both, outcome and process features of resilience results in inadequate and

inaccurate assessment of social resilience. Hence, there is a need for developing a

social resilience framework that can integrate key process-oriented indicators for

adequate measurement of social resilience.

Further, due to the multi-faceted nature of resilience and complex

interconnections between social resilience characteristics, there is pronounced

difficulty in carrying out an effective and feasible resilience measurement within a

reasonable timeframe to achieve comprehensiveness and greater accuracy. Moreover,

the wide-ranging disparity of social resilience characteristics in different contexts

along with spatial and temporal variations, and complex inter-relationships among

these characteristics, challenge their direct measurability through a set of

comprehensive and definitive indicators. Therefore, new approaches should also

address the challenges in directly measuring social resilience due to their added

complexity.

Many outcome-oriented indicators of social resilience are primarily measured

using publicly available census data, which is not frequently updated. Hence, the


existing measures provide a limited and often not accurate assessment of social

resilience (Figure 2.3 shows the links to key knowledge gaps from the critical literature

review). A time and resource effective practical approach needs to be employed that

can use regularly updated data such as administrative data to measure social resilience.

One of the approaches to overcome these limitations is to use a surrogate approach, as

it can facilitate the identification of key facets of a resilience indicator and can use

regularly updated locally available administrative data sources to measure them.

Hence, a new method such as the surrogate approach to measure social resilience

needs to be tested to address these challenges in measuring social resilience to

disasters.

In order to methodically select key social resilience indicators from multiple

social resilience dimensions for developing a surrogate approach, an integrated

framework such as the ‘5S’ framework developed in this study can be used. The ‘5S’

framework is very generic and adaptable, and can guide the selection of key social

resilience indicators to develop a surrogate approach in any context.

Chapter 3: Surrogate approach to assess social resilience to disasters 34

Chapter 3: Surrogate approach to assess

social resilience to disasters

This chapter consists of three sections.

Section 3.1 provides an overview of the surrogate approach. This includes a brief

introduction to the surrogate approach and analogue models in different disciplines,

and an analysis of existing frameworks, to develop a surrogate framework to assess

social resilience in a disaster context.

Section 3.2 details the surrogate development framework for assessing social

resilience indicators in a disaster context in three key steps. These are aligned with

three phases in this research, which are detailed in Chapter 4 (Research methods). The

following three key steps are discussed briefly in this section: A. Selecting key

resilience indicators that require surrogate approach, B. Identifying potential

surrogates, and C. Selecting the optimum surrogates for application by evaluating and

ranking potential surrogates.

Section 3.3 summarises the surrogate approach and application in disaster

management.

Figure 3.1. Chapter 3 and key sections in the thesis structure


AN OVERVIEW OF SURROGATE APPROACH

3.1.1 Defining surrogates

The use of a surrogate approach is an innovative way forward to overcome the

challenges and limitations highlighted in measuring social resilience to disasters as

discussed in Chapter 2 (Cutter, 2016; Kulig et al., 2013; Sharifi, 2016; Ziyath et al.,

2013). Developing and translating surrogate measures to real world applications to

measure resilience will help in effective disaster planning and resource allocation.

The surrogate approach has been successfully used in multiple disciplines and

contexts, such as clinical medicine, biodiversity, and environmental studies (Barton et

al., 2015; Lindenmayer et al., 2015b; Mellin, 2011). In ecology, a surrogate is defined

as “an ecological element or process that is used to represent another aspect of an

ecological system” (Barton et al., 2015, p. 393). Lindenmayer et al. (2015b) defined it

as a “component of the system of concern that one can more easily measure or manage

than others, and that is used as an indicator of the attribute/trait/characteristic/quality

of that system” (p. 1030). Hence, the surrogate method is used as a proxy measure to

understand the environmental conditions (Lindenmayer & Likens, 2011).

Resilience surrogates are defined by Bennett et al. (2005) as proxies in assessing

resilience that can be selected from measurable attributes in a social–ecological

system. For the purpose of this study, a surrogate is considered as an indicator or set

of indicators that effectively represents another indicator that is intended to be

measured (Miguntanna et al., 2010; Rodrigues & Brooks, 2007). This method enables

easier and more cost-effective measurement than measuring the target indicator

directly.

3.1.2 The use of surrogate approach

Table 3.1 shows the application of the surrogate approach in different disciplines

such as clinical medicine, biodiversity, social-ecological systems, and water quality.

For example, due to limited resources to track the changes in ecological systems, the

application of surrogates has evolved over time as a necessary and a cost-effective

way to assess ecological processes and ecosystem responses (Lindenmayer et al.,

2015b). In the surrogate approach, understanding and selecting the most effective

surrogate(s) for the indicator of measure is very important (Rodrigues & Brooks,

2007), because it will help to understand when or under which conditions the


surrogates can be best utilised (O’Loughlin et al., 2018a). For example, in clinical

medicine, cholesterol level as a quantitative measure is used for inferring a patient’s

health status and risk of a disease (Barton et al., 2015).

Surrogates are useful for several reasons. For example, the time required to

measure the target system can be much shorter with the surrogate, which is easier to

measure than the target system, and the sample size or the extent of the measurement

entity is much lesser for surrogates than for the target system (Asher et al., 2015;

Baker, 2005). Collectively, the application of surrogates, as shown in Table 3.1, is

useful to contextualise the rationale behind the surrogate approach. The surrogate

concept can be transferred to measure social resilience in a disaster context to

overcome existing challenges of measuring social resilience assessment methods

(Levine, 2014; Saja et al., 2019), as highlighted in Chapter 2.

In the surrogate approach, the surrogacy relationship between the surrogate and

the target indicator is important. Surrogacy is defined as the extent to which a

particular set of features (surrogates) effectively represents another set of features to

be measured (target) (Rodrigues & Brooks, 2007). Further, “a surrogate should be

consistent and repeatable, in the sense that independent observers given the same

information would assess the surrogate in the same way” (Carpenter et al., 2005, p.

942). The conceptual analogues of surrogate models applied in other disciplines are

discussed in the next section.

Table 3.1. Some examples of surrogate approach in other disciplines

Area of

research Examples of surrogate approach Reference

Water quality

surrogates

Turbidity as a surrogate for water quality

parameters

Schilling et al.

(2017)

Public health

An organism, particle, or substance as a

surrogate to study the fate of a pathogen in a

specific environment to improve public health

(Sinclair et al.,

2012)

Biodiversity

conservation

planning

Conservation planning based on a set of

biodiversity features as surrogates (using

species accumulation index)

Rodrigues and

Brooks (2007)

(Lindenmayer et

al., 2015a)

Clinical

medicine

Surrogates (blood pressure) used to predict the

clinical endpoints (example stroke risk).

Biomarkers as substitute for clinical endpoints.

Aronson (2005)

(Fleming &

Powers, 2012)


Resilience for

socio-

ecological

systems

Observable features and models of socio-

ecological systems (using stakeholder

assessment, model explorations, historical

profiling, and case study comparison)

Carpenter et al.

(2005)

Resilience surrogates through development of

system models (systems are qualitative and the

relationship is quantified)

Bennett et al.

(2005)

Conceptual analogues for surrogate model in social resilience

This section explores how the surrogate approach is applied in other disciplines,

so as to conceptualise the surrogate approach to measure social resilience indicators

in a disaster context. There is evidence to suggest that existing social resilience

measurement methods have similar conceptual and methodological challenges in

ecology and clinical medicine.

For example, a systematic surrogate approach has been used in clinical medical

sciences for many decades (Barton et al., 2015). In clinical medicine, the causal

framework is built on the basic naïve, general, and composite models, with increasing

complexity of linkages, as shown in Figure 3.2. The naïve model links the effect of

the treatment (T) to the outcome (O), the treatment to the surrogate (S), and the

surrogate to the outcome. In comparison, the general model incorporates another

source of variability (U) on the surrogate and response variables. The most complex

type is the composite model that links additional covariates (denoted by X and L) and

their interrelated effects on the surrogate and response variables (Barton et al., 2015).

Similar models can be drawn for social resilience indicators and potential surrogates.

However, it is necessary to start with the basic naïve model in the social resilience

context, since the surrogate research has not been tested before to develop complex

composite models.

Figure 3.2. Graphical representation of surrogate approach models

reproduced from Barton et al. (2015, p. 394)


Barton et al. (2015) have applied the surrogate concept from clinical medicine

to ecology, and provided an ecological analogy for the surrogate model used in

medical sciences. Figure 3.3 highlights an example of conceptual analogous surrogate

models in clinical medicine, ecology, and in social resilience to disasters adapted from

Barton et al. (2015). For example, in clinical medicine, the stroke risk can be predicted

by measuring elevated blood pressure as a surrogate measure. In ecology, the desired

environmental state can be measured using lichen instead of measuring environmental

pollutants. Similarly, in the social resilience literature, social cohesion can be

measured using legal cases/complaints against neighbours or community members as

a potential surrogate instead of measuring social trust, which is an abstract and multi-

faceted social resilience indicator. The data can be accessed from regularly updated

data sources, such as administrative records available from local authorities or legal

enforcement authorities.

The use of a surrogate approach to measure social resilience can potentially

resolve the measurement issues in disaster resilience (Kulig et al., 2013). The surrogate

approach can be applied to social resilience measurement, similar to the use of the

surrogate approach to measure resilience in ecology. A robust surrogate development

framework is needed that can be used systematically to conceptualise, identify, and

evaluate surrogates in a disaster context, so as to overcome current conceptual

complexity and methodical constraint in measuring social resilience. The surrogate

Figure 3.3. Conceptual analogous surrogate model examples in clinical

medicine, ecology, and social resilience to disasters; adapted from Barton

et al. (2015) (p. 394)


thinking applied in ecology and recent evolution of surrogate frameworks that are

flexible for application in a multidisciplinary context will be used to guide this process,

as detailed in the next section.

A number of limitations were also identified in the application of surrogate

approach in other fields. For example, generalising the surrogates in different contexts

is one of the key challenges, which needs cautious interpretation of surrogate

relationship to the target (Grantham et al., 2010). However, O’Loughlin et al. (2018a)

highlighted that considerable variability and context dependency always exist in

environmental science and the application of similar surrogates in different contexts

can be overcome by adapting the framework through rigorous evaluation and

appropriate validation techniques. When a new surrogate is identified, it is necessary

to calibrate it with the existing surrogate for its robust application (Lindenmayer et al.,

2015b), which requires active learning and improvements in surrogate application

through continuous engagement of key stakeholders.

Another challenge in surrogate approach is the regular update of surrogate

information over time (Tulloch et al., 2016). Hence, the final selection of surrogate(s)

for real world application depends on the surrogate information sources that are

regularly updated, such as local administrative data. Potential surrogates can be

explored to measure sets of social resilience indicators by consulting key disaster

management stakeholders as a one-off exercise. Once the final set of surrogates has

been selected, they can be periodically updated in consultation with the relevant

agencies at the local level to capture any changes, so that resilience can be robustly

measured without repeatedly undertaking resource- and time-intensive processes.

3.1.3 Review of existing frameworks to develop a surrogate development

framework

There are a wide range of surrogate approaches proposed and used in

environmental science, such as Rodrigues and Brooks (2007), Carpenter et al. (2005),

Berkes and Seixas (2005), Bennett et al. (2005), Lindenmayer and Likens (2011), and

Lindenmayer et al. (2015b). Most of these approaches are very specific to the field of

study, as previously detailed in Table 3.1, and are very difficult to apply directly in a

disaster context. For example, Berkes and Seixas (2005) used case studies to identify

key factors influencing lagoon systems, and Bennett et al. (2005) used system models

to extract potential surrogates. These frameworks do not provide a common method


for developing a surrogate approach that can be easily adapted for other disciplines.

Two frameworks were selected for detailed review to develop a suitable surrogate

development framework to use in a disaster context.

(1) An adaptive surrogate framework by Lindenmayer et al. (2015b)

The first framework reviewed is the adaptive surrogate framework proposed for

the identification, evaluation, and application of environmental surrogates by

Lindenmayer et al. (2015b, p. 1033) (See Figure 3.4). This framework is a

comprehensive and adaptable framework compared to other surrogate frameworks,

which are mostly specific to the discipline. This framework was proposed based on

key lessons from applying varying methods of surrogate approach. It was also an

outcome of a collaborative and consultative process with many researchers who

developed surrogate approaches in environmental science (Lindenmayer et al.,

2015b).

The purpose of Lindenmayer et al. (2015b) framework is to guide methodical

identification, evaluation and selection of robust surrogates for application to measure

and monitor target ecological systems. The generic structure of the framework, which

Figure 3.4. Adaptive Surrogacy Framework

Lindenmayer et al. (2015b, p. 1033)


organises the key steps of surrogate development in a logical sequence for easy

adaptation to other disciplines, taking into account the iterative nature of surrogate

development, are the key strengths of the Lindenmayer et al. (2015b) framework

compared to other surrogate frameworks developed specific to a discipline. However,

the structuring of key steps 1 to 3 more consistently with key steps 4 and 5 can help to

adapt it easily to any context. Further, the Lindenmayer et al. (2015b) framework was

developed in the context of advancing the surrogate approach in ecology by unifying

different types of surrogate applications. Hence, this framework requires adaptation to

develop surrogates in new disciplines such as resilience measurement in disaster

management, which has not applied a surrogate approach to date.

Lindenmayer et al. (2015b) framework consists of eight key steps in developing

environmental surrogates:

1. Identify objectives: Firstly, it is important to define an objective that needs to be

addressed with respect to measuring an ecological system that is difficult to

measure directly.

2. Engage stakeholders: Identifying and engaging stakeholders in all phases of

surrogate development is important to capitalise on collective and diverse

perspectives.

3. Develop conceptual model for the target system: An important step in identifying

effective surrogates is to develop a good model of the target system with key

ecological processes.

4. Identification of surrogates: The identification of potential surrogates for the

selected ecological processes that are difficult to measure directly includes

designing a sampling strategy and key benchmarks or trigger points for

interventions.

5. Evaluation of surrogates: Evaluation of potential surrogates identified in the

previous step is critical. This includes: a) asserting the scientific validity of

surrogates; b) comparison of costs and benefits of surrogate measurement; and c)

undertaking a risk assessment.

6. Selection of surrogates: The iterative process between identification and evaluation

of surrogates will finally result in the selection of optimum surrogates that meets

the objective identified in the first step.


7. Application of surrogates: The selected surrogates need to be applied in real-world

scenarios.

8. Active learning: Active learning is a continuous process throughout all the key steps

for improving the surrogate development process.

Though all the key steps are relevant for developing surrogates in a disaster

context, some necessary adaptations in key steps need to be considered. Further, some

of the sub-steps proposed within very critical key steps such as the identification and

evaluation of surrogates (Key steps 4 and 5) need a detailed review to adapt them

appropriately in a disaster context. In order to see the relevance of these steps in a

disaster context, an indicator development framework proposed for vulnerability

assessment in disaster management was also selected for review in this study. A

comparison of the logical sequence and relevance of both frameworks can provide a

good understanding to structure surrogate development framework in a disaster

context. In the next section, review of a vulnerability indicator development

framework of Birkmann (2013) used in a disaster context is explained.

(2) Vulnerability indicator development framework by Birkmann (2013)

The second framework reviewed in this study is the vulnerability indicator

development framework, which is a concept that is closer to resilience in disaster

management. In the disaster management literature, Birkmann (2013) proposed nine

steps for developing vulnerability indicators, as shown Figure 3.5.

Figure 3.5. Vulnerability indicator development framework

Birkmann (2013, p. 94)


1. Defining goals: Indicator development process needs to start with defining or

selecting the relevant goal.

2. Scoping: Once the goals are defined, it is important to clarify the scope of indicator

by identifying the context, including the target groups and boundaries of

application.

3. Choose indictor framework: This involves identification of conceptual target

system and structuring the themes for the target system.

4. Define selection criteria: Next step involves selecting a set of general criteria for

identifying good-quality indicators.

5. Identify potential indicators: This is the key step, where a set of potential indicators

to measure the target system needs to be identified.

6. Choose a final set of indicators: Based on the criteria selected in step 4, potential

indicators need to be evaluated for selecting the final set of indicators for

application.

7. Analyse indicator results: Data collection for the indicator will be done in this step

to validate the applicability of the indicator results.

8. Prepare and present report: A final report will be prepared based on the analysis

results in step 7.

9. Assess indicators performance: The final step of indicator development is to assess

the performance of the indicator selected and applied in real context to further

improve its performance in the next step.

The comparison of key steps in both frameworks helped to select the key steps

required in a surrogate development framework for assessing social resilience to

disasters.

Comparison of key steps in both frameworks for adapting them to disaster context

A comparison of key steps in both frameworks is listed in Table 3.2, which

shows little variations in most of the key steps in both frameworks, although both are

used in different contexts. Some minor differences exist; for example, use of criteria

trade-offs for environmental surrogate evaluation as an integrated process within the

proposed framework by Lindenmayer et al. (2015b), as opposed to a separate key step


– ‘selection criteria’ by Birkmann (2013) to evaluate social vulnerability indicators

(Key step 4). Further, the final three steps (step 7, 8 and 9) in the vulnerability indicator

development framework (Figure 3.5) are: analysis of indicator results; preparation of

report; and assessment of indicator performance. In the Lindenmayer et al. (2015b)

adaptive surrogacy framework, these steps are not included, but general final steps

(step 7 and 8) are included as ‘application’ and ‘active learning’.

Table 3.2. Comparison of key steps in indicator/surrogate development frameworks

Key steps in vulnerability indicator

development (Birkmann, 2013, p. 94)

Key steps in surrogate development

(Lindenmayer et al., 2015b, p. 1033)

1 Defining goals 1 Identify objectives

2 Clarifying the scope by identifying

the target group and purpose 2 Identify and engage stakeholders

3 Identifying an appropriate

conceptual framework 3

Develop a conceptual model for the

target system

4 Defining selection criteria - Trade-offs integrated with key steps

5 Identification of potential indicators 4

Identification of potential

surrogates (benchmarks and

sampling approach)

6 Evaluation and selection of each

indicator using evaluation criteria 5

Evaluation of surrogates (scientific

validity, cost effectiveness, risk

assessment)

7 Analyse indicator results 6 Selection of surrogates

8 Prepare and present report 7 Application of selected surrogates

9 Assess indicator performance 8 Active learning throughout the

process and improvement

This comparison shows that key steps for the social resilience surrogate

framework can be developed from the adaptive surrogate framework by Lindenmayer

et al. (2015b) with proper contextualisation for use in a disaster context. Therefore,

the general key steps in surrogate development in a disaster context can include:

selection of resilience indicator to develop surrogates; identification of potential

surrogates; and evaluation of surrogates to select the best performing surrogate for

application.


The method of using the criteria for evaluation of potential surrogates in a

disaster context needs to be decided, since it is used in different ways depending on

the context. Defining criteria and evaluation were taken as separate steps (key steps 4

and 6) in the framework by Birkmann (2013), whereas the criteria trade-offs are

integrated in the Lindenmayer et al. (2015b) framework, since they are not explicitly

considered in surrogate evaluation.

The key steps 4 and 5 of Lindenmayer et al. (2015b) are the most critical in

surrogate development. Hence, some of the sub-steps proposed in two key steps (Key

step 4: Identification of potential surrogates and Key step 5: Evaluation of potential

surrogates) in the adaptive surrogate framework proposed by Lindenmayer et al.

(2015b) need further consideration when adapting them, to ensure compatibility with

the most relevant components for resilience measurement in a disaster context. Each

of the sub-steps that are adapted to formulate a surrogate development framework in

a disaster context is discussed below.

Key step 4 – Identification in Lindenmayer et al. (2015b) framework (Figure 3.4):

(1) Sub-step - Benchmarks/triggers (In key step 4 of Figure 3.4): In an ecological

context, benchmarking of surrogates is required to measure the target indicator, which

depends on the understanding of the study system (Lindenmayer et al., 2015b). The

measure of resilience is either based on the existing status (inherent resilience), in a

pre-disaster phase, or how people adapted in new efforts to recover from the disaster

impacts in the post-disaster phase (adaptive resilience) (Cutter, 2016). The baseline

status of social resilience can be measured by the potential surrogates, as proposed in

this research. However, the dynamic status of the baseline, which frequently changes

its status over time due to changing complex social systems, is an enormous challenge

to determine the baseline condition and benchmarking (Brown et al., 2018). Therefore,

in the disaster context, the surrogate measures can help to monitor the progress of

resilience status (Cutter et al., 2010), instead of benchmarking social resilience to a

threshold limit, which is often challenging. Hence, the understanding of the surrogacy

relationship between the surrogates and the target indicator is proposed to be included

as a sub-step, as it allows the establishment of a good measure of a target indicator

using existing resilience conditions through the surrogate approach.


(2) Sub-step - Sampling approach (In key step 4 of Figure 3.4): The measurement of

ecological surrogates can be done by a sampling approach, since surrogates are also

mostly measured through empirical methods. However, the sampling approach at the

initial stage of identifying social resilience surrogates can be difficult without a proper

understanding of surrogate measurement protocols. Even in developing resilience

surrogates in ecology, Carpenter et al. (2005) highlighted four different methods to

identify surrogates: stakeholder assessment; model development; historical profiling;

and case study comparisons. To decide a method for identification of surrogates for

social resilience measurement in a disaster context, it is most appropriate to be flexible

when deciding on an effective measurement method, since a surrogate approach for

resilience assessment in disaster management is still at the exploration stage. Hence,

the need for a sampling approach can be decided based on the proposed surrogate

measurement protocols. Hence, inclusion of surrogate measurement protocols as

another sub-step in the identification of potential surrogates is more appropriate to

measure social resilience using a surrogate approach in a disaster context.

Key step 5 – Evaluation in Lindenmayer et al. (2015b) framework (Figure 3.4):

(3) Sub-step - Scientific validity (In key step 5 of Figure 3.4): The evaluation of

scientific validity for environmental surrogates is a process using methods such as

experiment and observation (Lindenmayer et al., 2015b). However, the validity and

applicability of social resilience surrogates in a disaster context are either driven by

expert judgement or through locally driven approaches at the community levels (Tyler

et al., 2016). The initial identification of potential surrogates in this study was

proposed to be done through locally driven approaches, such as consulting disaster

management practitioners and policy makers working at the community and sub-

national levels. Hence, an evaluation of potential surrogates is proposed to be carried

out against a set of surrogate evaluation criteria such as accuracy and communicability,

with consultation with a wide range of disaster management experts from a variety of

cohorts (such as researchers, policy makers, and practitioners) to increase their

robustness and validity for application.

(4) Sub-step - Risk assessment (In key step 5 of Figure 3.4): A risk assessment for

identified ecological surrogates may be necessary, since the negative consequences of

applying the wrong surrogate for intervention decisions will be high (Lindenmayer et

al., 2015b). In a community disaster resilience measurement context, interventions and


investment decisions are necessary for most, if not all, of the potential surrogates based

on their baseline status. However, in reality, there is a need to prioritise resilience

interventions, due to resource constraints such as human and financial. Many

researchers such as Alshehri et al. (2015b), Carone et al. (2018), and Orencio and Fujii

(2013) ranked and prioritised resilience indicators using multi-criteria decision

making tools in a disaster context. Therefore, it is more appropriate to apply similar

approaches to rank potential surrogates to prioritise resilience building interventions.

Further, the selection of surrogates (key step 6 in Lindenmayer et al. (2015b)

framework) is the end result of the surrogate evaluation process (key step 5 in

Lindenmayer et al. (2015b) framework). Therefore, it is more appropriate to include

‘selection of surrogates’ as part of the key step – ‘evaluation of surrogates’ for

resilience assessment in a disaster context.

A surrogate development framework to measure social resilience to disasters

was created by considering the necessary adaptations to the Lindenmayer et al.

(2015b) framework, as explained above, for its application in disaster management.

From this analysis, a surrogate development framework was proposed for testing, to

assess social resilience in a disaster context, which is discussed in the next section.

A SURROGATE DEVELOPMENT FRAMEWORK FOR ASSESSING

SOCIAL RESILIENCE TO DISASTERS

The proposed surrogate development framework to measure social resilience to

disasters is shown in Figure 3.6. The proposed social resilience surrogate development

framework has three components, which will form the three key steps in the surrogate

development process (see Figure 3.6):

Key step A. Selection of social resilience indicators for surrogate development

Key step B. Identification of potential surrogates for the selected key indicators

Key step C. Selection of optimum surrogates for application in a disaster context

Stakeholder engagement and continuous learning/improvement spans across all

three key steps of the framework. This is because consultation with key stakeholders

is important as a way to obtain the necessary and on-going inputs in each respective

key step (A to C). Further, it is important to continuously improve the framework from

the lessons and new learnings when they are applied in different disaster contexts. The

details of key steps (A to C) and their sub-steps are discussed in the next sections.


Figure 3.6. Surrogate development framework to assess social resilience in a disaster context


3.2.1 Selection of resilience indicators for surrogate development (Key step A)

In key step A, two sub-steps are proposed to guide the logical decision making

for surrogate approach: (a) identification of key social resilience indicators; and (b)

the decision for surrogate approach.

A.1. Identification of key social resilience indicators

There are numerous and wide-ranging social resilience indicators. For example,

Saja et al. (2018) identified 46 key social resilience indicators in five dimensions of

resilience. However, in a particular context, certain key indicators are more important

than others, and have to be prioritised, due to resource and time limitations in

measuring all the key indicators. A feasible number of key social resilience indicators

needs to be initially agreed. This can be done either through consultation with the key

stakeholders, or through a review of available information sources on social resilience

measurement before initiating the surrogate identification process. In this study, a set

of key social resilience indicators to apply a surrogate approach was identified through

a review of literature on social resilience indicators, and is detailed in Section 4.3.

A.2. Decision for surrogate approach

As shown in Figure 3.7, the key reasons guiding the choice of surrogate

approach include:

a) When the process of existing measures for assessing a target indicator through direct

measurement is time and resource intensive (in financial and human terms)

(Lindenmayer & Likens, 2011); and/or

b) When the existing indirect measures use publicly available census data, which

mostly produce index based quantitative measures that are limited with good

qualitative interpretations to provide a good enough resilience measurement

(O’Loughlin et al., 2018b).

It was found in the review of existing social resilience measures that the two

reasons explained above are mostly true for process-oriented resilience indicators.

Further, outcome-oriented resilience indicators often provide limited measures, since

they use census data, which is not frequently updated. In such situations, outcome-

oriented resilience indicators can also be measured by using surrogates with the help

of most frequently updated administrative data.


The context-specific nature of resilience means that the measurement of

resilience is bound by temporal and spatial scales (Mitchell & Harris, 2012). Hence,

the surrogates identified in a disaster context should be applicable in different

geographic contexts and in different disaster phases. Once the surrogate approach to

measure social resilience to disasters is conceptualised, the objective of surrogate

measurement can be defined by setting the boundaries/scope, such as the geographical

and disaster context, within which the identified surrogates will be applicable. Once

the decision for a surrogate approach is made, the next key step (key step B in the

surrogate development framework in Figure 3.6) is to identify potential surrogates to

measure the selected social resilience indicators.

3.2.2 Identification of potential surrogates for selected indicators (Key step B)

Key step B is focused on the identification of potential surrogates for the selected

key indicators. This is an important step in the surrogate development framework,

because identifying as many potential surrogates as possible at the initial stage will

help to select the best surrogates through evaluation. It is important to check the

relationship between the surrogate and the target indicator and assess how stronger

and sensitive this relationship is, in order to ensure the initial validity of all potential

surrogates. Key step B has the following three sub-steps (as shown in Figure 3.8) and

each of them is explained below.

Figure 3.7. Surrogate approach decision flowchart (Key step A)


Figure 3.8. A framework to identify potential surrogates for social resilience indicators in a disaster context (Key step B)


B.1. Explore all potential surrogates,

B.2. Establish the surrogacy relationship, and

B.3. Determine protocols for surrogate measurement.

B.1. Exploring all potential surrogates

There are three elements when exploring all potential surrogates:

(1) Selection of a context:

Initially, a context for exploring surrogates needs to be established. Establishing

context includes selecting a site (geographical context), which includes urban/rural, or

coastal/inland/mountain area, as measures will differ depending on geographic

settings, and type of disasters (single or multi hazard). Resilience measures are mostly

context specific, which often vary with different types of disasters and the target

group/entity (local authority level, sub-national, or national), although there are many

common indicators and surrogates that can be identified (Ostadtaghizadeh et al., 2015;

Sharifi, 2016). Detailed explanation of the selection of case study context in this

research is explained in Section 4.5.

(2) Sampling of experts in disaster management:

Experts who are local policy makers and practitioners working in disaster risk

reduction can be purposefully sampled for data collection. For example, the experts

should have a minimum of three years of experience in disaster management projects.

Many bottom-up and locally driven frameworks developed to measure indicators

highlight the need for consulting key stakeholders relevant to the study, in order to

obtain good insights into the phenomenon being explored (Reed et al., 2006). This will

assist in making the frameworks and the measures widely applicable in practice.

Detailed explanation of the sampling method is given in Section 4.5.

(3) Consult experts for data collection:

Once the context is set and the experts are selected, an appropriate data

collection method to consult experts needs to be decided. Many common expert

consultation methods in a disaster management context, such as through key informant

interviews, interest group discussions, and participatory workshops, can be done in the

selected study locations (Keating et al., 2017). In this research, one-on-one interviews

and group interviews were used in the selected case study areas, and are explained in

Section 4.5.


B.2. Establish surrogacy relationship

In the next step, once all the potential surrogates are identified, the surrogacy

relationship for each potential surrogate with the target indicator needs to be

established. There are three key elements in this sub-step B.2:

(1) How surrogate relates to target indicator (surrogacy relationship):

Surrogacy relationship is defined as the extent to which a particular set of

features (surrogates) effectively represents another set of features of the target

indicators (Rodrigues & Brooks, 2007). This relationship between target indicator and

the proposed surrogate needs to be established from discussions with the experts.

Experts can be requested to provide examples for the proposed surrogates to establish

their relationship with the target indicator. A descriptive explanation can be given by

providing the quotes from the interview transcripts or from discussion records and to

link it with the existing literature or secondary data sources such as reports.

(2) Strength of the surrogacy relationship:

The quality and effectiveness of surrogate(s) depend on the strength of the

surrogacy relationship between the target indicator and surrogates (Grayson et al.,

1996). The higher the strength of this association between the surrogate and the target

indicator, the lesser the likelihood of misinterpreting the inference of the target

indicator (Lindenmayer et al., 2015a). Due to the multifaceted and dynamic nature of

social resilience, the most suitable method to understand the surrogacy relationship is

to use open-ended questions to explain the surrogacy relationship. Some challenges in

quantifying resilience, such as a lack of good explanation, also exist in quantifying the

surrogacy relationship. Therefore, interview participants can be requested to establish

the link between each surrogate and the target indicator , so that rich qualitative data

can be collected to overcome the limitations in quantifying the surrogate relationship

(Jones & Tanner, 2017).

(3) Sensitivity of the surrogacy relationship:

Measuring social resilience indicators in a disaster context through potential

surrogates can become complex due to multifaceted linkages between multiple

surrogates. The relationship of a potential surrogate with the target indicator is

complex in nature, such as; one surrogate is connected to the target indicator or to

other surrogates, and how changes in the surrogacy relationship may impact the target

indicator (Lane et al., 2015). It may become difficult to measure the target indicator


with the proposed surrogate, if there is a complex relationship between the identified

surrogates. Therefore, the aim should be to identify a unique set of key facets of the

target indicator which can be used as surrogates, so that the inter-linkages between

surrogates are minimal and do not make their measurement complex.

B.3. Protocol for measuring surrogates

One of the key justifications for the use of a surrogate approach is to enable the

use of frequently updated data sources so as to overcome the limitations in direct

measurement or publicly available census data sources. Therefore, the protocol for

measuring potential surrogates needs to be identified. Hence, the protocol to guide the

measurement of surrogates involve the following three elements:

1) Type of measurement:

The new approaches to assess and monitor resilience at regular time intervals

increasingly recognise and recommend a mix of qualitative and quantitative

approaches. However, more emphasis is given to provide enough evidence with the

existing data to measure the surrogates (Schipper & Langston, 2015). Therefore,

provision of sufficient qualitative interpretations of the existing status of the surrogate

will assist in translating the surrogate measure into resilience enhancement activities

for practical implementation. This has to be decided through expert consultation and

past experience.

2) Method for accessing accurate data:

The data collection method for a surrogate approach should be easier than that

used for direct measurement of resilience, such as household surveys, since they are a

resource and time intensive process and should be from sources that are more

frequently updated than census data. Therefore, the method of accessing available data

can include: primary data collection methods using a sampling approach from the

existing projects in the selected regions, or secondary data such as the regularly

updated administrative data from local state departments and authorities; a

combination of both can be explored for measuring resilience through surrogates

(Tyler et al., 2016).

3) Reliable sources for surrogate data:

In most of the local government and administrative bodies, the most frequently

updated administrative data is available, since it used by the authorities to monitor


their project goals and progress (Tyler et al., 2016). Therefore, it is necessary to source

data from non-traditional sources (for example administrative or project data), when

the publicly available census data is outdated, which has been the case in most

developing countries.

After the completion of key step B in the surrogate development process, which

includes three sub-steps as explained above, the next step (Key step C) is to evaluate

identified potential surrogates and rank them to select the optimum set of surrogates

for application.

3.2.3 Selection of optimum surrogates for application (Key step C)

Key step C involves the selection of the optimum surrogates for application, and

is the final step in the surrogate approach (as shown in Figure 3.6). The surrogates

must be evaluated independently against a set of criteria to determine optimum

surrogates that perform well in all criteria (Lindenmayer et al., 2015b). In the next

step, the ranking of surrogates can be done using multi-criteria decision making.

Finally, a set of surrogates can be selected based on priority rankings and contextual

requirements to achieve the final objective of assessing social resilience.

In this research study, the key step C of the surrogate development framework

(Figure 3.6) was extended to include all required elements for testing, as shown in

Figure 3.9. It has three distinct sub-steps: C.1. Evaluation of potential surrogates

against key surrogate evaluation criteria; C.2. Ranking of surrogates; and C.3.

Selection of optimum surrogates for application. These are detailed below.

C.1. Evaluation of potential surrogates

When the identification of potential surrogates for the selected key social

resilience indicators is completed in key step B, these have to be evaluated against a

set of key criteria to select the surrogates that have optimum performance in all criteria.

This sub-step has three components as shown in Figure 3.9: (1) Five surrogate

evaluation criteria; (2) Selection of experts for judgement; and (3) Design of

measurement (evaluation scale). A brief explanation of the above three components is

provided below. A detailed explanation of the research method for this sub-step is

detailed in Section 4.6.


Figure 3.9. Sub-steps to select optimum surrogates to assess social resilience indicators in a disaster context (key step C)


(1) Five surrogate evaluation criteria

The identification of all potential surrogates in key step B through a consultative

process with disaster management practitioners and policy makers will ensure their

initial validity for application in practice. In this step, the robustness of surrogates for

practical application can be judged by consulting higher level experts working in

disaster management at national and international levels. Some key criteria for

selecting good surrogates include accuracy, sensitive to the change, cost-effective,

easy to communicate, and capable of being updated regularly (Birkmann, 2006;

Custance & Hillier, 1998). Each surrogate should be assessed against each criterion

independently. The five selected surrogate evaluation criteria are discussed below in

detail.

Development of surrogates need to be guided by a set of key qualifying criteria

for the selection of rigorous surrogates to measure social resilience effectively. Five

key evaluation criteria were established from the set of surrogate selection trade-offs

proposed by Lindenmayer et al. (2015b), in evaluating the environmental surrogates

and standard indicator development criteria used in the vulnerability assessment

context by (Birkmann, 2013). The key criteria in trade-offs proposed by Lindenmayer

et al. (2015b) include: accuracy, communicability, cost-effectiveness, and time

sensitivity. These are similar to the criteria used in a disaster vulnerability assessment

context. For example, good surrogates should be able to easily measure the target

indicator, be sensitive to different time periods, easily understandable to users, and

cost-effective for collecting data (Maclaren, 1996). Hence, the following five key

criteria to evaluate surrogates in measuring the target social resilience indicators were

established by drawing from the surrogate and disaster management literature as

highlighted above (Figure 3.10- Surrogate evaluation criteria pentagon).

1. Accuracy of the surrogate: The link between the indicator and the target is an

important criterion for making a decision on effective indicators (Mitchell, 2013).

Further, the precision of a measurement implies the level of credibility of the indicator.

A surrogate is credible when it has high accuracy in predicting the target indicator,

and it should be able to provide appropriate information on the target indicator

intended to be measured.


Hence, the level of accuracy between the surrogate and the target resilience indicator

is an important criterion in selecting best surrogates. Experts can determine accuracy

by understanding the perceived strength of the surrogate with the target indicator.

2. Cost-effectiveness of the surrogate: The cost of measuring an indicator is typically

related to time, personnel, and logistics costs associated with data collection,

processing, and analysis (FAO, 2008). These costs may vary significantly based on

the target indicator of measure and data collection method. Often, the use of low-cost

indicators may imply difficult trade-offs in terms of their accuracy and credibility,

which needs to be considered in selecting indicators. Calculating the cost of collecting

any given indicator is relatively straightforward, but the benefits associated with that

additional piece of information may be difficult to define and quantify (FAO, 2008).

Hence, each surrogate needs to be examined for cost-effectiveness, and they can be

based on readily available data or be available at a reasonable cost through a brief

consultative process with key stakeholders.

3. Time sensitivity: It is necessary to clearly identify temporal variations of resilience

surrogates in a disaster context. An effective indicator should be able to measure the

progress or trend in different time periods (Mitchell, 2013) and be responsive to

change. In a disaster context, a surrogate should be able to respond in all three key

phases (preparedness, response, and recovery), so that the resilience characteristic is

consistently detectable, without losing its robustness. Surrogates should be capable of

being updated at regular time intervals (Donnelly et al., 2007), particularly in the

preparedness phase so that they can be monitored periodically. Time sensitivity of the

selected surrogate is therefore another criterion for surrogate evaluation.

Figure 3.10. Surrogate evaluation criteria pentagon


4. Measurement complexity: Availability of data and accessibility of the data are

important factors to consider when selecting any indicator (EC, 2015). Availability of

accessible data so that they become easily applicable in practice is often emphasized

by practitioners (Birkmann, 2006). When a surrogate is selected, it is important to

understand what data is available to measure the surrogate indicator and can be easily

obtained to predict the target resilience indicator. It is important to find the right

balance between the accuracy of data and the limited availability of data (Birkmann,

2006).

5. Communicability: Any indicator should be able to communicate the required data

to a wide range of stakeholders (practitioners, policy makers, and researchers) engaged

in measuring social resilience. An indicator should be simple to communicate while

being scientifically sound and valid (Mega and Pedersen, 1998). The task of any

indicator is also to relay complex information in an accurate and understandable

manner so that decision makers can make informed decisions (Donnelly et al., 2007).

(2) Selection of experts for judgment:

A number of experts who have experience in disaster management projects and

research can be approached to evaluate the potential surrogates against the five

surrogate evaluation criteria. To increase the validity of evaluation results, a wide

range of experts from different cohorts of the disaster management sector, such as

practitioners, policy makers, and researchers, can be invited to provide their

judgement.

(3) Design of measurement:

The measurement to evaluate potential surrogates is a five-point Likert scale that

will denote the degree of agreement from very high to very low. For each of the social

resilience indicators, experts will be requested to provide their judgement on the Likert

scale as to how they think each surrogate will perform against each of the five criteria.

The ordinal qualitative scale used for rating can be converted for quantitative analysis

in multi-criteria decision making to rank the potential surrogates.

C.2. Ranking of potential surrogates

The ranking of potential surrogates is determined by a multi-criteria decision making

process. This sub-step consists of three elements and is discussed below. A detailed

explanation of each element with its research method is provided in Section 4.6.


(1) Criteria weighting:

Based on the importance of each criterion (Figure 3.10), different weights can

be applied to each of the five surrogate evaluation criteria. If the evaluator thinks that

all five criteria are equally important, an equal weight can be applied. The experts can

also be asked to rate the importance of criteria through pair-wise comparisons. The

consolidated pair-wise comparisons of experts can be analysed using multi-criteria

decision analysis techniques, such as the Analytical Hierarchical Process (AHP), to

determine the weights for the criteria.

(2) Multi-Criteria Decision Analysis (MCDM):

Once the independent evaluation is completed, a multi-criteria decision making

method can be employed to select the best surrogate indicators to measure the required

social resilience characteristic (Ziyath et al., 2013). The Multi-Criteria Decision

Making (MCDM)/Multi-Criteria Decision Analysis (MCDA) tool is designed to

evaluate several possible decisions or items against multiple, but often conflicting,

criteria that can identify the best possible decision.

(3) Ranking of potential surrogates against the criteria:

Hence, an MCDA method such as PROMETHEE (Preference ranking

organisation method for enrichment evaluation) can provide a ranking for potential

surrogates evaluated by experts. A sensitivity analysis can also be carried out with

equal weights and consolidated weight obtained from experts to understand the

influence of criteria weights in surrogate ranking. Ranking preferences of different

cohorts of experts can also be analysed to understand if there are any differences in

preferences of potential surrogates.

C.3. Selection of optimum surrogates for application

Finally, in practice, not all potential surrogates perform well when evaluated

against all the criteria. Hence, selecting the best indicators require compromises in

their performance against five surrogate evaluation key criteria: accuracy, cost-

effectiveness, time-sensitivity, measurement complexity, and communicability. The

selection of surrogates that perform optimally in all criteria can reduce the chances of

poorly performing surrogates being selected for application (Lindenmayer et al.,

2015b). This can limit the risk of errors in interpreting the target resilience indicator.

Hence, multiple surrogates can also be selected based on the final ranking against all

five evaluation criteria, for comprehensive measurement of the target indicator.


However, the priority should be given to apply the first ranked surrogates, which were

selected through a robust identification and evaluation process.

SUMMARY

The surrogate approach has been successfully tested in a number of fields such

as ecology and clinical medicine, to predict parameters that are complex for direct

measurement. Surrogates can adequately represent a target indicator that is difficult to

measure, since surrogates are described through the identification of key facets of

target indicators. A conceptual surrogate framework to measure resilience indicators

was proposed and include three key steps: A. Selecting key resilience indicators that

require surrogate approach; B. Identifying potential surrogates; and C. Selecting the

optimum surrogates for application by evaluating and ranking potential surrogates.

This framework will help to systematically select surrogates for real world application

and hence, will assist in effective resilience investment decision making.

The application of surrogate approach to measure social resilience has not been

tested in the Disaster Risk Reduction (DRR) sector to date. Measuring social resilience

indicators using a surrogate approach will address the knowledge gap in resilience

measurement in a disaster context. The framework that was developed in this research

to measure social resilience by using surrogates will provide a new method to measure

resilience indicators using a surrogate approach in other dimensions of community

resilience, such as in the economic, physical, and institutional dimensions.

Once a set of surrogates is selected for a geographical context, they can be

regularly updated to continuously inform the resilience status of the community. It will

help the disaster management policy makers and practitioners to devise appropriate

strategies for enhancing resilience by effectively measuring it by operationalising a set

of surrogates to measure social resilience to disasters. The selection of optimal

surrogates will assist in overcoming conceptual and methodical challenges of social

resilience assessment.

Chapter 4: Research Method 62

Chapter 4: Research Method

This chapter describes the research method adopted in this study to investigate

the key research question, and to achieve three research objectives as detailed in

Section 1.5: to conceptualise, identify, and evaluate surrogates to measure social

resilience indicators in a disaster context. This chapter is divided into seven sections

as shown in Figure 4.1.

The chapter starts with an explanation of the philosophical position of this

research project (Section 4.1). Section 4.2 provides an overview of the research

methods adopted to achieve the defined research objectives, including the selection of

a mixed methods approach. Section 4.3 discusses the selection of five key social

resilience indicators to develop a surrogate approach. Section 4.4 discusses the ethical

considerations of this research study.

The next two sections detail the research method for two phases of research.

Section 4.5 discusses the use of case study interviews in phase I (qualitative study) to

facilitate the exploration of potential surrogates to measure the five selected social

resilience indicators. Section 4.6 discusses the use of an online survey questionnaire

in phase II (quantitative study) to evaluate potential surrogates identified in phase I.

In each section, respective components of the surrogate framework, data collection

technique, and data analysis methods are discussed.

Finally, this chapter concluded with the summary of research methods adopted

in this study (Section 4.7).


Figure 4.1. Chapter 4 and key sections in thesis structure


PHILOSOPHICAL POSITION OF RESEARCH

A consistent and well-articulated set of epistemological, ontological, and

methodological assumptions will guide researchers to formulate an appropriate

philosophy to frame the research question, research objectives, research methods, and

the interpretation of research findings (Saunders et al., 2015). Research philosophy

defines what knowledge is generated and its nature, and the key assumptions of world

views. Research design is informed by the beliefs and assumptions that underpin

various research philosophies to explain and justify the methodological choice,

research strategy, and data collection and analysis techniques (Saunders et al., 2016).

A research study is designed based on three key philosophical fields - ontology,

epistemology, and methodology. Ontology is a set of assumptions about what exists

in the world, epistemology is a way of understanding the assumptions on what exists,

and the methodological approach is determined by a collective consideration of both

the ontological and epistemological perspectives (Henn et al., 2009).

Ontological position

Ontology refers to the concerns of what exists and the nature of being. There are

two positions of ontology – realism and nominalism (Neuman, 2014) or objectivism

and subjectivism (Saunders et al., 2019). According to Saunders et al. (2019, p. 135),

“ontologically, objectivism embraces realism, which, in its most extreme form,

considers social entities to be like physical entities of the natural world, in so far as

they exist independently of how we think of them, label them, or even of our awareness

of them”. On the other hand, subjectivism is often associated with constructionism,

that views the motivations of social actors and the context in constructing the reality

(Bryman, 2015; Saunders et al., 2019). This research is mostly guided by a

constructionist ontological position, as it builds on a social phenomenon through the

views of social actors.

Epistemological position

Epistemology is the issue of knowledge production and how we know

scientifically what we know (Neuman, 2014). There are many research philosophies

based on epistemology, and the key ones are: positivism, realism, interpretivism, and

pragmatism (Saunders et al., 2019). The following philosophies are briefly described

as follows:


Positivism advocates for the development of knowledge based on careful observation

and measurement of the objective reality that exists in the world (Creswell & Creswell,

2017);

Realism asserts that reality can be understood through the use of appropriate methods,

and scientific conceptualisation is a way of knowing that reality (Bell et al., 2018);

Interpretivism considers the differences between humans in their role as social actors

and objects, and seeks to understand the world of research subjects from their point of

view (Saunders et al., 2019);

Pragmatism opens the door for different world views, varying assumptions, and

multiple methods of actions, situations, and consequences rather than precursor

conditions (Creswell & Creswell, 2017). The philosophical assumptions of

pragmatists mostly align with mixed method research strategies. For example, this

approach is not confined to one philosophical system and reality, does not subscribe

to an absolute unity of world view, and aims to provide the best understanding of the

problem by looking at what and how to undertake research.

Different experts and stakeholders in disaster management can have multiple or

conflicting priorities in resilience measurement, and thus, in resilience investments

decision making. However, it is important to balance different viewpoints with the

objective of resilience measurement, as there is a need for a consistent approach and

method to measure it in any context. Therefore, a resilience measurement research

needs to balance different perspectives and viewpoints. This makes it very difficult to

label any social resilience research to particular epistemology. Yet, aligning the

research to a most appropriate epistemological position will help to build a strong

philosophical foundation to the research that makes the selection of a robust research

method.

Social resilience is a multi-faceted concept, and its measurement requires a

broad understanding of the context and key characteristics, the role of social actors,

and their perspectives. Many existing approaches to measure resilience in a disaster

context often fail to provide a broader picture of social resilience status. Many

limitations in the existing resilience measurement approaches are increasingly

apparent in recent research studies which predominantly aim to quantify resilience.

These limitations have led to the adoption of social constructivist approaches. For


example, by Endress (2015) as an analytical perspective to identify social processes in

their historical context, which can help to unlock the difficulties in measuring complex

and dynamic resilience characteristics, such as social resilience. Table 4.1 below from

Creswell and Creswell (2017) summarises four alternative combinations of

epistemological positions, strategies of inquiry, and research methods.

Table 4.1. Alternative combinations of knowledge claims, strategies of inquiry, and

methods (Source: Creswell and Creswell (2017) (p. 17))

Knowledge

claims

Research

approach

Strategy of inquiry Methods

Postpositivist Quantitative Experiment design Measuring attitudes,

rating, behaviours

Constructivist Qualitative Ethnography design Field observations

Emancipatory Qualitative Narrative design Open-ended

interviewing

Pragmatic Mixed methods Mixed methods

design

Closed-ended

measures, open-

ended observations

In the context of different philosophical paradigms, this research is most closely

aligned with a pragmatism epistemological viewpoint, as resilience and its

measurement (research subjects in this study) can be influenced by perspectives of the

disaster management experts and the views expressed by the key stakeholders in

disaster management, including communities. The pragmatism approach allows for

different viewpoints to generalise findings by looking at the best options available for

researchers, mostly through mixed method, as was adopted in this research. A

pragmatic approach adapts abductive reasoning that evaluates prior findings from an

inductive process to enable a workable solution, which is very common in sequential

exploration strategies in mixed method research (Morgan, 2007). The next section

provides a brief explanation of different research methods and strategies of inquiries,

and further explains mixed method as a research method selected for this study.


RESEARCH METHODS AND STRATEGIES

This study utilised the mixed methods approach, as it is aligned with the

pragmatism epistemological position, discussed in Section 4.1.

In general, research can be exploratory, descriptive, and explanatory, according

to the purpose of research (Neuman, 2014). There are three types of research designs:

qualitative research, quantitative research, and mixed method research. Each design

has its own merits and demerits and is selected based on the context of the research.

At a technical level, it is a matter of selecting the best-suited research design (research

methods and tools) to discover the phenomena being studied, and qualitative and

quantitative methods can be combined concurrently or sequentially in a single research

study design (Henn et al., 2009). This research adopted a mixed method exploratory

research strategy to examine complex and little understood phenomena in order to

develop primary ideas to measure social resilience through a surrogate approach.

(a) Qualitative research

Qualitative research is described as an investigation of concrete case evidences

in a particular temporal and spatial setting, starting from the expressions and activities

of people in that context (Flick, 2014). Qualitative research techniques are associated

with collecting and analysing a set of narrative information (Teddlie & Tashakkori,

2009). Some of the qualitative research strategies include: ethnography, grounded

theory, case studies, phenomenological research, and narrative research (Creswell &

Creswell, 2017).

(b) Quantitative research

Quantitative research is a deductive method for testing objective theories of

social reality by examining the relationship among variables (Bell et al., 2018;

Creswell & Creswell, 2017). Its techniques are associated with collecting and

analysing a set of numerical data (Teddlie & Tashakkori, 2009). Some of the

quantitative research strategies include: survey research and experimental research

(Creswell & Creswell, 2017).

(c) Mixed methods research

Mixed method allows the collection of a set of evidence that is richer and

stronger to allow in-depth exploration of the research phenomena under investigation

to address the research question (Yin, 2014). Mixed method bases the program of


inquiry on collecting diverse, complementary and affirmative types of data that

provide a better understanding of the research problem (Creswell & Creswell, 2017).

There are three common mixed method research designs based on whether the studies

take place simultaneously or sequentially. The three mixed method research designs

are:

- Sequential explanatory;

- Sequential exploratory; and

- Concurrent triangulation.

The data collection method and role of qualitative and quantitative methods in

each mixed method design are summarised in Table 4.2. The founding research

philosophy of this study is mixed method sequential exploratory research design,

which is discussed in detail next.

Table 4.2. Three designs of mix method research (Boeije, 2009)

Designs Data collection Role of qualitative/quantitative

research

Sequential

explanatory

QUAN

qual

Quantitative data are

collected and analysed,

followed by qualitative

data. Priority is usually

unequal and given to

quantitative data

Qualitative data are used to explain

unexpected outcomes of

quantitative research. Qualitative

part is used to augment quantitative

data.

Sequential

exploratory

QUAL

quan

Qualitative data are

collected and analysed

first, followed by

quantitative data. Priority

is usually unequal and

given to the qualitative

data

The qualitative part is used to

develop theory and explore

relationships between phenomena.

Concurrent

triangulation

QUAN +

QUAL

Quantitative and

qualitative data collected

and analysed at the same

time. Priority is usually

equal and given to both

forms of data

The qualitative part is used to

confirm and cross-validate the

findings of the quantitative part.


4.2.1 Mixed method research - sequential exploratory strategy

In this study, sequential exploratory research strategy is adopted (Figure 4.2).

The sequential exploratory research design is chosen because the surrogates to

measure social resilience (the key phenomena of the study) need to be explored

initially (Boeije, 2009). This study aims to apply a surrogate approach to measure

social resilience indicators in a disaster context. Surrogates need to be explored

initially and then evaluated against a set of standard criteria for final selection. The

research objectives of this study guide the selection of sequential exploratory research

strategy in the mixed methods research approach, which is based on the needs of the

study to initially explore surrogates and then to evaluate it (Sharp et al., 2012). The

pragmatic perspective of research provides a greater flexibility to mix research

methods to initially explore to understand and then to interpret different viewpoints

(Greene & Hall, 2010).

This strategy involves the initial phase (Phase I), as shown in Figure 4.2, a

qualitative data collection and analysis (through interviews), followed by quantitative

data collection and analysis (through online questionnaire survey) that builds on the

results on the qualitative phase (Creswell & Creswell, 2017). Mixed method strategy

is useful because the emergent surrogate identification in the exploratory qualitative

phase can be further validated through an evaluation in the quantitative phase. This

strategy helps to generalise the results for wider contexts. It helps to explore a

phenomenon initially and to expand the qualitative findings in the next phase to

increase the validity of the selected findings in the qualitative phase (Creswell &

Creswell, 2017).

Initially, five social resilience indicators were selected to develop a surrogate

approach based on the review of literature. This helped to achieve Research Objective

1 (RO1): “To select key social resilience indicators that require surrogate approach

by developing a ‘5S’ social resilience framework in disaster management”. This is

Figure 4.2. Sequential design integrating qualitative and quantitative research

methodologies (Creswell & Clark, 2007; Flick, 2014; Hyde, 2006)


aligned to key step ‘A’ in the surrogate development framework shown in Figure 3.6.

The method for this step is explained in Section 4.3.

Followed by an extensive literature review, data collection using a mixed

method research approach was designed to identify, evaluate, and select surrogates in

two sequential phases. The mixed method adapted a sequential exploratory research

strategy that included a qualitative research method using interviews in phase I, and a

quantitative research method using an online survey questionnaire in phase II. The

purpose of phase I was to achieve the Research Objective 2 (RO2): “To identify

potential surrogates to measure key social resilience indicators in disaster

management”, using interviews at the community and sub-national levels. Phase I of

this research is aligned to key step ‘B’ in the surrogate development framework shown

in Figure 3.6. The detailed research method for phase I is provided in Section 4.5.

The analysis of data collected in phase I resulted in the identification of potential

surrogates for the five selected social resilience indicators. In phase II, surrogates

identified in phase I were evaluated against five key surrogate evaluation criteria to

rank potential surrogates that aimed to achieve Research Objective 3 (RO3): “To

evaluate and select optimum surrogates for application by ranking the potential

surrogates against surrogate evaluation criteria”. Phase II of this research is aligned

to key step ‘C’ in the surrogate development framework shown in Figure 3.6. The

detailed research method for phase II is explained in Section 4.6.

An overall research process with research methods for each phase of this study

is presented in Figure 4.3. It consists of key research methodological steps across each

phase and their corresponding research objective. Key steps in each phase shown in

Figure 4.3 are listed below with the expected outcomes (in italicised dot points).

1. In-depth literature review & research design phase: (to achieve RO1)

1.1. Literature review of social resilience frameworks and indicators

Identification of social resilience indicators that require surrogates

1.2. Selection of research method and design for developing surrogates

2. Phase I: Data collection (interviews) to identify surrogates: (to achieve RO2)

2.1. Field data collection (group interviews and 1-on-1 interviews)

2.2. Interview data analysis

Identification of potential surrogates to measure social resilience indicators


3. Phase II: Data collection (survey) to evaluate surrogates: (to achieve RO3)

3.1. Multi-expert evaluation of surrogates using online questionnaire survey

Evaluation of surrogates based on five surrogate evaluation criteria for

selecting robust surrogates

3.2. Multi-expert multi-criteria survey data analysis

Rank and select the final surrogate(s) based on their performance using multi-

criteria decision analysis

Figure 4.4 shows the overall research methods in each phase of this study,

aligned with research objectives and key steps in the surrogate development

framework (A to C) as shown in Figure 3.6. The surrogate development framework

component in each phase of the study is integrated into one overall research

framework.

Literature review and research design phase: Identification of five key social resilience

indicators that require surrogates for their measurement.

In Phase I: Three sub-steps in identification of surrogates are included:

- Identification of all potential surrogates using a case study research

- Establishing surrogacy relationship

- Exploring protocols for measuring potential surrogates

In Phase II: Three sub-steps in selecting surrogates identified in phase I are included:

- Multi-expert evaluation of potential surrogates against five evaluation criteria

using an online survey

- Ranking of potential surrogates using Multi-Criteria Decision Analysis

(MCDA) method

- Selection of optimum surrogates for application

The research method, data collection technique, and data analysis method for

each phase of this study are discussed in detail in the following sections (Sections 4.3,

4.5 and 4.6). Section 4.4 provides a brief summary of ethical considerations of this

research in both phases.


RO2 RO3

To identify potential surrogates to

measure key social resilience

indicators in disaster management

To evaluate and select optimum

surrogates for application by

ranking the potential surrogates

against surrogate evaluation

criteria

To select key social resilience indicators

that require surrogate approach by

developing an inclusive and adaptive

social resilience framework in disaster

management

RO1

Figure 4.3. Overall research process with key steps in three phased research


Figure 4.4. Expanded research framework with sub-steps for testing surrogate approach (In phases aligned with ROs)


SOCIAL RESILIENCE INDICATORS SELECTED FOR DEVELOPING

SURROGATES

At the initial stage of this research (literature review and research design), a

detail review of social resilience assessment frameworks was undertaken to identify

social resilience indicators that need surrogates due to the complexity of measuring

them directly. This process resulted in a ‘5S’ model social resilience framework and

five indicators were selected to develop a surrogate approach, which was to achieve

RO1 as shown in Figure 4.5.

In this study, a ‘5S’ inclusive and adaptive social resilience framework was

developed from the critical analysis of existing social resilience frameworks, as shown

in Figure 2.5. The proposed framework can guide the selection of appropriate social

resilience indicators and their adequate operationalisation based on the context of its

application. This framework was used to select a set of key social resilience indicators

to develop a surrogate approach.

Although the majority of the indicators among the 46 indicators require a

surrogate approach, five key indicators were selected (one indicator for each social

resilience sub-dimension) for developing surrogates in this research. The five

following social resilience indicators were selected for developing a surrogate

approach as a testbed:

To select key social

resilience indicators

that require

surrogate approach

by developing an

inclusive and

adaptive social

resilience framework

in disaster

management

RO1

Figure 4.5. Extract from the overall research process for the literature

review and research design phase (RO1)


1. Access to transport facility in a disaster situation to measure social mobility;

2. Social trust during disasters and recovery to measure social cohesion;

3. Learning from past disaster experience to measure social competence;

4. Involvement and equity for people with specific needs in different phases of

disasters to measure social equity; and

5. Existing cultural and behavioural norms in relation to disaster risks and

managing disasters to measure social beliefs.

These were selected based on the following three general criteria for deciding the

surrogate approach:

1. Process resilience indicators that are not very static (dynamic, frequently change

over time); or,

2. Complex to conceptualise and cannot be easily measured quantitatively (e.g. social

trust, community competence); and,

3. When resilience is measured using indirect methods, existing frameworks mainly

use data from publicly available census data sources, which is often not able to

accurately and adequately measure the target indicator.

In this study, a fourth criteria was also included to ensure the wide applicability of the

findings, namely, indicators that are relevant to multiple disasters and different

geographical and socio-economic contexts. Finally, one indicator was selected for each

of the five social resilience dimensions in the ‘5S’ social resilience model developed

by Saja et al. (2018).

Table 4.3 shows the detailed descriptions and justification of five key indicators

that were selected for developing surrogates. As previously stated in Chapters 1 and 2,

direct measurement of the above key resilience indicators often requires detailed

household surveys, which are costly and time consuming. These indicators are

geographically very dynamic, varying greatly from household to household and among

social groups. They also change frequently with time. For example, regular and

detailed community surveys are required to identify multiple transport facilities

available and required for disaster evacuation. This information needs to capture

unique household requirements needed to evacuate people with special needs, such as

disabled, elderly, children, and people who require special assistance. Therefore, a set


of surrogates that can be mainly measured through regularly updated administrative

data is needed to overcome existing limitations in resilience measurement, instead of

outdated and less frequently updated census data. In most of the developing countries,

census data is not often updated. The local authorities and administrative state

departments, however, update their administrative data regularly. Hence, new

approaches are needed to find ways to identify key facets of social resilience indicators

that are difficult to measure. The key facets of these indicators can be potential

surrogates for which the data can be obtained from the regularly updated data sources.

A list of existing measures for the five selected indicators in the 31 social

resilience frameworks analysed in this study is provided in Appendix A. Most of the

existing measures have used census statistics data that are available from the public

databases. They are not adequate to provide a good measure of social resilience as

discussed in Chapter 2 (literature review). Further, there were no relevant measures

found for some of the key social resilience indicators in the 31 social resilience

frameworks analysed in this study, such as equity for people with specific needs and

cultural/behavioural norms.


Table 4.3. Details of social resilience indicators selected for developing surrogates

# Indicat

or

Characteristic Link to social resilience to

disasters

Justification for selection of the indicator to measure social

resilience

References

1

Access

to

transpor

t facility

in a

disaster

situation

Social mobility

during disasters

for effective

evacuation and

disaster

response

Ability of greater mobility of

people individually and of

families collectively can

enhance social resilience to

disasters and improve response

effectiveness at the time of

disasters, particularly during

evacuation.

There are many ways disaster evacuation can happen,

including using private vehicles or mass transport systems or

there may be a need for access to emergency transport

facilities. Access to transport is mostly attributed to vehicle

ownership. Disaster evacuation researchers also found that,

evacuation behaviour is interconnected to many factors such

as socio-economic status, race, ethnicity, and local language

ability. Availability or access to special needs transport is

another crucial element with regard to mobility of people in a

disaster context.

Adger et al.

(2005),

Burton (2015),

Kotzee and

Reyers (2016),

Peacock et al.

(2010), Tierney

(2009).

Frailing and

Harper (2017)

2

Social

trust

during

disasters

and

recovery

Social cohesion

that help

improve

response and

quicker recovery

Social cohesion is positively

interrelated to social resilience,

although the strength of this

association varies. Social

cohesion and cohesion among

social entities proved to be an

important factor in effective

coping and response to any

disaster.

Social trust is an integral part of social cohesion. Trust

among the community members during a disaster is very

important, because it positively facilitates coordination and

cooperation for effective disaster response and provide access

to resources. Trust by the community in local authorities is

also another key factor to consider in measuring social

cohesion. A positive strong correlation exists between the

level of trust and engagement of community members by

emergency management authorities on risk communication

and actions.

Leykin et al.

(2016), Aldrich and Meyer (2014b),

Ainuddin and

Routray (2012),

Paton (2007),

Lorenz (2013),

Townshend et al.

(2015)


3

Learnin

g from

the past

disaster

experien

ce

Social

competence in

managing

disasters

Communities exposed to

disasters should have a good

knowledge about their risks and

have the capacity to acquire

reliable and precise information

of risk assessment to solve

emerging problems.

Past experience and learning from the past helps the process

of understanding future emerging risks. Past experience and

exposure to a disaster helps to enhance social resilience,

which increase the competency of the community to prepare

for, and to face the next disaster. Greater importance is given

in many social resilience frameworks to community

competence as one of the primary sets of adaptive capacities

to build social resilience. Coping style influences social

resilience and problem-focused coping style is helpful for

effective disaster response and recovery.

Norris et al.

(2008), Leykin et

al. (2016),

Miller et al.

(1999)

4

Involve

ment

and

equity

for

people

with

specific

needs

Community

inclusiveness

and social

equity in

disaster

preparedness

initiatives

Equitable gender and social

relations can contribute to

enhance social resilience.

Inclusion of people with special

needs in the process and

provision of equal opportunities

in disaster planning and

recovery are important to

enhance social resilience.

Exclusion of the most vulnerable and less resilient segment

of the population has been highlighted as the most pressing

issue in disaster response and recovery programs, due to

difficulty of access as well as purposeful negligence. Ethnic

minority, race, caste, cultural disparity, and any other forms

of marginalisation based on socio-economic status of the

population will contribute to lowering social resilience,

predominantly in resource allocation and provision of equal

access to needs and services.

Cutter et al.

(2003),

Enarson (1998),

Kotzee and

Reyers (2016),

Begum (2008)

5

Existing

cultural

and

behavio

ural

norms

Local cultural

norms and

social belief

systems dealing

with disaster

risk

Cultural beliefs exert a certain

level of influence on the

interpretation of disaster risks by

a community. Resilient building

initiatives must be undertaken

with greater cultural sensitivity

to help increasing capacities,

and so as not to lower existing

resilience.

Not only human resources and physical assets, but local

culture and social beliefs can play an essential role in

determining social resilience. Social beliefs need to be

positively capitalised in communities that are oriented with

their own local culture and faith systems. Different cultural

groups within the same community may differ in their

preferences in resilient building activities and also in disaster

response actions.

Eiser et al.

(2012),

Wilkinson (2015),

Kwok et al.

(2016),

Ostadtaghizadeh

et al. (2016)


RESEARCH ETHICAL CONSIDERATIONS

Research ethics is concerned with the procedures that should be applied to

protect research participants and to regulate the relationship of researchers with them

(Flick, 2014). Ethics approval from QUT was obtained, since this research was

conducted with people and their data. This research was a human-centred, which was

conducted with selected disaster management experts and people with disaster

management experience. Therefore, people were invited to take part in interviews and

an online survey.

Ethical compliance for this research was submitted to the QUT Office of

Research Ethics and Integrity (OREI) for ethical clearance. The ethics application

under the category of ‘Negligible-Low risk’ was approved by the Queensland

University of Technology (QUT) Human Research Ethics Committee (HREC). The

approval UHREC Reference number is 1700000832. The approval from QUT-HREC

is provided in Appendix B.

This research was considered a negligible or low risk study, because:

o Foreseeable risk to participants was no more than "inconvenience" due to the

limited amount of time spent on interviews and completing the online survey;

o Questions to participants were general in nature and did not relate to business/

past or current project practices;

o Participants’ reaction to questions in interviews and survey was not likely to

cause any harm or distress;

o Participation in interviews and surveys was on voluntary basis and results

reported remain anonymous; and

o In the publication of research results, participants of interviews and online

surveys were de-identified.

4.4.1 Ethical considerations in interviews (Phase I)

The semi-structured questions during interviews focused on the identification of

potential surrogates in general. Hence there was no personal information collected.

Any personal information was treated as confidential and aliases such as “participant

in interview #1” were used instead of the real names of participants in all transcripts

and publications.


All participants were given time to read the “Information for interview

participants” sheet and were also provided with a brief explanation about the research,

its benefits and purposes prior to starting interviews. Interviewees were reminded that

their participation in the interviews is strictly voluntary. If they feel discomfort, they

have the right to withdraw from the interview. Interview participants were requested

to seek permission from their organization/superior, when needed, before participating

and abide by confidentiality and privacy rules as per their organisation’s requirements.

A consent form was signed by each participant prior to starting interviews to indicate

that they agreed to participate in the interview on a voluntary basis.

For group interviews, a list of members attached to divisional disaster

management committee were obtained from the Divisional Secretary. The researcher

selected the participants for each group interview based on their experience (minimum

of three years of experience in disaster management). The selected interviewees were

directly contacted by the researcher to check their willingness to participate in the

group interview. Further, for group interviews, participants of a similar employment

level within the organisational hierarchy were formed into a group for the interviews

to minimise their discomfort or concerns at speaking on the topic in front of others

who are at a higher level than them. The researcher encountered some difficulties to

fix the time for group interviews that was convenient for all the participants who

agreed for group interviews. Hence, some participants did not turn out at the scheduled

time for group interviews. When participants who were selected for group interviews

were unable to participate in the group interview, one-on-one interview was conducted

with them at a later date that was convenient for them.

4.4.2 Ethical considerations in online survey (Phase II)

The online survey was anonymous, and personal identifiable information such

as name or contact details were not collected. All requirements in participating in the

online survey were written in the survey email invite and respondents were requested

to go to the online questionnaire link to respond, if they wished to participate. All

participants were sent “Information for survey participants” sheet as an attachment to

the email to read before they decide to respond to the survey questionnaire. The survey

targeted mid and senior level individuals who had a minimum of three years of direct

work experience in disaster management. This was ensured by the first question


included in the questionnaire, “Do you have a minimum of three years of direct work

experience in disaster management work”.

PHASE I – QUALITATIVE CASE STUDY RESEARCH FOR

IDENTIFYING POTENTIAL SURROGATES (INTERVIEWS)

4.5.1 Identification of potential surrogates: Case study as a strategy of inquiry

The case study research method is identified as a key strategy of inquiry to

unlock the complexities in social phenomena (Yin, 2014). A case study is an intensive

research process defined as an in-depth study of a social phenomenon within a unit of

observation; in this case, the unit of analysis is an urban division (Swanborn, 2010).

At the local level of this case study, disaster management planning and project

implementation are done in Sri Lanka division-wise.

While every country has its own definition of an ‘urban’ locality, ‘urban’ is

defined in the Sri Lankan context as “a smallest administrative division which has a

minimum population of 750 persons, a population density greater than 500 persons per

square kilometre, firewood dependence of less than 95 % households, and well-water

dependence of less than 95% households” (Weeraratne, 2016, p. 4). There is an

increasing emphasis for enhancing resilience in urban areas due to rapidly changing

disaster risk landscape (Maduz & Roth, 2017). The location of this research in urban

areas has wider applicability to other urban contexts in the world. Figure 4.6 shows the

research process in phase I, which was to achieve research objective 2 (RO2).

RO2

To identify potential

surrogates to

measure key social

resilience indicators

in disaster

management

Figure 4.6. Extract from the overall research process for Phase I (RO2)


A key justification for undertaking case study research is its ability to investigate

a new approach in detail. The use of a surrogate approach in a disaster context is an

innovation of this research, and requires an in-depth exploration of the surrogate

approach in real-world context (Henn et al., 2009; Yin, 2014). A case study approach

can help to maintain a pragmatic position that assumes a reality can be constructed and

its understanding can be developed through social experiences (Mills et al., 2017).

There are many strengths in case study research that enables the measure of

abstract concepts such as resilience to actual experiences and evidences (Neuman,

2014). Some of the strengths of case study research in the context of measuring social

resilience in disaster context include (Neuman, 2014):

o Conceptual validity, which helps to identify concepts/surrogates for measuring

social resilience, an abstract phenomenon;

o Extension of the surrogate approach to measure social resilience in a disaster

context;

o Ability to make social mechanisms more visible in order to deconstruct the key

facets of social resilience indicators

o Ability to capture complexity and trace processes in a social context;

o Enables the study to fine-tune measures of abstract concepts of resilience to

new experiences.

Further, the exploration of surrogates in multiple case study locations uses

multiple sources of evidence, maintains a chain of evidence, and creates a case study

database that enables cross-case synthesis (Yin, 2014). It allows diverse insights, since

data were obtained from a wide range of participants. Multiple cases allows

triangulation to reduce bias (Yin, 2014) and avoid focusing on the predominant views

within the specific study group.

Interviews as a data collection technique

Interviews with disaster management practitioners and policy makers working

at the community and sub-national levels were selected as a data collection technique

in the case study areas to allow easy identification of a set of potential surrogates to

assess selected indicators. Interviews can provide more detailed and important insights

for the research objectives being studied (Hancock & Algozzine, 2016), which is to


identify potential surrogates. Therefore, interviews were selected as the most

appropriate data collection tool for phase I of this research.

Interviews is one of the widely used technique for collecting qualitative data,

different qualitative interviewing strategies from diverse disciplinary perspectives

have resulted in a wide range of interview-based data collection techniques (DiCicco‐

Bloom & Crabtree, 2006). This includes face-to-face and one-on-one interviews,

telephone interviews, focus group interviews, and email internet interviews (Creswell

& Creswell, 2017). Survey interviewing, in-depth interviewing, and life story

interviews (Gubrium et al., 2012) are also used. For exploratory inductive research

objectives that focus on what and how social processes, interview techniques using a

one-on-one method are best suited (Johnson & Rowlands, 2012), because they can

delve into how surrogates are linked to the target indicator (surrogacy relationship)

and how they can be measured (measurement protocols). In this research, face-to-face

groups and one-on-one interviews were used for initial exploration of potential

surrogates, since this method can uncover deeper information from the knowledge and

experience of the respondents (Johnson & Rowlands, 2012).

Group interviews: Some researchers have characterised ‘group interviews’ with

limited interaction among participants as focus groups, and some others view a focus

group as a ‘group discussion’ which gives more importance to interaction between

participants than to the group interviews (Alasuutari, 2008; Bryman, 2015). A focus

group is also a type of group interview, but primarily seeks to generate qualitative data,

from the interaction between group members (Davis, 2016). In this research, group

interviews are defined as interviews with more than two participants, where the

interviewer listens to participants’ opinions individually, without any interaction with

other members in the group (Davies, 2010; De Ruyter, 1996). In group interviews,

participants are able to connect to their own experience and can build on what has been

already described by other participants. The interviewer as a moderator can encourage

each participant to offer new ideas (Axinn & Pearce, 2006) and limit the responses

when it is being repeated. Further, a group setting encourages each participant in the

group to diffuse diverse ideas influenced by other group members (Patton, 2002),

whereas in individual interviews, the same ideas can be repeated by every interviewee

and may not contribute new information. However, group interviews have some

drawbacks compared to one-on-one interviews, such as domination by some members


in the group. These limitations were overcome in this study by providing every

member of the group an opportunity to answer every question; this was set as a ground

rule before the interviews commenced.

Individual interviews (one-on-one/in-person): In order to gather information-rich

samples that are geographically and institutionally widespread, one-on-one interviews

were undertaken in addition to the group interviews, since it was often difficult to get

participants together in one place to undertake a group interview. Individual interviews

followed the same process and structure (semi-structured questionnaire format) of the

group interviews. However, during individual interviews, each respondent had more

time than in group interviews to explore potential surrogates for each social resilience

indicator. A total of 39 one-on-one in-person interviews were held, including 20 at the

sub-national level (Case study #4). The remaining 19 were held in the divisional case

study locations (Case studies 1 to 3).

Both one-on-one and group interviews followed the same structure and process.

An interview guide with five semi-structured questions for each of the five resilience

indicators was used, based on three sub-steps to identify potential surrogates, as shown

in Figure 4.4. Interview questions were structured as follows (detailed interview guide

for each of the resilience indicators is provided in Appendix F).

Appendix C shows the letter of recruitment for interview participants; Appendix

D shows the participant interview consent form used in this study to obtain a signed

consent for participation in the interview; Appendix E shows the participant

information sheet provided to participants before the interview to familiarise them with

the interview processes and ethical guidelines.

Creswell and Creswell (2017) highlighted four key aspects in selecting study

locations and study participants for data collection using interviews:

1. Study setting - where the research was conducted;

2. Study actors - who participated in interviews;

3. Study events - what the participants were interviewed about;

4. Study process - to know how the work of actors evolved within the research setting.

Each aspect is discussed below.


Study area and location (study setting)

The Kalmunai Municipality of Ampara district in the Eastern Province of Sri

Lanka was selected as the case study area for this research. Sri Lanka is one of the

countries in Asia that is highly vulnerable to multiple hazards, such as tsunamis,

cyclones, floods, and landslides. Sri Lanka was rated among the top 10 most affected

countries in the climate risk index for 2017 by Germanwatch (Eckstein, 2018).

Kalmunai Municipality was selected because it is frequently exposed to multiple

disasters such as flooding, tsunami and cyclone (UNHABITAT, 2015; Zubair et al.,

2005) and the researcher could access a wide range of disaster management experts

working in this region.

Kalmunai municipality is the most populated local authority in the coastal area

of Ampara district and can also be characterized as ‘Urban’ area by it’s highly densely

population (JICA, 2008). In Sri Lanka, all areas administered by Municipal and Urban

councils constitute the urban sector and ‘urban’ area is a Grama Niladhari Division,

“if it has a minimum population of 750 persons, a population density greater than 500

persons per km2, firewood dependence of less than 95 % households, and well-water

dependence of less than 95% households” (Weeraratne, 2016, p. iv). The population

density per square kilometre of the Kalmunai municipality is as high as 5,000 per

square kilometre whereas the national average is 299 people per square kilometre

(GoSL, 2018; JICA, 2008). The total population within Kalmunai Municipality is more

than 110,670 and most of the people live within 1.5 km range of the area from the

seashore of Indian Ocean in the East (GoSL, 2018).

The selection of urban context is also justified, because the statistics highlight

that 54% of the world population lives in urban areas and 16% of them live in large

cities around the world (WorldBank, 2014). Hence, the findings from this research will

have a wide applicability, since the data collection for this research was done in the

urban context, compared to being undertaken in a rural context. Further, the historical

records (Jameel, 2009) show that a series of cyclones hit South-eastern coast of Sri

Lanka in 1980s and 90s. This includes severe and medium category cyclones in 1845,

1891, 1907, 1921, and 1978. Although the damage and loss data is not available for

these cyclone events, 1978 cyclone was reported to be the major cyclone in the

historical records available, and it took three months to bring back to the normalcy.

Three major floods were also reported in this area: in 1933, 1957, and 2010. The 2004


Tsunami was the major devastating disaster in the history of this area, which killed

more than 3,000 people and entirely wiped out half of the coastal areas of Kalmunai

Municipality, leaving thousands of families, homeless and houses severely damaged.

It took more than five years for recovery from the Tsunami damages.

Four case studies were adopted. Among these, three urban study locations within

the Kalmunai municipality (local authority) were selected for this research. The three

case study locations selected are at the lowest state administrative level, which is

termed a ‘Divisional Secretariat’. This also allows multiple sources of evidence and

convergence of data in a triangulating fashion, which helps exploratory studies in

theory development (Yin, 2014).

Figure 4.7 shows the geographical area of the three study locations (Divisional

Secretariate administrative divisions), namely:

Figure 4.7. Case study locations in Sri Lanka map


Case study area 1: Kalmunai Tamil (denoted as ‘KMT’),

Case study area 2: Sainthamaruthu (denoted as ‘SM’), and

Case study area 3: Kalmunai Muslim Division (denoted as ‘KMM’)

The fourth case study area was selected at the sub-national level as ‘Ampara

district and Eastern Province’, which comprises other local authorities, including three

case study locations. The fourth case study location was at a sub-national level to reach

disaster experts and professionals with a holistic and national understanding of

disasters. They also work in three case study areas while working at the sub-national

levels in the district and province, which ensured wider representation of samples. In

total, the selection of four case studies as a whole provides a good representation of a

wider geographical region from community to sub-national levels, where there is a

wealth of disaster management experience.

The structure of Sri Lanka’s administrative and devolved political power

structures and their linkages is shown in Figure 4.8. It depicts the relationship of case

study locations to the lower and sub-national levels in the central administrative

structure. It also shows the parallel devolved governance structure at local authority

and provincial levels. In this study, interview participants were sampled at two levels:

Figure 4.8. Administrative and authority governance structures

in Sri Lanka


sub-national and lower (community) levels of administration and authorities, since

both entities have roles in disaster management in their respective levels. The district

level comes under the central government administrative structure, and the provincial

level is part of the devolved political power structure.

Sample selection (study actors): Purposive sampling strategy for interviews

Purposeful sampling is more appropriate for qualitative research that uses an

interview method for data collection, as individuals who have experience in the central

phenomenon of the research can be reached more effectively (Creswell & Creswell,

2017). The basis for a purposive sampling strategy is justified if the random selection

fails to yield the most informative samples, and therefore samples in this study can be

chosen for their knowledge which is representative of a larger population (Alasuutari

et al., 2008). Theoretical sampling, where individuals are selected according to their

ability to contribute new insights to the research phenomenon being investigated

(Flick, 2014) was the starting point in a purposive sampling strategy for this research.

Sample selection was further intensified with the identification of samples with

different levels of experience, from administrative to community work, and with wider

variations such as senior to mid-level officers (Flick, 2014).

a) Whom to select for interviews: The sampling strategy began with information-rich

participants (Marshall & Rossman, 2014) in relation to the research phenomenon,

often referred to as ‘key informants’, who are considered to be knowledgeable about

the subject (Faifua, 2014). The sample should inform the research objective

sufficiently to gather a substantial set of information that can then be content

analysed. A list of members in both divisional and district disaster management

committees in each case study location was obtained from the relevant divisional

and district government offices (theory-based samples). The final selection of

samples was done by selecting people who are active members of divisional/district

disaster management committees and have a minimum of three years of experience

in disaster management work and/or are key decision makers in disaster

management planning. Further, the inclusion of ‘information rich samples’ was

ensured by selecting those who have more than three years of experience in disaster

management projects at the sub-national level, as shown in Figure 4.9.


b) How many interviews: The total selected sample size for phase I was 50, including

a minimum of ten for each study location (i.e. one DS division). According to Bion et

al. (2000), the optimum number of interviews for a research study lies somewhere

between 15 and 25 individual interviews, and for groups, the optimum number can be

between six and eight groups. Since this study combined individual and group

interviews, it was planned to reach 50 interviews. Table 4.4 provides a detailed

overview of the number of selected interview participants for each case study location,

and Tables 4.5 (a) and (b) details the summary of participant profiles, such as their

administrative positions and affiliations.

A total of three group interviews (one in each case study location) and 39 one-on-one

interviews were arranged. Dynamics of the response can be increased by including

diverse stakeholders (Flick, 2014). In this research, three to four members in each

group participated in group interviews in three DS divisional case study locations (case

studies 1 to 3). A group interview technique was used in this research with selected

government officers who work in the same office (divisional secretariat) and in the

same geographical area. Some participants may gain more confidence in a group

setting from the responses of other participants compared to one-on-one interviews

Figure 4.9. Sample selection process diagram


(Axinn & Pearce, 2006). Group interviews are efficient because group members

stimulate each other, interviewer-respondent relationship can be sustained, they are

less likely to stall compared to one-on-one interviews and effective when the time-

frame is limited (Fontana & Prokos, 2016).

Table 4.4. Samples selected for interviews in all study locations

Table 4.5. Interview participant profiles: (a) designation/position and (b) affiliation

As shown in Table 4.5 (a) and (b), the experts interviewed in this study are a

good mix of senior-level highly experienced government and project management staff

members. 60% of interviewees hold high level positions such as Project Director,

Project Manager, Project Officer, and Project Consultants. Mid-level/community level

Interview participant

affiliation in case study areas

Group interviews

(# of experts)

Individual

interview

Tota

l

Kalmunai T (Case study 1) 1 Group (3 experts) 7 10

Sainthamaruthu (Case study 2) 1 Group (4 experts) 6 10

Kalmunai M (Case study 3) 1 Group (4 experts) 6 10

Case study 4 includes Ampara

District and Easter Province N/A 20 20

Total interviews and

participants

3 Group interviews

(11 participants)

39

participants 50

(a) Interview participants’ designation (n=50)

Project Director 18 36%

Project Manager 5 10%

Project Officer 2 4%

President/Executive Director 3 6%

Project Consultants 5 10%

Village Administrators 12 24%

Social Service Officers 2 4%

Divisional Disaster Relief

Officers

3 6%

(b) Participants’ affiliation (n=50)

Central Government 28 56%

Local Government 2 4%

NGO (Local) 11 22%

NGO (International) 7 14%

Multilateral agencies such as UN 2 4%


government officers, village administrators and leaders of community organisations

were 34% and 6%, respectively. The majority of the participants were from high level

positions, since almost all the participants in case study 4 and two to three participants

in each division were senior and experienced officials. The expert sample also included

60% from the public sector, 22% from local/community-based organisations, and 18%

from international agencies.

Table 4.6 shows the affiliation of interview participants in divisional

administration and their respective positions in the divisional disaster management

committee. The sample selection for each case also ensured the inclusion of interview

participants from eight different functionalities/departments involved in disaster

related activities, such as social services, disaster relief services, village administrative

services, and planning services of a divisional administrative area, as shown in Table

4.6.

Table 4.6. Selected participants for interviews in one study location (DS division)

# Divisional Disaster

Management Committee Position Affiliation No

1 Divisional Secretary (DS) Co-Chairman Administration

service 1

2 Asst. Divisional Secretary

(ADS) Secretary

Administration

service 1

3 Assistant Director of

Planning (ADP) Convener Planning Service 1

4 Administrative Officer

(Village Divisions)

Coordinator

for village

divisions

Local Administration 1

5 Social Service Officer

(SSO) Member Local Administration 1

6 NDRS officer (National

Disaster Relief Services)

Coordinating

Assistant

Disaster Mgt. Centre/


7 Village division

administrators

Representing

village

divisions


8

Representative from the

civil society organisation

(CSO)

Member

representing

CSO

Civil Society

Organisation (CSO) 1

Total 10


Study events - What the participants were interviewed about

1. The interviewer provided an introductory explanation of the social resilience

indicator for which the surrogates were explored.

2. To kick-start the interview, interview participants were asked to briefly explain the

current context of the resilience indicator in the case study area (This question was

to help the participants to start thinking of key facets of the resilience indicator).

3. Next, participants were asked how the target resilience indicator was assessed. This

question was to explore potential surrogate measures to assess the target indicator.

4. From the proposed key facets of the resilience indicator (which can be a potential

surrogate to that indicator), participants were probed on what key attributes indicate

a target resilience measure for gauging the surrogacy relationship. This was to

ensure that the proposed key facet of the resilience indicator is very closely related

to the target indicator.

5. Finally, the interviewer probed how the data can be obtained to measure the

proposed potential surrogate measure to understand the measurement protocols.

Study process – How interviews were conducted

Interviews were conducted from September to November 2017. All the

interviews were conducted in Tamil except for two interviews which were done in

English. The majority of the participants were from the Eastern Province of Sri Lanka,

and their native language is Tamil, and they are not necessarily fluent in English. Each

one-on-one interview lasted approximately 45 minutes, whereas the group interviews

took up to 60 minutes. All the interviews were audio recorded, which allowed the

researcher to focus on the interview questions and helped to record all the data. On

average, 10 minutes was spent on potential surrogates for each indicator. Interview

questions were posed, followed by additional probes to break the silence, and

responses were audio-recorded for analysis (Leeuw, 2008). Semi-structured interviews

with open questions were used in this research, since these are much more flexible and

more likely to yield new ideas than structured survey-type interviews or less structured

interviews (Axinn & Pearce, 2006).


4.5.2 Phase I – Interview data analysis

The qualitative phase of the research using interviews, is a cyclic process

between data collection, data analysis and sampling (Boeije, 2009; Vaismoradi et al.,

2016). However, the quantitative phase, using an online survey, is linear. In

exploratory mixed method research design, connected data analysis is done to

generalise findings (Creswell & Creswell, 2017). The complete representation of a

seven-step sequential, exploratory, mixed method data collection and analysis model

is shown in Figure 4.10 (Creswell & Creswell, 2017).

The case studies were analysed individually, with interview audio recordings

transcribed, and then professionally translated from Tamil to English language. The

interview transcripts yielded large volumes of text. Hence, the content analysis of

interview data was done using the Leximancer data mining tool.

Leximancer as a textual data analysis tool

Leximancer software is a computer-based lexical analysis data mining tool, used

for content analysis of qualitative data, and can produce thematic strings. Leximancer

was used because it provides an automated data analysis based on lexical text

properties, and it is efficient in handling large sets of qualitative data without bias

(Sotiriadou et al., 2014). Further, Leximancer is not only more effective than manual

coding, but also identifies a broad range of concepts thereby increasing the consistency

and reproducibility that helps analysts to derive greater insights (Angus et al., 2013;

Penn-Edwards, 2010).

Figure 4.10. Sequential exploratory design (qualitative to quantitative)

(Creswell & Creswell, 2017)


Concepts in Leximancer are collections of recurrent words that travel together

throughout a text document, and a concept map is generated using a frequency of co-

occurrences of concepts (Leximancer, 2017). A set of concepts that formed a theme

was identified from the lexi concept maps in a systematic fashion across the entire data

set for each case study. The concepts were then translated into meaningful themes and

a set of correlated themes was then combined to create a higher-order theme. Each

higher-order theme was then interpreted in terms of potential surrogates that are

relevant to the intended target indicator of measure in this study. The process of

thematic synthesis of case study data is shown in Figure 4.11 (Cruzes & Dyba, 2011;

Cruzes et al., 2015). The levels of interpretation to develop surrogates followed a path

of abstraction from text to higher-order themes through the interpretation of concepts

and themes generated in Leximancer analysis.

Analysis method of interview transcripts in Leximancer

Interview data was fed into Leximancer in order to develop concepts and theme

maps. The initial content analysis of interview data for each case study generated

concepts and themes in three platforms: concept cloud view, concepts integrated into

themes map, and analyst synopsis with the descriptions of concepts and themes as

shown in Figure 4.12. A sample concept, theme, and higher-order theme map is shown

in Figure 4.13. Each node is a concept, each circle around a set of similar concepts is

a theme, which are automatically generated in Leximancer. The researcher identified

similar or closer themes to create a higher-order themes shown by solid square or

rectangular lines in Figure 4.13.

Increasing level of

abstraction

Figure 4.11. Levels of interpretation in thematic analysis


Themes and concepts were inter-related and their meaning was interpreted with

the help of interview data in the description of each concept. Identification of concepts

and themes by inferring the content of the interview data is similar to the identification

of patterns and typologies suggested by Flick (2014). At the end of the analysis, the

aim was to extract all potential higher-order themes that can be used as potential

surrogates to measure the five selected social resilience indicators. Thematic analysis

was conducted independently for each case study. A relevant quote selected from an

interview participants in each case study is provided below to describe the relationship

between the higher-order theme and the target indicator.

Distinct case analysis: The interview transcripts for each case study were collated into

a single document for individual analysis. Among many case study analysis

techniques, pattern matching (concepts and themes), explanation building, and cross-

case synthesis are used to analyse interview data which is the method that was also

adopted for this study (Yin, 2014). The concepts/themes map using Leximancer was

generated for each case study. From these maps, a thematic analysis was carried out

by combining the most closely related themes for each case study independently to

identify a set of meaningful higher-order themes for each case study. The higher-order

theme was considered as a potential surrogate for each case study area. Further, each

Figure 4.12. Process of case study data analysis using Leximancer


of the higher-order themes built from the distinct case study analysis is explained with

the interview transcript data. The key step of identifying higher-order themes was done

by combining most closely related themes generated by Leximancer. Figure 4.12

shows the process of reaching higher-order themes from the interview data for each

case study.

Cross-case synthesis: In order to build a new body of knowledge, a synthesis of the

multiple cases was necessary. This was undertaken using the case study analysis

results. Figure 4.14, adapted from Yin (2014) shows the analysis structure of multiple

case studies using cross-case synthesis to develop surrogates based on the higher-order

themes generated from the different case analyses. Finally, unique themes from the

different case analyses were cross tabulated to identify commonly occurring higher-

order themes across all case studies. The final set of unique higher-order themes was

taken as potential surrogates for each of the five selected social resilience indicators.

At the end, explanation building in multiple-case studies was carried out to interpret

the potential surrogate using the surrogacy relationship and protocols for measuring

potential surrogates from interview transcripts.

Figure 4.13. Sample Leximancer concept/theme maps from interview data


PHASE II – QUANTITATIVE SURVEY TO EVALUATE AND RANK

POTENTIAL SURROGATES (ONLINE SURVEY)

In a sequential exploratory mixed method research design, a quantitative

research method is employed based on the qualitative findings. In phase II, an online

survey was used to evaluate potential surrogates identified in phase I. Experts with a

Figure 4.14. Cross-case synthesis diagram adapted from (Yin, 2014)

RO3

To evaluate and select

optimum surrogates for

application by ranking the

potential surrogates against

surrogate evaluation criteria

Figure 4.15. Extract of overall research process for phase II (RO3)


minimum of three years of experience in disaster management were invited to provide

their insights as to how the proposed surrogates can measure the target indicator, since

the aim of phase II was to evaluate potential surrogates against five key criteria to

select optimum surrogates. The research process of phase II to achieve RO3 is shown

in Figure 4.15.

4.6.1 Phase II data collection through online survey

The research framework shown in Figure 4.4 was used to select optimum

surrogates by evaluating and ranking potential surrogates identified in phase I.

Evaluation and selection of surrogates: Survey as a strategy of inquiry

Questionnaire, interviews, and structured observations are some data collection

techniques that can be used in the survey research method (Pickard, 2013). For the

purpose of this study, which aims to obtain multiple expert opinions to select the best

variables based on a set of criteria, since the target respondents are disaster

management experts such as researchers/academics, practitioners, and policy makers,

who are working in different geographical locations, and an online survey

questionnaire was chosen as the most appropriate research method, as it can reach a

larger sample. Hence, reaching experts through an online survey data collection tool

was used for the phase II of this research.

Four key elements were highlighted by Sapsford (2006) in planning for a survey

method in research, as illustrated in Figure 4.16: Problem definition – to decide what

type of answers are required; 2. Sample selection – deciding who/what is to be counted;

3. Design of measurement – deciding what is to be measured and how; and 4. Concern

for respondents – ethical responsibility preventing harm/discomfort to the respondents

(Sapsford, 2006).

Problem definition: Survey questionnaire as a data collection technique

Online surveys are conducted by inviting potential respondents by email to

complete a questionnaire embedded in a web link and completed online (Bryman,

2015). The survey was designed in ‘Key Survey’, an online platform for surveys

provided by QUT, where this study was undertaken. An email was sent with an

introduction to the survey and inviting individuals to the URL (survey link) to

participate, if they agree to participate. The email also requested a response if the user

declined to participate in the survey (“opt out”). Two follow-up reminder emails after


the first publication of the survey were sent to increase the participation in the survey

(Saleh & Bista, 2017), when there was a low response rate after the first follow-up

email after four weeks’ time. The second follow-up email was sent after another two

weeks and the survey was closed eight weeks after the initial launch. The survey was

open from September to October 2018.

Purposive sampling for online survey

The online survey was also used as a purposive sampling strategy to assist in

targeting the desired sample population. The survey respondents were selected based

on their expertise and/or experience, such as people who had a minimum of three years

of experience working in disaster management, because this was the best way to elicit

the views of persons who have specific expertise and experience (Appendix I.1 for the

participant profile questions used). Three cohorts were used in the sampling

population: practitioners, policy makers, and academics/researchers, in order to elicit

their preferences for selecting the surrogates and to explore different perspectives.

There is no standard formula to calculate the required number of samples in non-

probability sampling. According to Oppenheim (2000), 100 respondents are sufficient

for most questionnaire research. However, according to Pickard (2013), 100 is rather

Figure 4.16. Survey design elements (Sapsford, 2006, p. 34)


high for dissertation research. Since guidance was not available in relation to the

sample population number, it was estimated to reach 200 samples with the aim to reach

a sufficient number of responses, aiming for 100. A relatively small number of experts

work in disaster management and particularly in resilience building at sub-national and

national levels in a country (approximately around 20-25 experts in each country),

which is a key limitation to reach a large number of samples.

A total of 208 experts were approached as part of the purposive sampling

approach. The list of 208 email addresses was finalised for this survey, mainly from

three sources: the directory of humanitarian organisations in Sri Lanka, email

addresses from the disaster management forum in Sri Lanka/Regions such as South

Asia, and email addresses from past regional disaster related workshops attended by

researcher.

The survey specifically targeted experts working in disaster management, such

as practitioners and policy makers from UN and NGOs, government officers working

in disaster management, and donor organisations. A list of emails was collected mainly

from the disaster management contact directory in Sri Lanka. However, many experts

who worked in disaster preparedness projects after the 2004 Tsunami in Sri Lanka

have left the country to take assignments in other countries.

Survey respondents/experts’ profile

A total of 66 experts responded to the survey. The response rate for this survey

was 32%. Among the experts who responded to the survey, 84% (n=55) have a Masters

degree qualification, and 69% (n=45) have more than five years’ experience in disaster

management. 56% of the experts were from international/local NGOs/UN agencies

and 25% were from the university/research organisations. The cluster of experts

included 52% practitioners (n=34), 23% policy makers (n=15), and 26% researchers

(n=17), making the responses inclusive and representative of all key segments in the

disaster management domain. In terms of the geographical location of the experts, 33%

were from Sri Lanka and 29% were from the South Asia region excluding Sri Lanka

(Table 4.7). Hence, the responses are largely influenced by a South Asian context, and

share similar socio-economic characteristics.


Table 4.7. Key characteristics and profile of survey respondents

Categories Key characteristics # of experts % of

experts

Educational

qualifications

PhD 15 24

Masters degree 40 60

Bachelors degree 7 10

Other 4 6

Experience

Less than 3 years 8 11

3-5 years 13 20

5-10 years 19 29

More than 10 years 26 40

Employment

Researchers 17 26

Practitioners 34 52

Policy makers 15 23

Affiliation

Government department 7 11

Local NGO/ Community Based

Organisation

6 10

International NGO 23 35

UN agency 7 11

Private sector/donor agency 4 6

Research organisation/institute 5 7

University 12 18

Other 2 3

Location

(country/region)

Sri Lanka 22 33

South Asia (Except Sri Lanka) 19 29

Australia/Pacific 4 7

American continent 5 7

Europe 2 4

Africa 6 8

Middle East 2 4

South East Asia 6 8

Questionnaire

There are many methods of scaling the responses to questions, and the most

appropriate method depends on the way the question is worded (Punch, 2013). The

questionnaire in this survey used the Likert scale item, which asked respondents to

identify the extent to which they agreed or disagreed with a given statement (Pickard,

2013). Likert scale is “a scaling technique that allows respondents to select a choice

that best demonstrates their level of agreement with a given statement” (Pickard, 2013,

p. 213).

In this survey, the respondents were invited to scale their responses on a Likert

scale of 1-5, in terms of the suitability of surrogates in measuring the target indicators

against the five evaluation criteria. Type of scale that is most suitable for this

questionnaire will request the degree of agreement of experts based on their evaluation


of each surrogate for each criterion. Each response can be structured on a five-point

scale response system, because it can produce more information, and respondents can

select their choice in a stable and straightforward way (Punch, 2013; Singh, 2007).

(See Table 4.8 below for scale of responses for each criterion). Appendix G shows the

sample email for participant recruitment, Appendix H shows the participant

information sheet sent with the email, and Appendix I.2 shows the complete set of

questions used in a key survey platform. The questionnaire consisted of three parts:

the first part had eight participants’ profile questions, the second part had 15 questions

in total (each of the three surrogates had five criteria ranking for five indicators), and

the third part had ten criteria pair-wise comparisons. It was estimated to take 20 to 25

minutes to complete the survey.

Table 4.8. Five points scale continuum for each criterion

Criteria for evaluation Five points scale along the scale continuum

1 2 3 4 5

Can measure the target

indicator accurately

Strongly

disagree

(Not

very

good)

Disagree

(Not

good)

Neither

agree or

disagree

(Neutral)

Agree

(Good)

Strongly

agree

(Very

good)

Can be transferred in

different contexts

Can be used to measure in

different phases of a disaster

Can easily communicate the

target indicator

Is cost-effective in measuring

the target indicator

Selection of scale for pair-wise comparison of criteria

A scale is required to indicate the extent to which one criterion is more important

or less important than another (Saaty, 2008), in this case, with respect to evaluating

surrogates to measure social resilience indicators in a disaster context. The nine scale

intensity proposed by Saaty (2008) was reduced to a five scale intensity of importance,

since the intermediate scales may not be needed for comparing five surrogate

evaluation criteria. The five scale is more convenient for survey respondents for

evaluating surrogates against criteria, compared to the nine scale, when the majority

of the survey respondents are not researchers. Table 4.9 below shows the scale used to


indicate the criteria preferences for evaluating surrogates in measuring social resilience

indicators in disaster management.

The pair-wise comparison method was executed in the following steps:

o Five surrogate evaluation criteria were formed into ten pair-wise choices in the

‘Key Survey’ platform as shown in Appendix I.3.

o The disaster management experts selected for evaluating potential surrogates

were also invited to evaluate the criteria in the same survey form; and

o Independent judgement of each expert on criteria was recorded in the key survey

platform on scale of 1–5, as explained in Table 4.9 below.

Table 4.9. Fundamental scale of absolute numbers and the scale used in the survey of

this study adapted from Saaty (2008) (p.86)

Scale in

the survey

Intensity of

Importance

in Saaty

scale

Definition Explanation

1- Equally

important 1 Equal importance

Two activities contribute

equally to the objective

2- Less

important

2 Weak or slight Experience and judgement

slightly favour one activity

over another 3 Moderate

3- Moderately

important

4 Moderate plus Experience importance and

judgement strongly favour

one activity over another 5 Strong importance

4- More

important

6 Strong plus An activity is favoured very

strongly over another; its

dominance demonstrated in

practice 7

Very strong or

demonstrated

importance

5- Extremely

important

8 Very, very strong The evidence favouring one

activity over another is of

the highest possible order of

affirmation 9

Extreme

importance

4.6.2 Phase II – Survey data analysis

In phase II of this study, multi-experts evaluated each surrogate against each

criterion on a five-point Likert scale. Inputs from experts were analysed using Multi-

Criteria Decision Analysis (MCDA) for ranking potential surrogates.


Survey data were analysed by aggregating all completed responses in the online

questionnaire. An initial analysis was carried out on respondent characteristics.

Respondent characteristics included the current or past role of respondents in disaster

management activities, the number of years of experience in disaster management

projects, and the affiliation of past/current portfolios such as a researcher/academic,

practitioner, or policy maker. Finally, response scales from each respondent for each

surrogate against five criteria were aggregated into variables in a multi-criteria matrix.

The values in the matrix were used for multi-expert multi-criteria decision analysis.

Multi-Expert Multi-Criteria Decision Analysis (ME-MCDA)

Cinelli et al. (2014, p. 146) defined Multi Criteria Decision Analysis (MCDA)

in assessing sustainability “as a tool to support the process of decision making by

taking into consideration multiple criteria in a flexible manner, by means of a

structured and intelligible framework.” MCDA is a good tool to make transparent

decisions in a structured manner for problems with complex information (Brinkhoff,

2011). The nine stage MCDA process shown in Figure 4.17 is briefly described below

(Hyde, 2006; Majumder, 2015):

1. Selection of disaster management experts: Experts were selected as per the

sampling criteria discussed in Section 4.6.1.

Figure 4.17. Nine stages of MCDA process adapted

from Majumder (2015)


2. Identification of surrogate evaluation criteria: Five key criteria were identified from

the literature review on surrogate evaluation as described in Section 3.2.3.

3. Identification of surrogates as alternatives: Alternatives are the three potential

surrogates that were identified in a case study research in phase I of this study.

4. Selection of MCDA method: Among many MCDA techniques, Preference Ranking

Organization METHods for Enrichment Evaluations (PROMETHEE) method was

selected in this study for analysis of survey data. PROMETHEE can convert Likert

scale values collected using survey questionnaires into numerical values for

analysis (Carone et al., 2018).

5. Assignment of criteria performance values: The values were five-point Likert scale

(1-5) from very good to not very good.

6. Assign weights to the surrogate evaluation criteria: Initially, equal weight was

assigned for all five criteria to obtain ranking of surrogates.

7. Ranking the potential surrogates against multi-criteria: Visual PROMETHEE was

used to rank the alternatives. Each expert ranking was compiled to Group Decision

Support System (GDSS) feature to obtain the final ranking.

8. Analyse ranking with different weights to the criteria: The criteria weights were

changed to the consolidated criteria weights obtained from Analytical Hierarchy

Process (AHP) method from the pair-wise comparisons of criteria done by the

experts in this survey. The ranking with new weights were compared with equally

weighted criteria ranking.

9. Making final decision: Potential surrogates were ranked to select the best

performing surrogates. Rankings obtained with equally weighted and expert

weighted criteria were compared.

PROMETHEE as a MCDA tool to analyse survey data

Among many multi-criteria decision analysis techniques such as Analytical

Hierarchy Process (AHP), MACBETH (Measuring Attractiveness by a Categorical

Based Evaluation. Technique), PROMETHEE-GAIA, and Technique for Order

Preference by Similarity to an Ideal Solution (TOPSIS), PROMETHEE-GAIA (GAIA

stands for Geometrical Analysis for Interactive Aid as a complementary analysis tool

in PROMETHEE) has been used by many researchers to select the best actions based


on multiple criteria. Verheyden and De Moor (2016) from their research on a process-

oriented social responsibility indicator for mutual funds, found that PROMETHEE

was the most appropriate methodology from the set of MCDA tools in terms of

academic and professional applications.

Table 4.10, produced by Verheyden and De Moor (2016) compares four MCDA

methods against their overall robustness, ease of implementation, transparency and

ease of understanding, and extensiveness of sensitivity analysis in the context of

qualitative indicator applications. PROMETHEE has positive aspects against many

attributes of MCDA methods. The fundamental idea of outranking methods such as

PROMETHEE is that an action is ranked higher than others, if that action performs

better than other actions in the majority of the criteria, and that action does not perform

worse in other criteria (Fernández, 2013). Carone et al. (2018) used PROMETHEE to

rank communities based on the assessment of selected social resilience indicators. In

this study, PROMETHEE was selected as the MCDA technique for the evaluation and

ranking of potential surrogates to measure social resilience indicators. PROMETHEE-

GAIA supports the comparison of alternatives between assessments at different stages

of information (dynamic re-evaluation), which is a major advantage compared to all

other MCDA techniques (Cinelli et al., 2014). The PROMETHEE method is also

useful because it can provide software supported data management and supports

comparison of scenarios for different weights for criteria and their visualisation. Visual

PROMETHEE software was used for this analysis.

Table 4.10. Comparison of Multi-Criteria Decision Analysis (MCDA) methods

(Verheyden & De Moor, 2016) (p.80)

Comparison Attribute AHP MACBE

TH

PROME

THEE

TOPS

IS

Overall robustness for application X X X

Professional application

Extensiveness of sensitivity analysis X

Academic application

Ease of implementation X X X

Transparency and ease of understanding X X X


PROMETHEE ranking

PROMETHEE I ranking provides a partial ordering of alternatives (in this case

potential surrogates), whereas PROMETHEE II method will give full ranking of

surrogates by calculating net flow values. In this study, PROMETHEE II ranking was

used, as it can provide a complete ranking of potential surrogates based on the net

flow. The PROMETHEE II ranking calculation procedure is as follows (Brans & De

Smet, 2016):

Step 1: Formulation of a difference matrix:

A difference matrix is produced between two data points i and i’ from the raw data

inputs. Similarly, the differences for the ith alternative with respect to other alternatives

are determined. This involves determining the differences between different pairwise

alternatives for each criterion:

𝑑𝑗 = 𝑦𝑗(𝑖) − 𝑦𝑗(𝑖′)

Where 𝑦𝑗(𝑖) and 𝑦𝑗(𝑖′) are the data points of alternatives i and i’ for criteria 𝑦𝑗

Step 2: Selection of preference function 𝑃𝑗 (𝑖, 𝑖′):

Some preferential parameters such as preference and indifference thresholds need to

be defined for each function. This is done through the selection of preference

function 𝑃𝑗 (𝑖, 𝑖′). There are six generalised preference functions available in

PROMETHEE for the user to select based on the type of criteria as shown in Figure

4.18: (1) Linear, (2) V-shape, (3) Usual, (4) U-shape, (5) Level, and (6) Gaussian. The

‘Usual’ and ‘Level’ preference functions are best suited for qualitative criteria.

For this research, the ‘Level’ function is used for the net outranking flows for each of

the experts. According to PROMETHEE guidelines (VP, 2013), the ‘Level’ preference

function is a good choice for qualitative criteria such as the 5-point scale if it is needed

to differentiate smaller deviations from larger ones. In the ‘Level’ preference function,

indifference = 0, and there is a strong preference for an action as soon as there is a

difference. If the preference is between 0 and 1, then the preference value is 0.5. For

the Group Decision Support System (GDSS), the ‘Linear’ function is used, since the

net flow is a quantitative value.

Step 3: Calculate the leaving and the entering outranking flows as given below

Leaving (positive) flow for ith alternative


∅+(𝑖) =1

𝑚−1 ∑ 𝜋(𝑖, 𝑖′) (𝑖 ≠ 𝑖′)𝑚

𝑖′=1

Entering (negative) flow for ith alternative

∅−(𝑖) =1

𝑚−1 ∑ 𝜋(𝑖′, 𝑖) (𝑖 ≠ 𝑖′)𝑚

𝑖′=1

The leaving flow states how much an alternative dominates the other alternatives,

while the entering flow expresses how much an alternative is dominated by the other

alternatives.

Step 4: Determine the net outranking flow:

For each alternative to rank the alternatives using PROMETHEE II (complete)

ranking, a net outranking flow is calculated as follows.

∅(𝑖) = ∅+(𝑖) − ∅−(𝑖)

Step 5: Determine the rankings:

Figure 4.18. Preference functions used in PROMETHEE (Hyde, 2006; Kilic

et al., 2015).

Legends in the graphs (P(x) is preference, x is difference, x1 is indifference

threshold, x2 is preference threshold)


The ranking is determined of all the considered alternatives depending on ø (i) values,

which is the net phi values. The best performing surrogate (alternative) will have the

highest ø (i) value according to PROMETHEE II (complete) ranking. In this study,

PROMEHTEE II ranking was employed using the net flow calculation as shown in

step 4.

Step 6: Group Decision Support System (GDSS):

The potential surrogates were ranked to select the best performing surrogates.

Multi-expert group decision flow chart shown in Figure 4.19 depicts the flow of inputs

from experts to the final decision-making matrix using PROMETHEE Group Decision

Support System (GDSS) algorithm (Brans & De Smet, 2016; Ishizaka & Nemery,

2013). Initially, the Likert scale evaluation value of each expert for each surrogate

against the five criteria was entered into PROMETHEE independently (the expert

Figure 4.19. Multi-expert multi-criteria group decision support system

flowchart


evaluation value was based on 1-5 Likert scale). By running the PROMETHEE

analysis, potential surrogates were ranked for each expert and net flow values for each

surrogate was obtained.

In the next step, GDSS algorithm was implemented in PROMETHEE to obtain

final ranks by taking each expert as a criterion and each surrogate as an alternative.

The net flow value was then entered into PROMETHEE again for each surrogate

against all 66 experts as criteria. The final ranking was obtained by performing the

multi-criteria decision analysis in PROMETHEE again from steps 1 to 5, explained

above, using the experts as criteria.

Step 7: GAIA (Geometrical Analysis for Interactive Aid) analysis:

Further to the PROMETHEE ranking, the GAIA (Geometrical Analysis for

Interactive Aid) in Visual PROMEHTEE software provides a complementary visual

analysis of the results produced in PROMETHEE. Each surrogate is represented by a

point in the GAIA plane, and its position is related to its evaluations on the set of multi-

criteria in such a way that actions with similar profiles are closer to each other. Figure

4.20 shows an example of PROMETHEE inputs for actions (actions 1 to 4) in the 1-5

Likert scale against four Decision Makers as criteria used in GDSS (DM1 to DM4),

their ranking based on net flow (Phi), and GAIA plot representation of four actions

and criteria. The GAIA plane plot helps to analyse the actions and criteria (Decision

Makers) as follows (based on the example in Figure 4.16) (Brans & De Smet, 2016):

- Similar actions are placed close to each other and vice-versa (for example, action

1 which is ranked last, is represented far from other actions in Figure 4.20)

- Vectors pointing to the same direction as the decision axis (red arrow) have

positive correlation, and the opposite direction has negative correlation (for

example DM2 has conflicting views compared to DM1, DM3, and DM4).

In this study, GAIA representation visuals were used to undertake two more

distinctive analyses based on the evaluation of potential surrogates to see: (1) how

different years of experience and experts from different cohorts aligned with the

overall PROMETHEE ranking results; (2) how their preferences varied with the

overall PROMETHEE ranking. For the first analysis in GAIA, the experts with varying

number of years of experience were clustered into four cohorts: experts with less than

three years of experience (< 3Y), three to five years of experience (3-5 Y), five to ten


years of experience (5-10 Y), and expert with more than 10 years of experience (10

Y). Similarly, for the second analysis in GAIA, the experts were clustered into three

cohorts based on their work affiliations: practitioners, policy makers, and researchers.

Criteria weight calculation

Analytic Hierarchy Process (AHP) is an MCDA technique with the ability to

provide a consensus based on the individual expert judgments, mostly done through

assigning weights for indicators and criteria. AHP has been used by many researchers

to determine weights for a set of indicators; for example, Alshehri et al. (2015b) used

Figure 4.20. Sample PROMETHEE inputs, ranking, and GAIA plot


the AHP method with disaster management experts to determine weights for different

dimensions and characteristics of community resilience to disasters. In this study, AHP

was used for determining the weight for the surrogate evaluation criteria.

Determining the criteria weight using Analytic Hierarchy Process (AHP)

In order to determine the weights for the surrogate evaluation criteria in a

participatory method, the disaster management experts who evaluated potential

surrogates were also asked to provide their preferences for pair-wise comparisons of

the five surrogate evaluation criteria. The results obtained in the key survey platform

were analysed using the latest Analytic Hierarchy Process (AHP) tool by Goepel

(2013), which is being increasingly applied in the assessment of sustainability studies

(Markelj et al., 2014; Petrini et al., 2016). The analysis of inputs obtained in the survey

to determine criteria weights was done in the following steps:

o The pairwise comparisons of each expert were recorded into a matrix developed

in Excel by Goepel (2013) to determine the weights using AHP process.

o The Consistency Ratio (CR) was calculated to establish the consistency of the

judgements provided by the experts.

o The judgements with the reasonable consistency ratio (CR < 0.2) were selected

and aggregated into a separate group decision making process in AHP.

o The final weight for the criteria was obtained and used in the PROMETHEE

multi-criteria decision making tool to evaluate the potential surrogates to measure

social resilience against these weighted criteria.

o PROMETHEE ranking was obtained for a set of potential surrogates to measure

five selected social resilience indicators with the new criteria weights and

compared with the PROMETHEE ranking obtained with equal criteria weights.

CR < 0.2 is an acceptable and permissible level of consistency (Ho et al., 2005;

Park & Youngchul Kim, 2014; Wedley, 1993), though Saaty (1990) recommends CR

< 0.1. Consistency becomes a concern due to the limited capacity of humans to keep

their judgements in pair-wise comparisons consistent with more than three criteria

when executing AHP (Asadabadi et al., 2019).

In this survey, the CR value for 14 respondents among the total of 66 survey

respondents was < 0.2. The consolidated AHP results of 14 responses showed a 71%

consensus rate of pair-wise comparisons of criteria. Since the survey in this study was


done only once, it was not possible to request responses again so that pair-wise

comparisons could be improved to increase consistency in their responses. Therefore,

it was decided to select the responses which had CR < 0.2 to determine the

consolidated weight for the surrogate evaluation criteria.

For example, in a Focus Group Setting, five to eight decision makers are

acceptable for AHP decision making (Udie et al., 2018). Alshehri et al. (2015b) invited

16 respondents and Udie et al. (2018) invited 19 respondents through a survey using

AHP to determine the weights for resilience dimensions in a disaster context, and to

assess vulnerability to climate change in critical infrastructure. The sample size is not

a limitation to carrying out AHP, as different studies have used very small sample

sizes, and there are many studies with smaller purposeful samples. Since this study

combined the PROMETHEE tool for ranking potential surrogates and AHP to

determine the surrogate evaluation criteria weights, the number of respondents for both

survey components were the same, 66. However, only the CR of 14 respondents was

less than 0.2, which is an acceptable level of CR determined in this study as explained

above.


SUMMARY

This chapter discusses a mixed method research design and sequential

exploration as a research strategy for this study. This research study was carried out in

two key phases. This chapter provided an overview of research philosophy and

methods, mixed method research strategy, data collection tools, and finally the method

of data analysis for each phase.

To achieve Research Objective 1 (RO1), five key social resilience indicators

were selected from the literature review and research design phase to apply surrogate

approach from the ‘5S’ adaptive and inclusive social resilience framework developed

in this study, based on set of surrogate decision criteria (Section 2.3). The ‘5S’ social

resilience framework consists of many indicators that are either not easily measurable

or if measured, would have used data obtained from publicly available census data,

which most often do not provide an adequate measure for process-oriented resilience

indicators.

In the qualitative phase of the research (Phase I - Interviews), to achieve

Research Objective 2 (RO2), a case study utilising interviews was adopted to explore

potential surrogates by consulting disaster management practitioners and policy

makers in the selected case study locations. The data collected from four case study

areas in the Eastern coast of Sri Lanka were analysed using Leximancer – a text data

mining tool that generates concepts and a themes map. A set of higher-order themes

were identified as key facets of the target resilience indicator, which can be used as

potential surrogates to measure each of the social resilience indicators.

In the quantitative phase of the research (Phase II - survey) to achieve Research

Objective 3 (RO3), an online survey was used to evaluate the potential surrogates

identified for the five selected social resilience indicators in phase I. A number of

disaster management experts, mainly from the practitioner, research and policy making

community, were invited to evaluate the potential surrogates against five key surrogate

evaluation criteria. The responses were then analysed using PROMETHEE – a Multi-

Criteria Decision Analysis (MCDA) method, to rank the potential surrogates based on

their overall performance against all five surrogate evaluation criteria.

Chapter 5: Interview analysis-Identification of potential surrogates 115

Chapter 5: Interview analysis-Identification

of potential surrogates

This chapter is divided into six key sections to present key findings from phase I of

this study (interviews) as shown in the chapter structure in Figure 5.1. Interviews were

utilised in Phase I of this study to identify potential surrogates to assess five selected

social resilience indicators.

Sections 5.1 to 5.5 present potential surrogates for the following five indicators:

Social mobility and access to transport in a disaster context (Indicator #1);

Social trust in a disaster context (Indicator #2);

Learnings from the past disasters as social competence (Indicator #3);

Involvement of people with specific needs as social equity (Indicator #4); and

Cultural norms/behaviours as social belief (Indicator #5).

The analysis of interview data for each indicator is structured in each section as

follows:

1. Higher-order themes for case study #1 (Kalmunai T division)

2. Higher-order themes for case study #2 (Sainthamaruthu division)

3. Higher-order themes for case study #3 (Kalmunai M division)

4. Higher-order themes for case study #4 (Sub-national level)

5. Final set of surrogates identified from cross-case synthesis: A brief discussion

about the potential surrogates, their relationship to the target indicator, and

measurement protocols to assess each indicator are provided.

6. Summary of findings

Final section 5.6 presents a summary of the overall findings and includes a final set of

six potential surrogates for each of the five social resilience indicators. This section

also provides a table of three selected potential surrogates for evaluation in the next

phase (Phase II) through an online survey.


Figure 5.1. Chapter 5 and key sections in thesis structure


SURROGATE MEASURES TO ASSESS SOCIAL MOBILITY AND

ACCESS TO TRANSPORT IN A DISASTER CONTEXT (INDICATOR #1)

The surrogate development framework was operationalised to identify potential

surrogates for measuring the first social resilience indicator selected in this study –

social mobility and access to transport in a disaster context. Social mobility is an aspect

of movement of people individually or collectively in a disaster situation through

available means such as transport facilities (Yamamoto et al., 2018).

5.1.1 Higher-order themes for case study #1 - Kalmunai T division (KMT)

In case study #1, 12 themes were generated and four combined themes (higher-order

themes) were identified (Figure 5.2) based on connected concept nodes in Leximancer

concept maps. These include:

1. Awareness programs and early warning drills;

2. Social support through disaster management committees;

3. Evacuation places;

4. Transport for people with specific needs.

Figure 5.2. Leximancer concept/theme maps from interview data

of case study #1

3

4

1

2


(1) Awareness programs and early warning drills: Many interview respondents

highlighted that people who have higher level of awareness attend early warning

and evacuation drills annually. The awareness programs focused on people with

specific needs, who are the priority during evacuation. The contents of awareness

programs include early warning response, evacuation routes/mode and procedures.

Three key concepts identified in this theme are (as shown in box 1 in Figure 5.2): a)

frequency of awareness programs; b) early warning drills focusing on evacuating

people with specific needs; and c) information dissemination sources (such as police

and trusted sources).

“Social service department has given awareness programs for the family

members and parents of the people with specific needs. They will be

transported through vehicles such as three wheelers” (Interview #41)

“If we issue an early warning for evacuation, people have public transport

facility in the main road. There is only one village that is problematic in terms

of evacuation because it gets isolated. But we have boats available in this

location” (Interview #04)

(2) Social support through disaster management committees: This theme

mainly refers to the support from the disaster management committees for people to

access transport facilities and other mobility assistance during disaster evacuations.

The most vulnerable people seek social support from neighbours or from the

community based organisations or committees such as disaster management

committee in their area. People also request evacuation assistance from state

institutions such as from the divisional area office or other essential service

departments such as hospitals or from the police.


availability of emergency support services (water, first aid, public transport such as

buses); b) availability and use of emergency numbers by people in the area; and c)

availability and active DMC committees in the division.

“From the disaster committees there are arrangements to get access to vehicles.

For sick people or emergency cases, accidents, or disasters, we can access

vehicles from hospitals and police. Since there is a hospital and military forces

are available within our division, we can get vehicles. There are officers from


the government departments who are also disaster committee members”

(Interview #27)

“Private vehicle owner’s data is available from us. We also have given numbers

to public transport providers. When a disaster happens, they were advised to

provide their support by providing vehicles” (Interview #28)

(3) Evacuation places: This is a stand-alone theme in this case study. The

identification of evacuation centres with adequate emergency facilities when an

early warning is issued, is a key factor for timely evacuation decision making. The

importance of awareness level among the public on where to evacuate in the event

of an early warning was also highlighted.


evacuation maps available in the division with evacuation centres marked; b)

number of evacuation places identified which have emergency facilities such as

toilets for people with specific needs; and c) awareness among people of the

availability of the evacuation places in their respective area.

“With the participation of the people, we have developed evacuation plans and

maps to identify safer places” (Interview #4)

“When disaster occurs, the government has decided and demarcated places of

accommodation during evacuation. Disaster Management Centre in some

evacuation centres, there are toilets and other facilities made available in the

existing evacuation places” (Interview #41)

(4) Transport for people with specific needs: Vehicles targeting people with

special needs is important for effective mobility in times of disasters. The priority

should be given to evacuating people with specific needs, and hence their transport

facilities need to be measured.


availability of own transport facilities for use by people with specific needs; b)

government arranged transport facilities such as emergency service vehicles; and c)

availability of vehicles with organisations for transporting people with special

needs.


“Disaster management committee in each division can organise some transport

facility for the people with specific needs such as elders, children and disable

people based on the transport services and facilities available within the area.

We can’t provide transport facilities to all people and we have to give priority”

(Interview #34)

“Big vehicles are not very much available and these big vehicles cannot be used

during disasters because we don’t have wider roads to use bigger vehicles

during disaster time, because of congestion of vehicles. Push bicycles or motor

bikes are the used widely for transportation at any time and particularly during

disasters. There are some cases with people with specific needs who cannot be

transported by motor bikes” (Interview #41)

5.1.2 Higher-order themes for case study #2 – Sainthamaruthu division (SM)

In case study #2, 10 distinct themes were generated. By combining closely related

themes, four higher-order themes were identified (Figure 5.3). These included:

1. Awareness and early warning;

2. Evacuation places with facilities for people with special needs;

3. Transport facilities and vehicle owners;

4. Narrow roads.

(1) Awareness and early warning: This higher-order theme is similar to the higher-

order theme (1) in case study #1.


disaster preparedness awareness programs/drills conducted annually by the disaster

management committee of the area; b) how many people participate in each of the

awareness programs/drill; and c) how many people carry emergency bag during the

drill.

“Evacuation routes and maps have been displayed in many villages. This

awareness has been done through programs as to how to evacuate and where

to evacuate. This was done using pictures and they were also given

instructions” (Interview #6)


“There are volunteers and youth groups in each GN division. We have provided

awareness training to them on how to evacuate people with specific needs. In

each GN there is a disaster management committee, and youth and sports clubs.

Elders, pregnant mothers, were helped by volunteers to move them to evacuation

centres” (Interview #40)

(2) Evacuation places with facilities for people with specific needs: This theme has

emphasised the importance of evacuation places with facilities for people with specific

needs. When people are aware about evacuation places that were already identified,

have emergency facilities, they are more likely to evacuate. Timely evacuation

requires access to transport facility, importantly for the people with specific needs.


number of evacuation places available; b) population of specific needs; and c) ratio of

evacuation places equipped with specific needs to population of specific needs.


of case study #2

1

2

3

4


“We have an evacuation centre now in our division. It is now being constructed

with facilities for special needs…… In the past we used schools and mosques as

evacuation centres. But now we can use this new evacuation centre built for the

purpose of accommodating people when disasters occur” (Interview #12)

(3) Transport facilities and vehicle owners: The vehicle ownership and their

willingness to support during disasters emerged as inter-related themes in this case

study. Access to transport facility for timely evacuation is linked to the vehicle

ownership and transports available in the area that could be accessed by the people

who do not own a vehicle.


type of vehicles owned by the local authorities that can be used during disasters; b)

public/common vehicle in the area; and c) mapping of emergency vehicle availability.

In case of flood, the availability of boats become important.

“Vans and bus owners are there in my area, a bus owner with five buses have

given his buses for evacuation services with free of charge as a voluntary

contribution. Similarly, tractor owners also have done the same. If we can get

the details of vehicle owners within the community to mobilise help for transport

facility during disasters” (Interview #6)

(4) Narrow roads: The theme on community infrastructure that are important for

social mobility, such as road networks and bridges connecting roads has emerged in

this case study. A vehicle needs to reach to the place of a person with specific needs

such as elders or disabled who live in the vulnerable area in order to evacuate them.

Similarly, the identification of proper evacuation routes are important, because some

roads may be narrow and larger vehicles cannot be used.


road conditions (narrow or wide); b) road network for effective evacuation; and c)

road features such as bridges and drainage.

“There are problems to use vehicles in massive scale due to narrow roads.…..

For example, if a person has a van, when they evacuate they will take some more

people with them. For cars and motorbike, they have to evacuate alone. Many

roads are narrow, and our bridges are very narrow that is not enough in a large

evacuation” (Interview #40)


“…. we have done an evacuation and response plan in Sainthamaruthu and we

have also identified safety places for evacuation. Based on this we have made

awareness among people about evacuation routes and places. When there is a

disaster early warning, people will move towards safety centres and identified

and demarcated evacuation centres” (Interview #9)

5.1.3 Higher-order themes for case study #3 – Kalmunai M division (KMM)

In case study #3, 10 themes were generated and four combined themes were identified

as higher-order themes (Figure 5.4). These include:

1. Vehicles available in the community;

2. Evacuation roads;

3. Evacuation in mosques and schools;

4. Early warning.

(1) Vehicles available in the community: This theme highlights the existing vehicle

capacity in the geographical area or region that can be utilised during disaster

evacuation. It is important to know the location where vehicles for emergency use can

be obtained.


mapping of state emergency vehicle availability in the division; b) availability of

community managed vehicles; and c) private vehicle owners that can be used during

emergencies.

“We should check whether there is a local level mapping for availability of

vehicles and resources at the community level. Where the vehicles are available

and how many are available in each village. We have to check if the available

vehicles have the provisions of evacuating people with special needs” (Interview

#42)


(2) Evacuation roads: Access to transport facility depends on the capacity of the

roads to cater to an evacuation during a potential disaster. This theme is related to

evacuation routes and planning. Evacuation by means of a vehicle also depends on the

road network and conditions of roads in the area.


evacuation route sign boards displayed on roads; b) availability of evacuation route

maps and plans with local authorities; and c) key road infrastructure availability such

as large bridges.

“We have to first talk about roads for effective evacuation. When people

evacuate, if there are any problems from the private lands that create problems

to roads, we provide compensation to those private owners and get the land for

roads to make the road network effective and smooth during disasters”.

(Interview #25)

(3) Evacuation to mosques and schools: This theme is related using public buildings

such as places of worships and schools in times of disasters. The availability of

facilities in mosques and schools for use during evacuation time is important if they

are demarcated as evacuation places.

Figure 5.4. Leximancer concept/theme maps from interview data of

case study #3

1

2

3

4



number of evacuation places available (Mosques/schools/public buildings/designated

buildings); b) awareness among the public of evacuation places; and c) ratio of people

evacuated and stayed in the public buildings as opposed to in houses of relatives.

“…. People went to mosques and schools, when there was a flood. The house

heads stayed back and removed belongings and safe guarded their belongings

from flood water. People with specific needs were taken by the family members

to mosques and schools” (Interview #29)

(4) Early warning: Early warning is emphasised by many interview participants in

case study #3 as an important facet for decision making for access to a transport facility

during disasters.

Four key concepts identified in this theme are (as shown in box 4 in Figure 5.4): a)

availability of early warning systems (number of early warning towers); b) public

addressing system such as loud speakers in religious places (example mosques and

temples); c) geographical coverage of early warning messages i.e. the percentage reach

to the population at risk; and d) drills conducted to test the performance of early

warning systems and messages.

“Early warning towers are built and early warning equipment are given to local

authorities. The towers were regularly tested and rehearsal programs are being

held frequently. People are trained and awareness raising programs were also

held. Evacuation boards are displayed as to how people should be evacuated”.

(Interview #32)

5.1.4 Higher-order themes for case study #4 – Sub-national level

In case study #4, nine themes were generated and four combined higher-order themes

were identified (Figure 5.5). These included:

1. Evacuation places in the area

2. Early warning systems and plan

3. Availability and access to transport facilities

4. Emergency information sources


(1) Evacuation places in the area: This theme highlights the importance of

designated safe areas as part of a disaster preparedness plan to increase the resilience

of communities.


number of evacuation places available (religious places/schools/public buildings); b)

availability of evacuation maps with evacuation places in public (roads/public places);

and c) availability of government demarcated safer areas in evacuation plans.

“People go to near-by places, common places like a temple, school,

community hall, and temporary centres. There are places that are identified

by Disaster Management Centre and village disaster management

committees (VDMC) for people to evacuate in times of disasters” (Interview

#17)


of case study #4

2

3

4

1


(2) Early warning systems and plan: The theme on available early warning

mechanisms and proper execution plan when there is a disaster, emerged in case study

#4, which was pointed out by many interview respondents.


availability of early warning systems (example, Tsunami early warning towers); b)

geographical coverage of reach to the population at risk; and c) testing/drills using

early warning towers (annual Tsunami Remembrance Day drill).

“Early warning group member details should be available at the DS and village

level, which is an important indicator. The availability of contact directory of

early warning group in GN divisions and they were given identity cards is

another indicator. There is also a software system with contact details of key

people in the early warning group at the division level” (Interview #5)

(3) Availability and access to transport facilities: One of the key factors to

understand the social mobility is to know the availability and access to transport

facilities in time of disasters. It was emphasised that vehicles should be driven through

narrow roads and avoid major traffic congestions during disaster evacuation.


availability of public transport (buses) and other special transport systems during

disasters (boats); b) private vehicles such as tractors or ambulances from hospital or

CBOs, in cities, school vans and three-wheelers; and c) access to smaller vehicle such

as three-wheelers.

“We do resource mapping for a village. We also collect information such as how

many vehicles and where they are in the households, how many tractors, how

many carts are and where they are available. When we do village development

plan, we include these information in emergency preparedness plans”

(Interview #5)

(4) Emergency information sources: Access to information is key to improving the

mobility and access to required transport facilities during disaster warnings. The

availability of contact numbers of emergency vehicles in the community disaster

management plan can be one of the measures of access to transport facilities when

disasters occur.



emergency contact details and directory (available in the disaster preparedness plan

document); b) early warning groups in the village or disaster level disaster

management committees; and c) availability of early warning systems (mobile apps

and towers).

“Nowadays the information system is dependent on mobile phones.….. Disaster

Management Committee, will have an information on how to access transport or

bus. Disaster Management Centre has some information district wise and people

can call DMC and ask them for help to access transport” (Interview #13)

“We can get those information it from researches or assessments. They are

posted in the divisional offices. The numbers that are needed to contact to get a

vehicle during an emergency are sometimes available in public or posted in

administrative offices. But it is very rare to display in community. We can do a

mapping out what are the emergency vehicles available, where to get them, and

who to contact, and how to contact them to get those vehicles. This should be

displayed in common places where people often go or access to their needs to

such as library, divisional offices, and local authorities and village divisional

offices” (Interview #3)

5.1.5 Final surrogates identified from cross-case synthesis for indicator#1

Table 5.1 below summarises the mapping of higher-order themes (taken as surrogates)

which emerged from thematic analysis of interview data of four case studies.

Table 5.1. Surrogates (higher-order themes) mapping for indicator #1

Case study

locations/area

Surrogates to measure social mobility in a disaster

context

S11 S12 S13 S14 S15 S16

Case study #1

(KMT) X X X X

Case study #2

(SM) X X X X

Case study #3

(KMM) X X X X

Case study #4

(sub-national

level)

X X X X

S11: Available transport facilities (targeting people with specific needs)


S12: Evacuation places and centres (including facilities for people with specific needs)

S13: Awareness programs/plans and early warning systems

S14: Evacuation routes and plans

S15: Social support systems (incl. disaster management committees and government)

S16: Emergency information dissemination and sources (including telephone numbers

to access transport)

For each of the potential surrogates, a brief description of the surrogate, its relationship

with the target indicator, and assessment protocols proposed by interview participants

from the synthesis of interview responses are summarised below.

1. Available transport facilities (targeting people with specific needs) (S11): This

surrogate was highlighted in all four case studies. Access to a transport facility during

a disaster should be planned to deal with the future disasters effectively. Timely access

to a required transport facility is necessary for effective evacuation and mobility during

and after disasters. The availability of transport facilities or timely accessibility for

transport facilities for People with Specific Needs (PwSN) was highlighted by many

interview participants in the case studies.

Further, the measurement of surrogate (S11) should include the assessment and

mapping of the ratio of vehicle ownership to population, availability of vehicles to

‘transport PwSN’ to the ‘PwSN’ ratio, and the ratio of vehicles available with

government institutions and private owners that can be used during disasters.

2. Evacuation places/centres (including facilities for people with specific needs)

(S12): This surrogate was highlighted in all four case studies. The existence of

demarcated evacuation places or purpose-built disaster evacuation centres can help the

population at risk to execute a planned evacuation during disasters. The level of

awareness among the people about the facilities at the evacuation centres will influence

the mobility of people with special needs who are the least resilient segment of the

population.

The surrogate (S12) measurement protocol can include: assessing and mapping

of evacuation places available (places of worship/schools/public buildings/designated

buildings) and its capacity, ratio of evacuation places to how many are equipped to

handle people with special needs, and the ratio between the total capacity of evacuation

places and the population at risk.


3. Awareness programs/plans and early warning system (S13): This surrogate was

also highlighted in all four case studies. People at-risk should be made aware about the

mobility and accessibility to transport facilities when required during a disaster

warning or response. Awareness programs can influence the way people evacuate and

the resilience of the community to disasters by planning for effective evacuation during

disasters. The dissemination of early warning messages at the right time, to the

population at risk can help people to evacuate to safer places. Effective risk

communication result in greater mobility and allow people to prepare for timely

evacuation.

The surrogate (S13) can be measured by mapping and assessing: the existence of

early warning mechanisms such as towers, sirens, and public address systems in places

of worship and schools, annual disaster evacuation drills (ratio of people who

participated to the population at risk and number of people evacuated carrying

belongings), number of training and awareness programs related to evacuation

conducted annually in places of worship and schools, and participation trends in

training and awareness programs conducted related to evacuation.

4. Evacuation routes and plans (S14): This surrogate was highlighted in case studies

#2 and #3. The existence of evacuation routes and plans help a community to be

resilient when disasters occur, so that the community can respond to the event

effectively. Most often the evacuation plan as part of the bigger disaster management

plan identifies the requirement for transport facilities for evacuating people in the event

of a disaster. The understanding of evacuation access routes and multiple evacuation

access points will contribute to greater mobility of the community as a whole, as

highlighted by many interview participants across the four case studies.

The surrogate (S14) can be measured by mapping the key access roads including

multiple access points that are connecting to major highways, and assessing the road

conditions, road infrastructure, and evacuation signage with instructions. Further,

village or street disaster management committees and existence of evacuation plans

available with these committees can also be assessed.

5. Social support system (including from disaster management committee and

government) (S15): The social support system (support through disaster management

committees and government officials) was selected as the fifth surrogate from case

study #1. The support for accessing transport and assisting in evacuation after disaster


warning from the neighbouring community and other support systems such as police

and security forces is another important factor for greater mobility. The support for

accessing transport and assisting in mobility of people to evacuate from emerging

disasters may also come from the community itself. People who are not at risk from

the emerging disaster, such as people living in less vulnerable areas within the same

community may come forward to assist people, who are living in highly vulnerable

areas.

The surrogate (S15) can be measured by assessing the resource maps available in

the village based disaster management plans, the existence of annually elected village

level disaster management committees, and the activities of disaster management

committees, helping to increase the mobility of people during disaster evacuation.

6. Emergency Information dissemination and sources (including emergency

telephone numbers of transport facilities) (S16): The emergency information sources

were selected as the sixth surrogate to measure access to transport facility from case

study #4. The dissemination of information to people at risk is an important factor for

timely access to a transport facility for evacuation during a disaster early warning.

Access to information such as emergency telephone numbers of the key organisations

involved in disaster response and recovery is needed to access required resources such

as emergency transport facility for evacuation.

The surrogate (S16) can be measured by sampling the area based social media

platforms/websites/groups, and assessing the active disaster information dissemination

services and programs, number of people who use the internet or other emergency

communication devices such as satellite phones and their usage during the disaster

warning period, and the usage of area based emergency call number services for

disaster information and requests for emergency transport facilities.


5.1.6 Summary of findings for indicator #1

The study initially identified 83 concepts and 41 themes from the four case

studies for indicator #1. Similar themes were then aggregated and six higher-order

themes were identified through cross-case synthesis which can be considered as

potential surrogates to measure social mobility. The summary of the synthesis is shown

in Figure 5.6.

The following three surrogates – S11: Transport facilities available (targeting

people with specific needs), S12: Evacuation places and centres, and S13: Awareness

programs and early warning systems, have higher validity, since they were identified

across all four case studies.

Another set of relevant higher-order themes (S14: Evacuation routes and plans,

S15: Social support systems, and S16: Emergency information sources) were found in

at least one of the case studies, can also be potential surrogates to measure social

mobility in a disaster context. Hence, six potential surrogates to measure social

mobility are highly reliable and practically applicable in similar contexts, since they

were identified in consultation with practitioners and policy makers who are highly

experienced in implementing disaster management activities at the community level.

Figure 5.6. Summary of synthesis for first social resilience indicator –

social mobility and access to transport facilities


SURROGATE MEASURES TO ASSESS SOCIAL TRUST IN A

DISASTER CONTEXT (INDICATOR #2)


surrogates for assessing the second social resilience indicator selected in this study –

social trust in a disaster context. Social trust in a disaster context is a phenomenon

which influences the behaviour of people with other social groups and associations in

the community in disaster management activities (Thoresen et al., 2018).


In case study #1, nine themes were generated and four combined themes (higher-order

themes) were identified (Figure 5.7) based on connected concept nodes in the

Leximancer concept maps. These include:

1. Performance of CBOs or Effectiveness of CBO activities;

2. Public-Government officer relationship;

3. Government office/local authority support for people;

4. Effectiveness of disaster relief work.

(1) Performance of CBOs or Effectiveness of CBO activities: When discussing

social trust in disaster management work, the role of Community Based

Organisations (CBOs) such as Rural Development Society (RDS) is very much

highlighted in the context of this study area.


trend of people in the village participates in CBO meetings; b) how many CBOs are

active and work in relation to disaster management work; and c) percentage of CBOs

participate in disaster preparedness work such as disaster early warning testing or

drills conducted annually.

“From the Rural Development Department, we can get information. We have

at least one CBO, Rural Development Society (RDS) in each division. If RDS is

inactive, then we use Women RDS (WRDS) if it is active. If WRDS is inactive,

we go for Sports club, youth club likewise. Most often we work with RDS and

WRDS, but if they are inactive we will have to approach other CBOs. They


follow up with the active CBOs. By this, we can measure the level of trust within

the community since the CBOs are within the same community” (Interview #27)

“We can measure trust based on the active and helping NGOs. The

organisations continue to be active, like SWOAD, World Vision who are

working all the time. Around, there are 65 CBOs registered. RDS, WRDS,

Sports clubs and youth clubs, and other clubs registered under departments.

Only 10-12 organisations are active. Based on their work we can measure the

trust. When disaster happen, RDS and WRDS and Civil Defence Group are the

three key CBOs in a village, who are already prepared to face disasters. They

have been elected by the people and they represent the people and they attend

to the issues when people need it. The society accepts their work since they work

for the society” (Interview #28)


case study #1

2

1

4

3


(2) Public-Government officer relationship: This theme mainly refers to the level of

trust by the people on government systems and mechanisms which largely depends on

the relationship of government officers with the public. The first point of contact for

many people is the respective village administrator, who is a public servant employed

by the state to identify peoples’ needs at all times.


the level of response from the public when Divisional Secretary (DS) office requests

support or help (fund and relief collection) people affected by disasters in other areas;

b) trend in people attending meetings/campaigns convened by the government officers;

and c) response rate to the requests for community help to resolve an issue (e.g.

compensation request for disaster damage).

“Public believe government officers more. People have much hope that the

relief and recovery needs will be fulfilled. People can get clear information only

from government officers. Some organisations go to some selected places. In

those selected locations, they only select few people and provide assistance.

This creates problems and leave out many people who are in real need. 70-80%

of the people who were affected did not receive the relief support. Other

organisations do not have much information and it is with the government

officers who are in the field. We received many complains about this in the past

during this relief distribution. Then we addressed these problems. Since we

have many information, so later NGOs approached us to work in a village to

support people who were affected by Tsunami” (Interview #41)

(3) Government office/local authority support for people: The concepts emerged

in this case study have given focus on the positive efforts by government officers to

build trust. Although CBOs help people faster in disaster response, all activities are

coordinated by the government departments and local authorities. CBOs and other

organisations have to work under the government leadership, sometimes a joint

leadership by civil society and state administration, particularly in disaster

management.


people assisted by government from the list of people identified with needs; b) budget

allocation from the state/local authority against required finances for the projects; c)


number of programs conducted/projects implemented by state/local authority from

the list of identified projects.

“For government, it has to work with circulars and it has its own capacity

limitations. Government has the responsibility to protect people from the

calamities. People think and believe that the government will somehow help

them at any point even after four months, because it is a permanent solution.

NGO solution is very short term and they will finish and go. People also know

that the government will not do things immediately” (Interview #39).

(4) Effectiveness of disaster relief work: The involvement of disaster related work

rely on the level of trust in the disaster management system. The trust largely

influences how the responsible stakeholders working in disaster management to

effectively implement their projects and coordinate with other stakeholders.


number Organisations actively working in disaster management; b) percentage of

assistance collected to support disaster victims outside, against the assistance within

the community (budget allocations by community based organisations); and c)

number of trained active volunteer base involved in awareness programs/information

dissemination.

“During Tsunami time, many help were done by NGOs. We guided and

facilitated NGOs through GN (Village divisions) and then DS (Divisional

administration) office. When people were staying in displacement centres, we

gathered information from people and held meetings with people and then with

NGOs to provide collected information. Complaints have come after Tsunami

disaster. Some organisations have given relief items to some segment of the

people created some suspicion. Some personal complaints have come due to

this suspicion. We guided and facilitated to solve them through DS office”

(Interview #41)

“During evacuation, disaster relief, and relief collection, people completely

collaborate. 90% of the people participate during the collection of money and

relief items for supporting disaster affected people in neighbouring villages or

other places. We do not only provide assistance to people but, also collect

money and relief support for other disaster affected people” (Interview #27).



In case study #2, eight distinct themes were generated. By combining closely related

themes, three higher-order themes were identified (Figure 5.8). These include:

1. Effectiveness of CBO activities/social service;

2. Level of support for awareness programs;

3. Government office (Divisional Administration Office) support for people.

(1) Effectiveness of CBO activities/social service: This theme is similar to the

theme (1) in case study #1.


active number of CBOs out of registered CBOs, number of projects implemented over

the past five years by active CBOs; b) membership and volunteer base of CBOs; c)

funding capacity of active CBOs; and d) reserve funding that can be utilised during

disaster time.

“CBOs, most of the time involve in relief work and community development

work. For disaster management purposes, there are local social service

organisations. There are other social service organisations too. DS office has

Rural Development Societies, Women Development Societies, Youth and Sport


of case study #2

2

3

1


Clubs, Elders society, Samurdhi Society were registered as social service

organisations. They do work with their own funding. There are district

registered organisations (NGOs). They get some funding and they also do

housing, roads, drinking water facilities, and toilet facilities” (Interview #12).

(2) Level of support for awareness programs: This theme emphasised the

importance of level of support given by the public to the awareness programs to

indicate social trust. The level of support provided by the local organisations to

implement awareness programs and the degree of public response for such awareness

programs or disaster drills show the level of trust by the public on the government

system and officers.


number of awareness programs conducted in the past years; b) number of people

participated in the awareness programs; and c) number of CBOs and volunteers

involved in organising the awareness programs.

“The level of trust depends on the work that the public institutions do in the

area. DS office conducts some programs such as awareness programs. We can

see how many people participate and attend those awareness programs related

to disaster management. We can measure the level of trust with the community

based organisations or societies, by measuring the contributions of people to

programs. For example, some organisations get more people when they ask for

awareness programs to come. Similarly, people will not come when some

organisations invite people to attend awareness programs” (Interview #40).

“When there is a disaster, there are some organisations who do some dedicated

work. For example, we have done a tsunami evacuation drill, a youth

parliamentarian group has done a good job. They have done a good effort on

evacuation drill, awareness, and training. Similarly, there is one society came

recently, when disasters happen, we have given a first aid training to the youth”

(Interview #16).

(3) Government office (Divisional Administration Office) support for people:

The support from government offices can measure the level of trust built with people,

as highlighted in this case study. The trust built during normal days will influence the

trust during the disaster management activities.



changes in number of requests/complaints to DS Office related to disaster relief or

recovery programs; b) number of people using Right to Information (RTI) services

for the past years; and c) appreciations or demonstrations/public protests/campaigns

held against the state institutions or projects in the past years.

“When disaster happens, people will approach DS office. Divisional Secretariate

(DS) will divert them to sectoral departments. If a road damaged, we will contact

Road Development Authority. If there is a problem in the lagoon, we will contact

the Municipal Council and they will do the work. They have given a call that

there is a water level increasing in lagoon and 15 houses are already under

water. I called Municipal Council and there was an excavator. The GN was in

the field with the community. When this type of work happens, people will trust

on the government system” (Interview #20).


In case study #3, nine themes were generated and four combined themes were

identified as higher-order themes (Figure 5.9). These include:

1. Effectiveness of community organisations;

2. Government office support for people;

3. Effectiveness of disaster relief system;

4. Complaint mechanisms available.

(1) Effectiveness of community organisations: This theme highlights the similar

theme (1) in case studies #1 and #2.


existence of coordination body and procedures to work in disaster management work;

b) how many petitions were made by people against the organisations; and c) number

of volunteers/human resources attached to the area organisations.


“When it comes to social trust, community based organisation that have a good

preparedness and services are very less in the area. There is less organisations

that have trained human resources and good preparedness plans that can

perform well during disasters. After disasters, organisations that have grown

well from other areas and outside Kalmunai come and do the work. The existing

organisations can do some work at the minimum level, but for disaster work there

is very less if none that can do disaster work effectively. There are some

organisations like sports club. There is no trained and effective organisation to

work in disaster time” (Interview #32).

(2) Government office support for people: This theme is related to trust built by state

officials with the public in normal time that have an impact on the trust in disaster

context. Although there are limitations from state institutions to immediately act after

a disaster due to bureaucratic regulations, the state mechanism for disaster

management and coordination before and after disasters are established in each village

and division. The divisional disaster management committee presided by the divisional

secretary is the central focal point for coordinating all the disaster management efforts.


response rate for the request to participate in public meetings/awareness programs; b)

Figure 5.9. Leximancer concept/theme maps from interview data of case

study #3

2

4

3

1


percentage of budget of the total spent/required by the government for disaster work;

and c) trend in complaints/legal cases/corruption petitions against government

officers.

“People believe that the government office will support immediately or even after

few days. For example we can see how the service by DS office on Wednesday,

how many people come to get the service in the DS office, will tell us the trust

people have on the government entities. People are aware more than us. Due to

media influence, people have more awareness and trust. People get the services

in the respective sector from the respective government counterpart” (Interview

#42).

(3) Effectiveness of disaster relief system: In this study, the concepts generated from

the interview data have focused on the trust built by the beneficiaries who received

disaster relief assistance. The level of satisfaction by beneficiaries by way of low

number of complaints lodged with respect to disaster relief system could be one of the

measures of trust.


duplication and gap of relief items provided to beneficiaries; b) number of

beneficiaries complaining of not receiving relief assistance; and c) trends of people

complaining lack of public services and resources in the area.

“The provisions of relief items will be same and the visibility will be different - I

mean their log are different. There is no coordination. Similar to resource

mobilisation and social cohesion, we can see how people and teams work

together in a normal situation if people work together, then during emergency,

they will work together. The work in the normal situation will reflect during

emergency. Since organisations need credit, how they work in normal situation

is important” (Interview #29)

(4) Complaint mechanisms available: The trend in complaints against an institution

could be a good measure of trust level with that institution. For example, how people

trust government institutions can be checked by the complaint trends against

government institutions.


trend of complains received by police or local authorities; b) public protests or publicly


expressing agitation against a state entity or CBOs; and c) the use of Right to

Information/Feedback box to complain.

“There are government officers who guide these societies, for example Rural

Development Society – Rural Development officer (RDO), for youth club – Youth

Service Officer (YSO), for Sports Club – Sports Officer (SO) are there. They can

provide a feedback or observations or complain about their respective

organisations within their purview. We can get some opinion from public about

community based organisations. We can ask opinions from community leaders

and other key community members about these organisations. We can get a good

status report, if we can ask from different segments of people whose roles fall

within those respective categories. There are suggestion boxes in the DS office.

We can check suggestion box, what proposals or complaints people have made.

We can also check with audit department. Audit department gets complaints from

many sections of people” (Interview #32)

5.2.4 Higher-order themes for case study #4 –Sub-national level

In case study #4, 11 themes were generated and four combined higher-order themes

were identified (Figure 5.10). They include:

1. Trust in information and early warning dissemination;

2. Functioning of disaster relief/management system;

3. Level of services and resources;

4. Level of trust in CBOs

(1) Trust on information and early warning dissemination: One of the important

aspects of trust building is through accurate information sharing such as during early

warning dissemination. People expect local authorities and respective state institutions

responsible for disaster management to issue an effective and timely early warning

message to the public, so that the public can take the necessary steps to prepare and

evacuate to safer places in due time.


Three key concepts identified in this theme are (as shown in box 1 in Figure 5.10):

Key concepts in this theme are: a) percentage difference in people evacuated between

authorised early warning issued by authorities against false early warning (rumours);

b) percentage of people who attended annual early warning rehearsals/awareness

programs; and c) number of phone calls made to DMC hotline/emergency number for

early warning confirmation.

“The trust building should start with proper warning. Met department has a

problem with the trust level on the warning and prediction. Early warning

government structure should be established with last mile reach. There are

excuses. I expect that the information should come to me. Timely warning is the

first step of trust. The first step of trust is whether people get the warning”.

(Interview #13)

“There was a Tsunami rumour two weeks back, 40 to 50 people called me. I told

them that there is no Tsunami. People believed in the rumour. The dissemination

of the early warning is not enough to the speed that the rumour spreads.

Figure 5.10. Leximancer concept/theme maps from

interview data of case study #4

2

1

4 3


………But normally the Police with Red Cross volunteers provide early warning

dissemination. There was no evidence for a Tsunami. The whole district was

affected by the rumour” (Interview #14)

(2) Functioning of disaster relief/management system: There are disaster

management committees formed in most of the villages under the Divisional Secretary

with the help of the Disaster Management Centre. However, the existence of such a

committee and a system does not alone build trust among the public. Instead they will

have to function in disaster preparedness phase and continue to prepare for disasters.

The level of trust in the committee and the system can be measured by its effectiveness

and regular functioning mechanisms.


annual trend in implementing Disaster Management (DM) related projects and people

support/participation; b) frequency of DM committee meetings per year and actions

implemented; and c) number of people trained in DM related programs such as first

aid.

“After 2004 Tsunami, the government and people started to talk about disaster

management. Government also started to work in disaster management. But

before that there were lots of rumours. There is a disaster management centre

hotline number - 117. We can see number of calls from public to see how people

trust disaster management activities. Like 117, at the national level we have 119.

People also see the media news….we have to say that the social trust has

increased. Sri Lanka government also has done many work at the field level. After

people face a disaster, the trust level increases. Now, the disaster management

committee and awareness programs are well established in the field level….we

don’t receive complaints directly. There are many CBOs and NGOs, but they are

not working in disaster management. Not many work is done in disaster

management currently” (Interview #26)

(3) Level of services and resources: People believe in a system based on the

satisfactory level of service and availability of resources within the system. For

example, for the public to trust the local authorities, they need to provide services to

the level of public satisfaction, and in return people will pay their taxes to the local

authorities. When there is lack of trust in the government system, people will start to

complain about the services and resources.



trend in the percentage of people who pay local authority tax annually; b) services such

as compensations for disaster losses promised and provided by the government; and c)

trend in problem reporting/lodging complaints to Police/DS office/Mediation board.

“The government offices are helping the people. But for local authorities, people

have to pay tax to services. One measure would be, number of people who pay

tax. If they trust they pay tax. If they ask people about common services that they

offer, for example solid waste, building approval, those things will show trust. In

the meantime the level of corruption is also high” (Interview #14)

“When there is some distribution or resource allocation, to implement a project,

there is no collaboration and coordination because it is very difficult to do. For

example to DS office and Municipal Council – how many service related petitions

have been lodged? We have formed a disaster management committee and we

need to coordinate” (Interview #11)

(4) Level of trust in CBOs: Since Community Based Organisations (CBOs) play an

important role in disaster management work and their close relationship with the

community, the level of trust determines the effectiveness in disaster preparedness and

response activities in a particular community.


functions of the Disaster Management Committee at the village level; b) trend in

volunteerism (capacity of volunteers and their projects); and c) ratio of participation

of members/CBOs in a regular public meetings in a division.

“Compared to the government institutions and local authorities, there is higher

level of trust with CBOs. When disaster happens, CBOs immediately do help and

relief work. People accept this and until people recover from disasters, these

CBOs continue to take care of peoples’ needs and get the complaints of people.

There is a maintenance body and leadership. There is a coordination and people

can lodge there complaints to this committee. Due to non-service of the

government, a new committee is formed. Those who need to do the work did not

do it, so the need for a new committee was realised” (Interview #11)


5.2.5 Final surrogates identified through cross-case synthesis for indicator #2

Table 5.2 below summarises the mapping of higher-order themes (taken as

surrogates) which emerged from the thematic analysis of interview data from the four

case studies.

Table 5.2. Surrogates (higher-order themes) mapping for indicator#2

Case study

locations/area

Surrogates to measure social trust in disaster

management

S21 S22 S23 S24 S25 S26

Case study #1

(KMT) X X X X

Case study #2 (SM) X X X

Case study #3

(KMM) X X X

Case study #4 (sub-

national level) X X X X

S21: Effectiveness of Community Based Organisations (CBOs)/social service

S22: Level of services and resources of local authorities/state support for people

S23: Functioning and effectiveness of disaster management systems and complaint

mechanisms

S24: Trust on information sharing /early warning dissemination

S25: Level of public support for awareness and public programs

S26: Public-government officer relationship


with the target indicator, and assessment protocols from the synthesis of interview

responses are summarised below.

1. Effectiveness of Community Based Organisations/social service (S21): This

surrogate was highlighted in all four case studies. Community Based Organisations

(CBOs) were highlighted as the key stakeholders in disaster management activities by

many interview participants across many case studies. Immediately after a disaster,

people have lack of trust on government mechanisms for help at least for the first three

days (72 hours). This fact was acknowledged even by the government officials,

because of rigid state procedures and bureaucracies do not allow the village level state

departments and officers to make decisions. They have to depend on the top for

decisions that will get delayed and people start to criticise the government although


they know this fact. However, people have more trust in CBOs that fill the gap in

disaster relief and recovery assistance during the time that government is unable to act

and mobilise resources.

Further, the measurement of this surrogate (S21) should include the assessment

and mapping of active and non-active CBOs in the selected region and their past

experiences in disaster management work. The assessment should also include their

present capacity such as the number of memberships/volunteers/networks and

resources in disaster management activities including the provision for disaster

management work in their constitutions.

2. Level of services and resources of local authorities/state support for people (S22):

This surrogate was also highlighted in all four case studies. An effective system of

providing service to the public by the government institutions and local authorities is

a key component of building trust with the public. The level of public service depends

on the commitment of the government staff towards providing the required services in

a timely manner and have sufficient resources needed to deliver the public services.

The surrogate (S22) can be measured by mapping and assessing the measure of

the projects implemented by state/local authorities against the total identified number

of projects, measure of the percentage of budget spent on projects against the budget

needed/ requested, and measure of the percentage of beneficiaries supported in disaster

management projects of the total identified people who needed help during disasters.

3. Functioning and effectiveness of disaster management systems/complaint

mechanisms (S23): This surrogate was highlighted in three of the case studies, except

case study#2. The social trust is built on the effectiveness of service delivery during

disasters and people weigh the support during the difficult times that they needed help.

The functioning of a disaster relief system with all primary stakeholders within a

community is key to building trust. The more people have trust, the less complaints

and vice versa. Hence, measuring the effectiveness of the disaster relief system by

understanding the existing disaster management committee structure and commitment

can be a surrogate measure of social trust.

The surrogate (S23) measurement protocol can include: measure of active

functioning of village/community level disaster management committee (such as

frequency of meetings, participation rate of committee members, and existence of a


disaster management plan) and measure the existence of complaint mechanisms

through the availability of a communication mechanism by the disaster management

committee, the availability of a complaint portal, and the trend in the number of

complaints received/demonstrations/other forms of resistance.

4. Trust on information sharing /early warning dissemination (S24): This surrogate

was only highlighted in case study #4. Information such as early warning messages

and evacuation procedures are very important for successful disaster management

work. The government entities such as Disaster Management Centre (DMC) is

responsible for issuing an early warning message to the public through an authorised

channel such as Police and DS offices. When an early warning message is issued, the

way people react depends on the trust with the early warning issuing mechanism or

dissemination authority. For example, if people evacuate from their places of living

when the disaster management centre issues a warning, it is a good indicator of trust

in the Disaster Management Centre. Similarly, the trust can be measured with other

social entities such as religious places, which also disseminate early warnings.

The surrogate (S24) can be measured by the percentage of people evacuated from

the vulnerable areas, measure the number of people self-evacuated/reached

displacement centre, percentage of people who participated in an early warning

rehearsal program or awareness programs conducted annually, and number of phone

calls made by the public to disaster hotline numbers such as 117 for early warning

clarification.

5. Level of public support for awareness and public programs (S25): The level of

public support for awareness and public programs was selected as the fifth surrogate

from case study#2. Social trust can be measured by the level of public support given,

when the disaster related awareness programs or other public programs are organised

in a community. The presence of number of community members invited is an

indication of the level of trust on the benefits of the program. The changes in number

of attendance annually can also be measured to see how the level of trust changes over

time. In addition to public participation, the measure can also include the involvement

of volunteers and community based organisations/social service organisation members

in organising such awareness programs/events to indicate the level of trust among the

public.


The surrogate (S25) can be measured by the number of awareness programs

conducted in the past five years, the trend in the number of people participating in

awareness programs, and the number of CBOs and volunteers involved in organising

such awareness programs.

6. Public-government officer relationship (S26): The public-government officer

relationship was highlighted as a key facet of social trust in case study#1. The way the

state officers responsible for the public service carries out their daily routine is critical

to build trust with the public. Their duty is done through the departmental work and

the trust is built on the effectiveness of their work. The relationship and the trust built

during normal working time will reflect in the trust during disaster work. Hence,

public-government officer relationship can be one of the surrogates to measure social

trust in a disaster context.

The surrogate (S26) can be measured by identifying the meetings or campaigns

convened by each of the government officers, calculating the percentage of attendance

of the public in such campaigns and meetings, the response rate/trend in support for

relief or fund collection campaigns by people in the area, and the trend in

complaints/legal cases reported against the government officers.



The study initially identified 91 concepts and 37 themes from four case studies

for indicator #2. Similar themes were then aggregated and six higher-order themes

were identified through cross-case synthesis which can be considered as potential

surrogates to measure social trust. The summary of the synthesis is shown in Figure

5.11.

The following three surrogates – S21: Effectiveness of Community Based

Organisations’ activities/social service, S22: Level of services and resources of local

authorities/State support for people, and S23: Functioning and effectiveness of disaster

management systems and complaint mechanisms, have higher validity, since these

were identified in at least three of the four case studies.

Another set of relevant higher-order themes (S24: Trust on information sharing

/early warning dissemination; S25: Level of public support for awareness and public

programs; S26: Public-government officer relationship) were found in at least one of

the case studies, can also be potential surrogates to measure social trust in a disaster

context. Hence, six potential surrogates to measure social trust are highly reliable and

practically applicable in similar contexts, since these were identified in consultation

with practitioners and policy makers who are highly experienced in implementing

disaster management activities at the community level.

Figure 5.11. Summary of synthesis for second social resilience indicator –

social trust


SURROGATE MEASURES TO ASSESS LEARNINGS FROM THE

PAST DISASTERS AS SOCIAL COMPETENCE (INDICATOR #3)


surrogates for assessing the third social resilience indicator selected in this study –

learnings from the past disasters as social competence. Social competence gained

through past disaster experiences is about how communities utilise their positive and

negative learnings to improve preparedness and recovery activities for future disaster

events (Hoffmann & Muttarak, 2017).




Leximancer concept maps. These were:

1. Evacuation drills

2. Reaction to early warning

3. Disaster awareness/knowledge level

4. New construction methods (e.g. Houses)


of case study #1

1

2

3

4


(1) Evacuation drills: An increase in participation in annual evacuation drills by the

population most vulnerable to disasters may indicate lessons learnt from the past

disasters. Disaster evacuation drills and exercises are carried out regularly – for

example in the case of Tsunami which is done annually.


participation of people (percentage increase after the disaster) in disaster drills; b)

number of people who carried disaster grab bag during the disaster drill; and c) the

level of support by CBOs and other stakeholders to conduct drills.’

“Disaster drills are conducted by the Disaster Management Centre. The

participation of people in these disaster drills has increased over time. People

show interest in learning new knowledge and experiences from the past about

Tsunami and other disasters. Disaster drills are conducted for the people

continuously. It was made clear to people during disaster drills and awareness

programs that how they should evacuate, what are the procedures for

evacuation, and what kind of relief support they will receive after a disaster”.

(Interview #41)

“In the past people ran and evacuated with a bag that contains important

documents and money. We have done a mock drill. People know now and they

were made more awareness. We have done the drills in all the coastal areas”

(Interview #28)

(2) Reaction to early warning: The changes in reaction to early warning before and

after a disaster could be a measure of learnings from the past disasters. Before a disaster

experience, people may not have any knowledge about the disasters, for example, the

tsunami in 2004 in Sri Lanka. After the 2004 tsunami, people react to early warnings

issued by the local authorities differently. The difference in reaction is due to lessons

from the past experience.


changes in issuing proper early warning messages to the population vulnerable to

disaster risks; b) changes in evacuation of people for evacuation early warning

messages; and c) participation of people in awareness programs held to provide new

knowledge about early warnings.


“People have good awareness now about early warning. When an early

warning message is issued, people become alert. We have done many

awareness programs and tsunami mock drills. We have two Tsunami early

warning towers in this region. People will not accept the early warning

messages from other people. They will only accept early warning messages

from religious places such as temples and mosques and from the DS (Divisional

Government Administration) office. However, sometimes, even when there was

a tsunami rumour, people moved immediately away from the disaster prone

coastal areas” (Interview #28)

(3) Disaster awareness/knowledge level: People learn from every disaster, a lesson

that helps them to prepare for future disasters. The efforts taken after a disaster to

raise level of awareness can indicate the learnings from the past disaster experience.


changes in participation in disaster awareness events by the public; b) involvement of

community based organisations in organising awareness events/interest shown; and

c) changes in disaster risk reduction activities before and after disaster.

“We have done awareness programs about disaster management. When we ask

people what will you do when disasters happen, people say that they are

prepared with the disaster grab bag. People tell us even before we do

awareness programs. They have clinic card, people will take medical cards and

other important documents and they have kept them in disaster bags they

already prepared. They also know which way to evacuate and what the routes

they should do evacuation are” (Interview #28)

(4) New construction methods (e.g. Houses): An example was highlighted for not

taking actions from the past flooding experience in this case study area. Problems

were identified with drainage and it was the primary cause of flooding. However, no

learnings from the past seem to be accounted to prevent flooding by solving the

drainage problems. This shows the lack of willingness to take the past experience into

account to address the emerging disaster related challenges.


changes in new housing constructions (incorporating DRR measures); b) public


infrastructure DRR integration projects such as drainage systems; and c) no

development in the identified risk areas.

“When concrete roads are constructed or people construct houses, they have

to lay pipes from their houses to the drainage canals. During the time of

construction, they should have kept a pipeline to send the water out of their

house to the drainage. People wait until the flood water gets into their houses

and later they will damage the roads to send the water out. This should have

been done before. If people have an awareness and proper knowledge on how

to divert and send the water out of their houses when they construct houses,

they would have placed water carrying pipes properly” (Interview #4)


In case study #2, eight distinct themes were generated. By combining closely related

themes, three higher-order themes were identified (Figure 5.13). They include:

1. Degree of government disaster preparedness work;

2. Disaster awareness/knowledge level;

3. Reaction to disaster early warning.

(1) Degree of government disaster preparedness work: Many government

institutions involve in disaster management work. New initiatives and projects are

initiated by government institutions before and after disasters. The changes in the

initiatives can provide a measure of learnings from the past disasters.


the changes in disaster preparedness projects before and after disasters; b) new

modifications in the government disaster preparedness plans after disasters; and c)

changes in government regulations/circulars/policies to mitigate and prevent disasters

damages.

“A twitter communication system is developed. In this twitter group government

officers are there. So they are included. People did not take risk reduction

measures to prepare for disasters in constructing their buildings and to reduce

the impact the losses. Government institutions have built new buildings taking

into account the disaster risks. People have no knowledge about the disaster


resilient housing construction. Even government institutions also lack this

knowledge. Even the masons who construct buildings are also not

knowledgeable” (Interview #16)

“Sometimes people do not take steps on their own. People have the mentality

that they expect the government institutions to do everything. People do not do

projects themselves very much, but depend on government or other projects

externally initiated. At the government level, have a very good preparedness.

We have a good preparedness plan. So, people are prepared because

government is well prepared to face future disasters” (Interview #20)

(2) Disaster awareness/knowledge level: People learn from every disaster a lesson

that helps them to prepare for future disasters and the efforts taken after a disaster to

raise the level of awareness can indicate the learnings from past disasters.


self-initiations by people at the individual/household level disaster preparedness


case study #2

1

3

2


activities; b) importance given to awareness programs such as new initiatives on

DRR; and c) increase in number of people participate; and d) participation in disaster

drills with new learnings such as preparedness kits.

“People who were affected by disasters have some experience, so we learned

lessons. We do awareness programs for children who are born after major

disasters. We have also done awareness programs for school teachers. Disaster

Management Centre in districts have video clips and practice about past

disasters. We have school awareness programs, combining public and school

children. We have also done awareness programs and discussions regularly with

the school principals. The awareness programs include the knowledge about

hazards and what precautions can be done. We have instructed people using

disaster examples happened recently” (Interview #12)

(3) Reaction to disaster early warning: Reaction to disaster early warning can help

to understand the learnings from the past disaster experiences. People learning from

the past disaster early warnings and improving their response to upcoming early

warning messages, is a key indication of learnings from the past disaster experience.


changes in early warning systems such as early warning tower, new disaster drills,

and early warning testing programs; b) differences in reaction to early warning

messages before and after disasters; and c) participation trends in early

warning/evacuation drills.

“The last 2004 Tsunami was not known to anyone. People went to see the sea

instead of going away. People only know that sea is coming to land. Now Disaster

Management Centre has done awareness programs for the last 13 years, now

people know what Tsunami is. People now have the knowledge, how to evacuate

when an early warning is issued. Earlier, it was told that when there is an

earthquake in Sumatra, it will take two and half hours to come to Sri Lanka. Now

we have made people aware” (Interview #40)


In case study #4, nine themes for case study #3 were generated and four combined

themes were identified as higher-order themes (Figure 5.14). This include:

1. Initiatives of disaster management committees or members;


2. New DRR programs (e.g. changes in buildings/infrastructure designs);

3. Changes in awareness knowledge;

4. Reaction to early warning.

(1) Initiatives of disaster management committees or members: Disasters often

highlight the need for active disaster management committees and members. Disaster

management committees can help to organise not only response to disasters

effectively, but can also play a key role in Disaster Risk Reduction (DRR). The

learnings from past disasters may trigger an inactive committee to work actively after

disasters based on past learnings. When there is no such committee dedicated to focus

on DRR work, there is an opportunity to establish such a committee after a disaster.


new committees established after recent disasters; b) new projects implemented after

recent disasters; and c) changes in committee work plans (improvements).

“People were already instructed about the evacuation locations and routes. We

also created a committee and they convene monthly meetings and discuss


of case study #3

4

3

1

2


disaster preparedness work. The committees function when there is a disaster”

(Interview #29)

“Since people are educated, they have a good knowledge. We have disaster

preparedness committee in each GN and with WRDS and RDS and other Social

Service Organisations, we do many awareness programs. These programs were

done with the disaster management centre to raise awareness and during the

committee meetings, we do awareness raising programs once in six months”

(Interview #19)

(2) New DRR programs (e.g. changes in buildings/infrastructure designs): The

past disaster experience helps the community to take proactive initiatives to reduce

risk from future disasters. For example, when water enters into the houses during flood,

people raise the foundation of the newly built houses. After 2004 Tsunami, people

changed their house designs and build upstairs houses. Similar type of disaster risk

reduction programs can be an indicator that people learned lessons from past disasters

and apply those lessons to reduce damages from future disasters by increasing their

resilience.


changes in housing/infrastructure designs in the disaster prone areas; b) records of new

DRR projects implemented after recent disasters; and c) new regulations/law/circulars

to reduce the impact of disasters.

“When we discuss about Tsunami, the land use in the coastal area has changed.

Now the coastal area was declared as a buffer zone. There were many houses

and buildings and now it is not the residential area. This new regulation has

imposed restrictions on building new constructions. There will be no permission

given in these area for buildings. Schools and other public buildings were not

given permission to operate and the law is enforced to restrict constructions in

this area. The area is also not allowed for construction, where people gather at

large numbers. People are now aware and on alert. New designs were also

introduced. The foundation level has increased in the flood prone areas. The

design of building constructions has changed to reduce future disaster losses”

(Interview #42)


(3) Changes in awareness knowledge: The changes in awareness campaign can be

explored to measure the improvements and level of learnings incorporated in new

disaster awareness program strategies. People may attend more enthusiastically in

awareness and training programs related to disasters based on past experience.


increased level of awareness programs and participation; b) increased volunteer

contribution and new stakeholders in organising awareness programs and disaster

related trainings; and c) different types of awareness programs (innovative).

“Many training programs were conducted and the knowledge has transferred to

different levels in the community. The training programs were done for many

segment of people like children, female, people with specific needs and other

different segment of population. There are many media work on DRR. We can

check them – if they produced CD or video, handbills for raising awareness”

(Interview #25)

(4) Reaction to early warning: Early warning plays a key role in disaster

management, particularly for disasters such as Tsunami. When an early warning alert

is issued to the population at-risk, the reaction of people may vary based on the past

learnings from disasters. In the village level disaster management structure, there is a

village level disaster management committee and there are sub-committees.


pre and post activities of early warning sub-committee under the village DM

committee; b) peoples’ reaction to disaster drills such as preparedness level in safe

guarding important documents; and c) number of people evacuated with and without

official early warning messages from local authorities.

“We can see their disaster preparedness activity. When there was an evacuation

during the tsunami rumour recently, people evacuated with important

documents. We could see them with these important documents when they

evacuate from their villages. This is an evidence from their past lessons. People

immediately went to schools and get their children, when there was an early

warning issued. People move children immediately from their houses as a

priority” (Interview #29)



In case study #4, seven themes were generated and four combined higher-order themes

were identified (Figure 5.15). These include:

1. Government action and community reaction to early warning;

2. Disaster evacuation/preparedness kit;

3. Disaster risk reduction work (New building structures);

4. Village level disaster management committee.

(1) Government action and community reaction to early warning: One of the key

components of disaster resilience strategy is to enhance effective early warning

generation, transmission, and dissemination to the last mile. The responsible authority

to issue early warning is mostly assigned to a state disaster management authority. The

past disaster experience highly influences not only on the government side, but also

among the community in reacting to the early warning.


changes in past to present investment in disaster early warning systems and awareness

Figure 5.15. Leximancer concept/theme maps from interview

data of case study #4

3

2

1

4

1


programs by the state; b) disaster drills and knowledge about evacuation locations; and

c) plan with early warning dissemination methods.

“When we issued a tsunami early warning even in the mid night time, people

evacuated timely. They have good preparedness. People who are geographically

and economically vulnerable, cooperate well for disaster drills. Sometimes we

expect 100 people, but around 300 people turnout during the drills. In some

areas, cooperation is very less and we think it is due to their lack of interest,

since their area was not affected much during the past tsunami. In the areas

which were affected badly, people give good cooperation” (Interview #26)

(2) Disaster evacuation/preparedness kit: Every disaster teaches different lessons to

people who are affected by those disasters and to the authorities who manage those

disasters. One of the learnings from the past to improve future resilience is to be better

prepare at the household level, to help people to recover back from the impact of

disasters quickly. Most of the interview participants were of the view that they noticed

during disaster evacuation drill, an increasing trend in people bringing a disaster

preparedness kit that is kept ready in their houses with important documents covered

in safe pockets.


number of households have prepared emergency kits; b) households have emergency

contact list and numbers; and c) sale of emergency preparedness kits/items by the local

shops.

“After fire incidents, how many people bought fire extinguishers? Items like first

aid kits and fire extinguishers can be checked in places that are sold. We can see

how many have been sold in the area. Laminating facility – how many people

have done laminated documents. Similarly we can check with the banks how

many people use safe lockers, after a disaster event” (Interview #10)

(3) Disaster risk reduction work (New building structures): Disaster Risk

Reduction (DRR) is the key driving theme in resilience literature. After a disaster,

DRR becomes a central talk in policy discussions and there is a tendency to forget

DRR after sometime. Therefore DRR should become a mainstreaming theme in all

development projects and budget allocations. Based on the past learnings from disaster


damages, new DRR initiatives need to be taken on board to reduce future disaster

losses.


availability of a disaster risk map in each division; b) trend in financial investment (or

separate budget allocation in DRR projects/number of projects addresses DRR; and c)

changes in infrastructure development policy/practice.

“We can check the financial investment by the community in their village

themselves, if there is an increase, then there is an interest shown and people

take into account the past disaster experience. Even in divisional meeting, a

community based organisation talks about DRR work, then it shows that they

have learnt lessons from the past to prepare for future disasters. Changes in the

number of DRR projects can be a good measure. We can financially measure the

level of DRR investment. If there are three agro wells constructed after a disaster

by the irrigation department, then there is lessons learnt. We can also check

number of letters comes from the community to DS office asking for DRR

projects” (Interview #5)

(4) Village level disaster management committee: It is essential that the committee

manages all activities related to disasters provides a coordinated and effective

mechanism to overcome challenges. When a disaster management committee at the

village level exists before a disaster, the changes in its function can be identified to see

how the past learnings from the previous disaster were taken on board. When there is

no such committee at the village or divisional level before the disaster occurs, the

indicator could be the formation of a new committee since the village could have felt

the need for such a committee to manage future disasters.


the availability of disaster stakeholder list and emergency contact list; b) the

availability of core group of people trained on critical disaster management skills such

as first aid; and c) function of the existing committee and projects implemented.

“Civil security committees are formed very recently. This was formed in every

village after the war to deal with security issues and to stabilise the villages

towards security issues. But now these groups have broadened their focus beyond

security issues and included disaster issues. There is a system and organisations


like us have given capacity building programs. These committees can deal with

all problems in a village, but they need to be given more capacity building. They

have to be fully equipped. After tsunami and floods, Disaster Management

Committees or units were started” (Interview #1)




case studies.


Case study

locations/area

Surrogates to measure learnings from the past

disaster experience (Emerging themes)

S31 S32 S33 S34 S35 S36

Case study #1

(KMT) X X X X


Case study #3

(KMM) X X X X

Case study #4 (sub-


S31: Reaction to disaster early warning

S32: Awareness and disaster knowledge level

S33: New DRR programs such as innovative construction methods

S34: Functional level of the disaster committee/members

S35: Disaster Risk Reduction (DRR) & Disaster Preparedness (DP) strategy

S36: Evacuation drills & disaster preparedness kits

For each of the potential surrogates, a brief description of the surrogate, its

relationship with the target indicator, and assessment protocols from the synthesis of

interview responses are summarised below.

1. Reaction to disaster early warning (S31): This surrogate was highlighted in all four

case studies. The key objective of any disaster preparedness activity is to save lives.

Early warning plays an important role in providing required information to people for

their necessary actions in a timely manner. An appropriate reaction to early warning


messages from the public enhances their resilience to disasters. The effectiveness of

reaction to early warning increases based on the learnings from the past disaster

experience. All four case studies have highlighted the reaction to disaster early

warning as a surrogate measure to learnings from the past, since there have been often

many improvements on early warning systems and peoples’ reaction to early warning

based on the past disaster learnings.

The surrogate (S31) can be measured by identifying the early warning systems

built after a disaster, the past real early warning instances, early warning/evacuation

drills, the reaction of the public by number of people who reacted appropriately, trend

in number of confirmation calls to disaster management centre after early warning

messages, and the number of people evacuated with and without official early warning

messages from local authorities.

2. Awareness and disaster knowledge level (S32): This surrogate was also highlighted

among three case studies #1, #2, and #3. The knowledge about disaster risks and

awareness of disaster risk reduction and preparedness methods are important factors

influencing disaster resilience of a community. When the awareness level of disasters

is higher, the community is well prepared to manage future disasters. In Sri Lanka,

2004 tsunami was one of the examples that showed that Sri Lanka as a country has

very minimal awareness and knowledge about tsunami. This disaster was a turning

point for major changes and a national policy shift in disaster management. Hence, the

key factor of influence for the lessons from past disaster experience could be attributed

to awareness and disaster knowledge of past disasters and future disaster risks.

The surrogate (S32) can be measured by awareness/training/capacity building

programs conducted over the past years, public participation trend in

awareness/training/capacity building programs, and the support by community based

organisation/other stakeholder engagement in organising the awareness programs.

3. New DRR programs such as innovative construction methods (S33): New DRR

programs such as innovative construction methods was selected as the fifth surrogate

from case studies #1, #3, and #4. The learnings from past disasters help people to

improve disaster resilience strategies. There could be many initiatives to reduce the

foreseen disaster risk and prepare for them. These measures can be structural and non-

structural. While people apply some DRR actions at the household level, for example,

people raise the foundation of their houses above the past flood level, or build a room


in the upper floor or reinforce the roof with stronger structures, when they construct

new houses and infrastructure. There could be many new innovative solutions

proposed for reducing the future disaster risks and preparing better for emerging

disasters. The learnings from the past disaster experience triggers people, responsible

community leaders and state officials to take necessary steps to avoid the previous

mistakes being repeated.

The surrogate (S33) can be measured by analysing the new projects implemented

after a recent disaster, the degree of integration of DRR activities mainstreamed in

those projects, percentage of new development proposals proposing DRR integration

measures, and documented learnings and past disaster experiences in the area selected

for resilience measure.

4. Functional level of the disaster committee/members (S34): This surrogate was

highlighted in two of the case studies #3 and #4. The existence of a disaster

management committee provides an importance assistance in enhancing community

competency and disaster resilience. The members in a disaster management committee

most often include key state officials and community leaders who can guide the

community towards effective disaster management activities. The actions taken by the

disaster management committee after a disaster based on learnings from the past

disaster experience could be a good measure of the competency of the community to

manage future disasters.

The surrogate (S34) measurement protocol can be undertaken by identifying if

there is a committee exists to manage disaster work, by measuring the records of the

activities implemented by the committee before and after a recent disaster, and the

measure of changes (improvements) in the committee itself after the occurrence of a

recent disaster based on the key lessons from its functions.

5. Disaster Risk Reduction & disaster preparedness strategy (S35): This surrogate

was highlighted only in case study #2. Disaster Risk Reduction and Disaster

Preparedness (DRR and DP) is the key for implementing resilience building activities

at the community level. DRR and DP strategy at the community and sub-national

levels need to be formulated based on the key learnings from the past disaster

experience. Central and local government play an important role in initiating and

implementing DRR and DP activities at the policy level and they have to ensure that

the policies are implemented that bring benefits to the communities at risk by reducing


the disaster risks. The changes in policy and projects of the government DRR/DP work

will indicate the incorporation of past learnings in the current or future project and

policy initiatives.

Further, the measurement of this surrogate (S35) should include the assessment

of the disaster risk map availability, measure of DRR and DP projects implemented,

measure of the budget spent on DRR and DP projects before and after a recent disaster,

and measure of new constructions that take into account DRR integration such as

elevating foundations/constructing high rise buildings in disaster prone areas.

6. Evacuation drills & disaster preparedness kits (S36): The participation in

evacuation drills and in possession of disaster preparedness kits was highlighted as one

potential surrogate to measure learnings from past disaster experiences in case studies

#1 and #4. The evacuation follows the early warning, when the disaster risk is

imminent. The evacuation drills as part of the disaster preparedness strategy help to

train people for effective evacuation in real time. The participation of people in

evacuation drill shows the interest of people in disaster preparedness activities. People

are trained to carry disaster preparedness kits during the evacuation that will give them

a practice to do it in a real disaster situation.

The surrogate (S36) can be measured by the number of evacuation drills

conducted before and after a disaster, number of people at risk who participated in the

drills before and after disaster, number of people who carried a disaster evacuation kit

among the people evacuated, the degree of participation of community based

organisations in disaster evacuation drills, and the availability of disaster

simulation/showcase in the area selected for resilience measure.



The study initially identified 84 concepts and 32 themes from four case studies.

Similar themes were then aggregated and six higher-order themes were identified

through cross-case synthesis which can be considered as potential surrogates to

measure social competence. The summary of the synthesis is shown in Figure 5.16.

The following three surrogates – S31: Reaction to disaster early warning, S32:

Awareness and disaster knowledge level, and S33: New DRR programs such as

innovative construction methods have higher validity, since they were identified in at

least three of the four case studies.

Another set of relevant higher-order themes (S34: Functional level of the disaster

committee/members, S35: Disaster Risk Reduction (DRR) & Disaster Preparedness

(DP) strategy, and S36: Evacuation drills & disaster preparedness kits) were found in

at least one of the case studies, can also be potential surrogates to measure learnings

from the past disaster experience. Hence, six potential surrogates to measure learnings

from the past disaster experience are highly reliable and practically applicable in

similar contexts, since they were identified in consultation with practitioners and

policy makers who are highly experienced in implementing disaster management

activities at community level.

Figure 5.16. Summary of synthesis for third social resilience indicator –

social competence


SURROGATE MEASURES TO ASSESS INVOLVEMENT OF PEOPLE

WITH SPECIFIC NEEDS AS SOCIAL EQUITY IN A DISASTER

CONTEXT (INDICATOR #4)


surrogates for assessing the fourth social resilience indicator selected in this study –

involvement of people with specific needs as social equity in a disaster context. The

involvement of people with specific needs who are most vulnerable and often

marginalised refers to the level of proactive engagement and provision of equitable

opportunities in increasing their resilience to disasters (Stough & Kang, 2015).



themes) were identified (Figure 5.17) based on connected concept nodes in


1. Disaster facilities with access priority to PwSN;

2. Available organisations and projects for PwSN;

3. Social safety programs/funds for PwSN;

4. Availability of committees and support services.

(1) Disaster facilities with access priority to PwSN: Most of the interview

participants highlighted that the key facility that require PwSN friendly access is the

evacuation centre. However, when there is lack of access to PwSN in normal situations,

interviewees raised concerns to raise interest among state and other responsible

stakeholders for investing in giving priority for PwSN to access facilities. Further, it

was highlighted that the key facilities such as the evacuation centres need to be

identified and assessed for measuring equity for PwSN.


PwSN access facilities to evacuation centres and other disaster related facilities such

as community centres; b) priority services available for people with specific needs; and

c) government policy initiatives to make the disaster facilities such as schools, access

friendly for PwSN.


“If there is a two story building, there is no provision for a disable person

to access that building. So there is no access and there is no thinking about

them in our society. There is no follow up monitoring and capacity building

efforts and support is not widely available” (Interview #4)

We also do some facilities such as for the person to access through wheel

chair in his house. If he uses toilet, then he or she can go to toilet and access

facilities without the support of others. We assess the needs of these people

and we do facilities such as hand rails to access toilets. We did not complete

it 100% but we do it satisfactorily” (Interview #28)

(2) Available organisations and projects for PwSN: In a community, there are many

Non-Government Organisations (NGOs) and other social service voluntary


of case study #1

3

4

1

2


organisations that implement community based projects. It was highlighted that there

are a number of projects implemented specific to PwSN in the target area.


availability and effective functioning of PwSN organisations; b) disaster related

projects implemented specific to PwSN by the government and non-government

organisations; and c) funding allocation from organisations for PwSN projects.

“There are some organisations but not in our area helping to PwSN. But, they

only help during disasters such as World Vision, Red Cross, EHED, and SERVO.

They do not function now and they are not here. It is only the government that

supports PwSN now” (Interview #35)

“When there is a disaster, we can check the list of relief projects or work. If they

are included properly then we can at least assure that they are given equal

opportunity. Samurdi is another mechanism for poverty alleviation program.

Government has reduced the number of Samurdi beneficiary list. Government

do not check their ability to earn enough income and awareness and education

to spend their earnings for the expenditure. This is available in circular and they

are implementing. But the problem is in monitoring” (Interview #4)

(3) Social safety programs/funds for PwSN: The availability of social safety

programs and funds/resources for people with specific needs is an important indicator

to measure the inclusion/equity for people with specific needs.


Government programs available for social safety funds targeting PwSN; b)

Government allocation for PwSN projects (including disaster related work); and c)

Non-governmental/charity/private sector initiatives or contributions for social safety

net programs for PwSN.

“There is a special project. There are many payments and schemes in the social

service department. There is a society in Kalmunai North and Kalmunai South.

There is a disable society and we give revolving fund without interest. There are

63 people with specific needs” (Interview #39)

(4) Availability of committees and support services: The involvement and equity

for PwSN can be enhanced when specific committees and support systems are

available to address the needs of PwSN in the state and community structure.



the level of representation of PwSN in disaster related committees; b) stand-alone

committees for PwSN or organisations formed by PwSN; and c) available support

services such as disability aids and building accessibility.

“Generally, the treatment for PwSN in normal routine life are subjected to

special consideration in all activities and projects. They are also given special

focus during disaster time. Data is available about PwSN such as elders,

disables, and the type of problems they have. When they are accommodated

in public places during evacuation, elders, disable, children, and women are

given support. There is an assessment of what type of assistance they require,

when they are accommodated in public places” (Interview #27)


In case study #2, nine distinct themes were generated from case study #2. By

combining closely related themes, four higher-order themes were identified (Figure

5.18). These include:

1. Social safety programs for people with specific needs;

2. Participation of persons with specific needs in training programs;

3. Availability of support services with access to people with disability;

4. Available committees and organisations for PwSN/representation of PwSN in

organisations or committees.

(1) Social safety programs for people with specific needs: Social safety programs

for elderly/persons with disability, plans and programs by the social service

department and availability of society or committee for persons with specific needs

are key considerations in the assistance to PwSN.


a) Government monthly/regular funding mechanisms for People with Specific Needs

(PwSN) under social safety net program; b) availability of disaster related specific


support to PwSN such as insurance scheme; and c) degree of external assistance to

PwSN.

“For elders beyond 70 years, every individual is given 2,000 rupees per

month. Through the elder’s society, social service department provide

spectacles for those who have eye problems. They are also given monthly

allowance as explained earlier. Their medical expenses and other specific

expenses are mostly covered by the government. For the people who are under

poverty level, they live in shelter. When their houses are damaged, they will

be assisted with food and other relief support” (Interview #40)

(2) Participation of persons with specific needs in training programs: The degree

of awareness and knowledge about disaster preparedness among the persons with

specific needs such as people with disability, elders, people with chronic illness,

children and pregnant mothers is an important measure of resilience. Since PwSN are


case study #2

1

4 3

2


highly vulnerable to disasters, they should be treated with special care and thus,

availability of training programs to build their capacity is important.


percentage of attendance by PwSN in disaster related awareness or training programs;

b) degree of participation of PwSN in disaster drills; and c) specific disaster related

trainings or awareness programs for PwSN.

“We provide them some technical training. We have information. They do not

have enough skills to compete with other people in the society. We expect some

qualifications for them to be appointed. There is a percentage of allocation of

recruitment in general recruitment procedure. There are many differently able

people who are very well skilled. They can be capacitated to develop their skills

to get jobs” (Interview #12)

“When they come to get these services in the DS office, many awareness

programs and information are passed to people over 70 years of age. When

there is known disaster before it happens, we can gather these people with

specific needs and educate them about impending disasters” (Interview #31)

(3) Availability of support services with access to people with disability: When

the accessibility of people with disability or elders are concerned, most often the first

thing the interview participants highlighted is the need for accessibility to buildings.

This may include ramps or lift services in buildings, so that people who come in wheel

chairs can easily access the services they wish to obtain.


availability of access facilities such as ramp, lift and hand rails in buildings; b)

provision of accessible aids for the people with specific needs; and c) support services

available for people with specific needs.

“There are grants and funding for constructing toilet facilities in their houses

and they are given supportive aids such as walking sticks and wheel chairs.

People with specific needs are also supported with grants to build houses.

There are projects under social service department such as 250,000 rupees

grant for PwSN” (Interview #40)

(4) Available committees and organisations for PwSN/representation of PwSN

in organisations or committees: The availability and effective functioning of a


committee specific to PwSN or a registered organisation run by PwSN such as

“Differently-able Persons Organisation” (DPO) can enhance the resilience of PwSN

to disasters.


the degree of priority given to PwSN in regular government programs; b) availability

of evacuation centres with support structure for PwSN; and c) Government policy

initiatives such as circulars or regulations for equity access for PwSN.

“Elders are living with children. There are elders’ society for each GN division.

In our division, there is a registered disable organisation. We can ask from these

societies whether they have been provided needed support during the past

disasters or to prepare for future disasters” (Interview #40)


In case study #3, seven themes were generated and four combined themes were

identified (Figure 5.19). These include:

1. Social safety programs for the most vulnerable people (Elders/Women Headed

Households);

2. Organisational and government support for PwSN;

3. Registered groups/committees for PwSN;

4. Planning and resources for PwSN.

(1) Social safety programs for the most vulnerable people (Elders/Women

Headed Households): Most of the social safety programs focus on two key segments

of people with specific needs such as elders and women headed households (widows).

Elders and widows can be physically highly vulnerable to disasters and require

immediate and special care. Government implements social safety net programs such

as monthly payment to elders over 70 years age and similar allowance for women

headed households who are under the poverty line.



representation of women in local organisations; b) care for elders (social safety

programs by state); and c) social safety financing for widows.

“Every month, elders above 60 years of old, get 2,000 rupees, people with

disability get 3,000 rupees by Social Service Department. In addition to

Public Monthly Allowance (PMA), social safety allowance (called

‘Samurdhi’) is also given to women headed households/widows and for the

families under poverty line. During disaster times also, the same funding is

given” (Interview #29)

(2) Organisational and government support for PwSN: The availability of

organisational (non-state) and government projects and support targeting PwSN can

ensure equity of resource allocation for PwSN, particularly for people with disability.

This includes programs or projects and resource support through specific interventions

related to disasters.


case study #3

4

1

2

3



Organisational services available to PwSN; b) support from government for PwSN;

and c) disaster specific services and support for PwSN.

“There are many access facilities for the people. When disaster response work

was done by some organisations, they collected information about people with

specific needs. They have designed and implemented specific programs for

these people” (Interview #32)

When there is a problem we temporary toilets for them with the help of MOH

and Municipal Council through respective organisations and NGOs and also

philanthropists. Government provides social service support. Government for

person 150/- per person item. Special allocations are done through NGOs.

Additionally not much done.” (Interview #30)

(3) Registered groups/committees for PwSN: The availability of registered groups

or committees that are set up for PwSN can enhance the support systems for PwSN in

a disaster context. However, the active functioning or participation of such groups or

committees in disaster risk reduction programs need to be considered, when measuring

the involvement of PwSN for enhancing disaster resilience.


available forums or committees for PwSN (specific to DRR); b) meetings organised

by the registered groups or committees in DRR; and c) projects/awareness programs

implemented.

“There is a group working with people with disabilities called ‘Human Link’.

This disability group is registered in government. They talk about disable people

rights and problems” (Interview #24)

“If a person with disability needs to attend a meeting, is the meeting place has

access to them. Are the people with specific needs organised in an organisation

or committee. Do they have a specific organisation formed for them? We can

see the structure? If people with specific needs to come to meeting,” (Interview

#25)

(4) Planning and resources for PwSN: Disaster Risk Reduction (DRR) plans in each

government administrative division need to consider the needs and vulnerabilities of

PwSN. The plan is an initial indicator, but also the resource allocations for effective


implementation of the plan can be considered as another measure to assess the equity

for PwSN in DRR programs.


availability of DRR plans targeting PwSN; b) resource allocation in general projects

for PwSN; and c) resource allocation and financing for PwSN for DRR projects.

“In the contingency plan, we have identified people who are more vulnerable to

disasters such as flood. There are around 1,500 people in such situations. Equal

access to resources is a problem. Because, we do not have enough resources.

Sometimes, it can be in Municipal Council. We can’t take decision alone. We

invite all stakeholders including community based organisations to a meeting.

When we do a project, we invite all important people for preparing the

contingency plan. When we do these meetings, we will take decisions with key

people in the community” (Interview #24)


In case study #4, ten themes were generated and four combined higher-order themes


1. Accessibility in evacuation buildings;

2. Early warning information for PwSN;

3. Priority criteria for people with specific needs in projects;

4. Government/NGO support services for PwSN.

(1) Accessibility in evacuation buildings: Evacuation is the key disaster response

component and timely evacuation of people can save many lives. However, evacuation

of people with specific needs is challenging which requires specific resources. For

example, people with disability requires additional physical and human resource

support to evacuate during disasters. Similarly, when they are evacuated to a pre-

identified evacuation building demarcated by the disaster management committee,

they should be with all accessible facilities for PwSN.


ratio of evacuation building with accessibility for PwSN and without accessibility; b)


regulations for construction of disabled/elders friendly buildings; and c) approval

mechanisms for construction of buildings with accessibility.

“The general circulars and rules will be same for local authority similar to

government systems. When new buildings are constructed, the access and lift

need to be ensured in the buildings. It is followed by the local authorities. We, at

the local authority will only approve the buildings that have the access for

persons with specific needs” (Interview #36)

(2) Early warning information for PwSN: Early warning is the key for effective

evacuation that will help people, particularly PwSN to prepare for timely evacuation.

Hence, the availability of multiple early warning messages that can cater to the needs

of different people such as people with disability (blind and deaf). A specialised early


case study #4

1

3

2

4


warning messages and ability to communicate with people with specific needs can

measure the involvement/equity of PwSN in disaster context.


availability of diverse early warning information dissemination methods for PwSN; b)

availability of awareness programs for PwSN of specific early warning systems; and

c) availability of access to information systems for PwSN.

“If people have multiple methods available to access early warning information,

it is an important measure. Certain Disable People Organisations (DPOs) are

well functioning. Active participation of them can be a good indicator. Early

warning should also include signs that can be understood. If there is a school

with special needs children, early warning messages need to target these students

through specific early warning signs that can be understood by them”. (Interview

#10)

(3) Priority criteria in projects for people with specific needs: The availability of a

set of criteria to prioritise the assistance for PwSN in the state and non-state funded

projects and services can be a measure of involvement and equity for PwSN in a

disaster context. The priority criteria is informed or decided through a circular issued

by the social services department at the national level to the sub-national offices, which

implements the projects for PwSN. There are also methods of assessing the needs of

the population based on a set of criteria and a scoring system. The specific questions

in the assessment forms provide priority and higher scores for persons with specific

needs.


number of projects implemented with the criteria for PwSN and without the criteria;

b) availability of official protocol for PwSN inclusion (Circulars); and c) number of

PwSN targeted projects.

“In my project there is a criteria. In the scoring system, the first and priority

criteria are these people – PwSN. The executive committee of community based

organisations also prioritise these people. The existence of criteria to prioritise

the needs of PwSN shows that the concerns of PwSN are prioritised” (Interview

#13)


(4) Government/NGO support services for PwSN: The key government support for

PwSN is the social safety net programs. It is regularly implemented by the respective

government ministries such as social service ministry that provide monthly and annual

assistance to elderly, persons with disability, women headed households, and for

people live under the poverty line.


resource allocation for PwSN from the government budget; b) NGOs or CBOs

available for PwSN; and c) number of government services available targeting PwSN.

“In each divisional office, there is a Social Service Officer (SSO). He has details

of people with specific needs who were supported with wheel chairs, livelihood

support and also how many organisations are working with them. There is a

regular allocation from the government, however, the government has imposed

many criteria to provide livelihood support and houses for PwSN” (Interview

#18)




case studies.


Case study

locations/area

Surrogates to measure involvement of persons with

specific needs in disaster management

S41 S42 S43 S44 S45 S46

Case study #1

(KMT) X X X X

Case study #2 (SM) X X X X

Case study #3

(KMM) X X X X

Case study #4 (sub-


S41: Social safety programs for PwSN

S42: PwSN Committees/groups or representation of PwSN in committees

S43: Organisations/projects for PwSN

S44: Participation of PwSN in training programs

S45: Disaster facilities with access/priority to PwSN


S46: Planning and resources for PwSN (Organisational/government support)


with the target indicator and assessment protocols from the synthesis of interview

responses are summarised below.

1. Social safety programs for Persons with Specific Needs (S41): This surrogate was

highlighted in all four case studies. Social safety programs are the key programs

available for PwSN to ensure their inclusiveness in a community. Through social

safety programs by the state, it ensures the rights of the PwSN and provide equal access

to resources to avoid discrimination for those people who need special care. In the

context of this research, some programs highlighted during the interviews include

monthly payments for elderly over 60 years age, women headed households, and

persons with disability. Social security funds established to support marginalised

families and families living under the poverty line is key to ensure equity for PwSN.

The measurement of this surrogate (S41) should include the measure of people

with specific needs in the division, the percentage of people with specific needs such

as elderly, women headed households, and disabled in disaster situations included in

social safety assistance programs, the trend in social safety program subscribers, and

identifying the disaster preparedness related needs that can be addressed through social

safety programs.

2. PwSN Committees/groups or representation of PwSN in committees (S42): This

surrogate was highlighted in three case studies (case studies #1, #2, and #3). The

involvement and equality for PwSN can be measured using the inclusion of PwSN in

committees in their own community. The level of representation of PwSN in general

committees working in the community shows that their voice is heard, they are

provided with equal opportunity and there is minimal discrimination. In some places,

separate groups are also formed specific to PwSN to ensure their collective ideas are

gathered in assessment planning, and implementation of projects.

The surrogate (S42) measurement protocol can include the measure of specific

committees available for PwSN to the PwSN population, the number of general

committees in which PwSN has adequate representation, and percentage of PwSN in

decision making positions in the divisional area.


3. Organisations/projects for PwSN (S43): This surrogate was highlighted in all four

case studies. The availability of projects targeting PwSN to ensure their needs are

properly catered is a key factor in measuring the equity for PwSN. The number of

projects specifically designed for PwSN by the government and NGOs can indicate

the level of consideration given to PwSN in a specific community. In addition to the

projects, there are also organisations formed for PwSN that represent the voice of

PwSN. The number of projects implemented by PwSN organisations and by the

government or non-governmental organisations is also a good measure of the

involvement of PwSN in a community.

The surrogate (S43) can be measured by the number of projects implemented

targeting PwSN alone, the number of organisations formed specific to PwSN, and the

percentage of people targeted under the projects to the total PwSN population.

4. Participation of persons with specific needs in training programs (S44): This

surrogate was highlighted only in case study #2. The involvement of PwSN in training

programs to build their capacity is a measure of ensuring equity for people with

specific needs in a disaster situation. Some training and awareness programs to raise

their awareness on disaster risk reduction such as early warning, evacuation and

recovery can help to enhance their knowledge to effectively undertake disaster risk

management. Awareness programs or disaster drills are widely conducted regularly to

prepare people for emerging disasters. However, knowing that the most vulnerable

segment of the population in any disaster context are people with specific needs, it is

important to highlight the availability of training programs specific to PwSN and their

participation in training/awareness programs.

The surrogate (S44) can be measured by the number of training/awareness

programs organised that target or include PwSN, percentage of participation of PwSN

in the general training/awareness programs to the total PwSN population, and the

frequency of the training/awareness/disaster drill programs conducted for PwSN.

5. Disaster facilities with access/priority to Persons with Specific Needs (S45): This

surrogate was selected as the fifth surrogate from case studies #1 and #4. In a disaster

context, one of the key factors for PwSN is the accessibility of infrastructure facilities

by PwSN. The availability of access to critical infrastructure such as hospital buildings


that need to be frequently accessed by people with specific needs is important to

measure equity for people with specific needs. In a disaster context, the accessibility

of evacuation centres for people with disability has been a concern. The availability of

latrines for men and women separately, for people with disability, and elder-friendly

access are some example of such disaster facilities to provide priority to PwSN.

The surrogate (S45) can be measured by the accessibility audit of critical

infrastructure such as hospitals and community buildings, accessibility audit for

evacuation centres including facilities for PwSN, and the availability of disability aids

for people with disability.

6. Planning and resources for PwSN (Organisational/government support) (S46):

This surrogate was highlighted as a key facet of the indicator – ‘involvement of persons

with specific needs in disaster management’ in case studies #3 and #4. The availability

of a plan and resources based on the plan from the government and NGOs can be a

surrogate to measure involvement/equity for PwSN. The social services departments

in each divisional area prepare an annual plan for supporting PwSN and allocate

resources accordingly to ensure the needs of PwSN are met. Social safety programs

are devised to assist the population with specific needs in order to reduce inequality.

However, they require additional resource allocation to fulfil their basic needs to live

with dignity.

The surrogate (S46) can be measured by the percentage of financial allocation per

annum from the total budget to PwSN of the division, the support from the NGOs

(percentage of contribution by NGOs to PwSN), and the allocation of resources for

PwSN at the divisional level (For example the budget and expenditure of the PwSN

organisations).



The study initially identified 81 concepts and 35 themes from the four case

studies for indicator #4. Similar themes were then aggregated and six higher-order

themes were identified through cross-case synthesis which can be considered as

potential surrogates to measure involvement of persons with specific needs in disaster

management. The summary of the synthesis for the fourth social resilience indicator –

involvement of PwSN as social equity is shown in Figure 5.21 below.

The following three surrogate – S41: Social safety programs for Persons with

Specific Needs, S42: PwSN committees/groups or representation of PwSN in

committees, and S43: Organisations/projects for Persons with Specific Needs, have

high validity, since they were identified in three of the four case studies.

Another set of relevant higher-order themes (S44: Participation of persons with

specific needs in training programs; S45: Disaster facilities with access/priority to

Persons with Specific Needs; and S46: Planning and resources for PwSN

(Organisational/government support)) were found in at least one of the case studies,

can also be potential surrogates to measure the equity/involvement of persons with

specific needs in a disaster context. Hence, six potential surrogates to measure the

involvement of persons with specific needs in disaster management are highly reliable

and practically applicable in similar contexts elsewhere, since they were identified in

consultation with practitioners and policy makers who are highly experienced in

implementing disaster management activities at the community level.

Figure 5.21. Summary of synthesis for the fourth social resilience indicator

– social equity


SURROGATE MEASURES TO ASSESS CULTURAL/BEHAVIOURAL

NORMS AS SOCIAL BELIEFS IN A DISASTER CONTEXT

(INDICATOR #5)


surrogates for measuring the fifth social resilience indicator selected in this study –

cultural and behavioural norms in disasters as social beliefs. The cultural and

behavioural norms in a disaster context refers to the influences of beliefs in the

interpretation of disaster risks (Eiser et al., 2012).


In case study #1, eight themes were generated and four combined themes (higher-order



1. Involvement of religious institutions in disaster relief and response

activities;

2. Religious places as evacuation centres with facilities;

3. Government disaster management plan integrating religious institutions/

faith-based organisations;

4. Role of faith-based organisations/leaders in social trust.

(1) Involvement of religious institutions in disaster relief and response activities:

Not all religious institutions are active in disaster relief and response activities.

However, in strong faith based communities, religious institution is one of the key

decision makers. It is very prominent in disaster situations that religious organisations

themselves extend support, not only to their own population, but to the neighbouring

population or sometimes people affected by disasters in other regions as well.


past records of disaster relief activities of religious institutions, b) availability of

Disaster Management (DM) volunteer force/network in religious institutions, and c)

availability of DM plans/DM committees by the religious institutions/faith-based

organisations.

“When there is a disaster, no ethnic or religious bias. When there is a

disaster people from different ethnicities and religious backgrounds come to

help people. The damaged houses were repaired by the international


organisations although they are Christians. During disaster time, there is no

division between ethnicity and religious backgrounds. There are religious

institutions providing relief items. For disaster relief, other ethnic people

send relief items to us. When disasters come, people come together to help

other community” (Interview #41)

(2) Religious places as evacuation centres with facilities: As religious institutions

play a key role, in many occasions people tend to seek safety in religious places when

they evacuate. People tend to take shelter in religious places. However, the facilities

available in religious places to accommodate people including the most vulnerable

population such as the people with specific needs is the key measure of social

resilience. Since some religious places may not be suitable for accommodating people

because they may lack basic facilities that could lead to more harm than good. Hence,

the measure of availability of religious places as evacuation centres with basic facilities

(Mosques, Temples, and Churches) can be a surrogate for existent religious/cultural

practices in relation to disaster management.

3

2

1

4


case study #1



percentage of religious places that have facilities for evacuation centres; b) percentage

of population that is regularly attached to religious places for their religious/social

activities; and c) percentage of people accommodated in all demarcated religious places

as evacuation centres.

“Religious places are also used as an evacuation places and people move

immediately to accommodate to these religious centres for first

accommodation and to safe guard themselves. There are Participatory Rural

Appraisal tools to understand these information. There are many tools that

could help us to understand what type of role the religious institutions can

play” (Interview #04)

(3) Government disaster management plan integrating religious institutions/

faith-based organisations: In strong faith-based communities, the religious

institutions and faith-based organisations play a key role in decision making on

community issues such as in disaster situations. In these communities, the role of

religious institutions and faith-based organisations/networks cannot be ruled out

when disaster management plans are developed by the local authorities. Hence, the

measure of role played by religious institutions and faith-based organisations is an

important factor in determining social resilience to disasters under the dimension of

social/cultural belief systems.


participation of faith-based organisations/religious institutions in disaster

management committee (members); b) involvement of religious institutions/faith-

based organisations in disaster management planning process; and c) participation of

faith-based organisations/religious institutions in disaster management activities by

government.

“When there is heavy rain or flood or even during droughts, there are social

and religious events organised. There are some special religious sermons

and prayers when there is a drought. People come together and it is a time

that people show a social cohesion and it is an opportunity for people to

gather for a common cause. When there is a disaster, people link it to

religious believes. For droughts, people link it to religious reasons.

Government also organise religious events to bring community together and


we can see how many people participate in those prayers and religious

functions. It shows their strength of the religious beliefs in disaster related

understanding and knowledge” (Interview #27)

(4) Role of faith-based organisations/leaders in social trust: This case study

emphasised on the trust on the faith-based organisation and religious

institutions/leaders. For example, the trust on early warning messages disseminated

from the religious institutions and leaders are more likely to be believed by the

community than from other sources when the community has stronger faith

orientations.


available faith-based organisations in the area; b) degree of social trust on faith-based

organisations (Degree of participation of people in religious events); and c)

participation of faith-based leaders in community issues.

“When there is a disaster early warning, temples and mosques and other

cultural centres are the first to disseminate information. Even in the night at

10 or 11 o’clock, the head of the temple informs to the public that they need

to be vigilant about an impending disaster. People believe these messages,

because it is coming from religious places. When a disaster information is

given by a normal person, people have more believe that the religious leaders

and spiritual leaders will not tell anything without confirmation. So there is a

stronger believe that people trust religious leaders and messages come from

religious institutions” (Interview #28)


In case study #2, nine distinct themes were generated. By combining closely related

themes, three higher-order themes were identified (Figure 5.23). These include:

1. Faith-based organisations/activities;

2. Early warning information/support by religious institutions;

3. Culture of women in the society.


(1) Faith-based organisations/activities: Faith-based organisations and their

activities have played an important role in disaster management work before and after

disasters. The motto of helping people becomes the centre of their activities. Hence,

faith based organisations comes under cultural division of the government

administration in the case study area.


ratio of faith based organisations to other community based organisations; b) level of

engagement of faith based organisations; and c) the work of community based

organisations in faith.

“Religious organisations will never do harm as they always try to help people.

For people, they show more interest in the religion. When there was war in this

country, people take self-defence in mosques for two reasons: one, people have


case study #2

1

3

2


more faith in their religion and second, when people stay in religious places, it

is very safe. So people seek shelter in religious places like mosques”. (Interview

#20)

(2) Early warning information/support by religious institutions: One of the nine

themes generated from this case study interview data analysis is ‘information’ that

indicate early warning and information. The other connected key concepts to

information/early warning is mosques (i.e. religious institutions) and their support in

disseminating information and early warning related to disaster risks.


role of religious institutions in Disaster Risk Reduction (DRR) for early warning and

information dissemination; b) support of religious leaders/active participation in DRR

work; c) facilities available in religious places as evacuation centres for the use during

evacuation; and d) early warning facilities available in religious places such as

mosques.

“The only information sharing method available now is to use siren. This society

is centred and linked to religious institutions and mosques. In our area, the

network and proximity of religious places available is a positive aspect, since it

is widely available and can be used for transmitting information very easily”

(Interview #31)

When during Tsunami early warning is issued asking people to move away from

the sea, people will go to temple even the Temple is in the coastal area that is

risky to Tsunami. We have pointed out in the evacuation map that people should

evacuate in the direction out of coastal area that is away from the Temple. We

have not shown the Temple as an evacuation centre, and we have indicated

school or community centre as an evacuation centre. But still people do go to

Temple in the wrong evacuation direction” (Interview #13)

(3) Culture of women in the society: Many interview participants in this case study

highlighted the cultural barriers exist for women to actively participate in disaster

related activities such as in evacuation drills and first aid training. The existent

cultural practice of women not participating in public forums and programs pose an

enormous challenge to prepare them for disaster risk and response activities.



women participation in disaster related training such as first aid; b) gender sensitive

disaster programs; and c) gender related restrictions and social norms.

“Sometimes religious and cultural beliefs can give greater impact. There are

some religious restrictions such as men and women do not mingle together.

For young girls, we need to find separate accommodation during disasters.

Generally women are separated and they will be hosted in a separate

evacuation centre. When we keep all men and women together, there are some

debates. People give more priority for religious beliefs and practice over than

any other issues” (Interview #12)


In case study #3, six themes were generated and three combined themes identified

(Figure 5.24). These include:

1. Cultural role of women in the society;

2. Awareness raising through faith-based activities;

3. Faith-oriented practices during disasters.

Figure 5.24. Leximancer concept/theme maps from interview

data of case study #3

1

2

3


(1) Cultural role of women in the society: The culture of women has been a dominant

factor in existent cultural/religious practices that impede disaster preparedness of the

community. Since case study #2 and #3 are Muslim dominated society (100% Muslim

urban areas), these two areas have similar themes.


degree of social activities of women that are considered to be acceptable among the

strong faith-based society; b) degree of religious norms followed with regard to

women’s dress; and c) degree of women participation in trainings such as first aid and

evacuation.

“In some societies, there are regulations or norms for women. Even in

emergencies, Muslim women do not go out without another person. We

experienced similar problems during previous disasters. More women such

as widows or women headed households lost their lives. My opinion is that

in the Muslim culture, if a women can dress properly within the religious

norms, she can go out to earn her living and search for livelihood. Access to

information is the key lacking part for women. They do not know what

happens during disasters” (Interview #42)

(2) Awareness raising through faith-based activities: The faith-based engagement

of the society has many positive factors in enhancing resilience. The strong faith-based

engagement is capitalised and can be used as an opportunity to raise awareness about

disaster impacts. Regular engagement in religious activities in public platforms are

good opportunities to raise awareness on disaster related messaging to the public,

which can have wider reach and internalised effectively by the community due to their

strong religious beliefs.


awareness programs held in religious institutions such as sermons; b) awareness

programs and capacity building trainings for religious leaders on disaster management;

and c) available faith-based mechanisms/engagement strategies by the disaster

management committee promoting disaster resilience.

“People mostly believe that the disasters and their damage are god’s will

and people do not most often show interest to prepare for disasters. Mosques

provide awareness programs. During Friday sermons, the religious leaders


and priests make people aware and give more knowledge to people about

disasters and preparedness activities” (Interview #29)

(3) Faith-oriented practices during disasters: Some communities which are rooted

in religious beliefs and strongly believe in their traditional religious practices, observe

some faith-based practices before, during, and after disasters. For example, in the

Muslim society, there has been a religious belief and practice of large gatherings for

prayers during drought seasons asking for rain from God. Similarly during flooding or

excessive rain or any other disaster, to plead with God through special prayers to

safeguard the community from danger. Similar practices are also practised in other

religions as well, such as in Hindu and Buddhism where there are special prayers and

alms giving ceremonies being in Temples.


available religious practices in different stages of a disaster cycle; b) participation of

the public in these religious practices; and c) participation of the population in annual

disaster remembrance days conducted by religious institutions.

“In Tamil culture, seeds are saved underground. When there is a disaster like

cyclone, it is safe under the ground. For example when there is no rain,

Muslims pray and gather asking their god for rain…. A prayer is the key

believe among Muslims, when they face hardships. The coordination between

faith-based organisations is a good indication. There are many factions

among Muslims based on faith beliefs and differences. When there is a

disaster, they coordinate and work. During the normal time, there are many

difference and divisions. Faith based divisions are in normal time reduce the

cohesion before disasters (during the normal times)” (Interview #32)


In case study #4, ten themes were generated and five combined higher-order themes


1. Available faith based organisations and support;

2. Culture in evacuation of women;

3. Cultural/religious barriers for preparedness;

4. Role of religious institutions in disaster management;


5. Trust in early warning messages.

(1) Available faith based organisations and support: The role of faith based

organisations in disaster management, particularly in disaster response and recovery

have been consistently highlighted by many interview participants during this study.

Sri Lanka, being a multi religious society, with strong religious beliefs rooted in the

community, the role of faith in disasters largely need to be capitalised for positive

outcomes, according to many interview participants. Although there are barriers for

enhancing community resilience to disasters due to some religious beliefs that

negatively impact disaster preparedness and resilience, the close interaction and belief

systems of the population with faith based organisation and leaders play a key role in

educating the public on the importance of disaster resilience.


case study #4

1

4

3

5

2



available faith-based organisations (Number of FBOs registered in the area); b)

projects involved or implemented in disaster management; and c) existence of a

disaster management structure within the organisation.

“There are many religious organisations within each religion. There is a

catholic youth union. These are very small organisations not a big NGO. A

support comes from the religious people and when a priest asks for support

using a loud speaker, they announce in the community, immediately support

comes. Because, when a priest says that they need to support people affected

by disasters, people immediately support. There is no need for confirmation.

When there is a fire in a house announced by a priest, people will not look at

where it happened. They will just support because the Priest or Moulavi asked

this help. Since this belief exists, immediately they collect money and support

when there is a disaster. The message is immediately passed through and there

is no need for confirmation, since their belief is so high” (Interview #13)

(2) Cultural/religious barriers for preparedness: There are many barriers for

disaster preparedness activities in a strong faith based community due to some myths

and misunderstood religious beliefs in relation to disaster preparedness. Although

many interview participants believed that religion can play a big role in advocating for

better disaster preparedness, the lack of understanding about disaster management

scientifically pose a serious barrier among the faith leaders and followers to use

religious practices to prepare for disasters. The culture of disaster preparedness should

be promoted decrying the barriers in religious practices.


participation of religious leaders in disaster preparedness projects; b) number of

disaster preparedness projects implemented by religious institutions/faith based

organisations; and c) percentage of religious leaders trained in disaster management.

“We can look at cultural practices positively. In some societies we can’t see

‘working together’ culture. In some cultures, there are some barriers in

preparedness. But, for disaster response, it is positive. There is a culture of

post disaster work, but they do not think that this could be better done for

preparedness work. For example, when there was a flood in Colombo, Council

of Muslim priests, a religious coordination body went and supported after flood


in post disaster work. They did not go after flood to check the preparedness

needs of families affected by past disasters to prepare for them for future

disasters. Donors are also emotionally influenced towards disaster response”.

(Interview #10)

(3) Culture in evacuation of women: Gender is another important parameter in

disaster management, particularly in the strong faith based communities. There are

many restrictions imposed on women in the name of following faith and cultural

norms. For example, women are afraid to evacuate alone during emergencies, when a

household is women headed, because of the fear that the community will see these

women in a different perspective. This fear makes women feel insecure and restrict

their mobility in disaster situations.


participation of women in disaster management related training such as swimming and

first aid; b) percentage of women headed households evacuated in the past disaster

early warning/past death tolls of women and other vulnerable groups; and c) trend in

the rate of crimes against women and Gender Based Violence.

“If we look at individual preparedness, there are restrictions due to cultural

norms and gender related issues for women. Women are more vulnerable, but

it is very difficult to train them for swimming in our culture. For example when

we do trainings for disaster evacuation such as swimming, we can check how

many women enrol in swimming training. This is a good measure. We can

also check in other training such as first aid training, how many women attend

these trainings” (Interview #10)

(4) Role of religious institutions in disaster management: The role of religious

institutions in disaster management has been widely acknowledged as an important

factor in a strong faith-based community. People tend to listen and trust more, religious

leaders than civic leaders. However, there is lack of initiatives by religious institutions

on disaster preparedness due to lack of understanding and knowledge about disaster

management among them. However, with limited understanding and knowledge on

disaster management, religious institutions can easily mobilise volunteers during

disasters due to their extended volunteer network developed on the basis of religious

engagement with youth and adults. They can become the forefront of disaster response

and relief activities such as evacuation, relief collection, and distribution.



available facilities in religious institutions for disaster preparedness; b) organisation

and participation in disaster management activities and projects; and c) available

disaster related support mechanisms in religious institutions.

“To accommodate people during disasters, mosques are very important. Since

they have two stories building, people go to mosques for evacuation. However,

preparedness was less in these religious institutions. There are information

centres such as Police, Government Administration office, Local Authority

Office. In some places, these institutions have not developed a good network

with CBOs and other organisations, such as religious institutions”.

(Interview #11)

(5) Trust in early warning messages: Early warning is the critical component of any

disaster response. Hence, the preparedness measures for early warning dissemination

to the communities that are most vulnerable to disasters become a key factor. In order

to act timely and properly, the community at risk should trust the early warning

message to take necessary actions according to the warning message disseminated.

However, there seems to be a lack of trust on warning messages among the community

issued by the media or government institutions due to past experience of false

warnings. When it comes to warning dissemination, the trust was built on religious

institutions and leaders for generations.


percentage of religious institutions as early warning dissemination centres; b) religious

institutions equipped with facilities for early warning; and c) degree of participation

of religious institutions in early warning drills.

“Unless the disaster comes to their step, they don’t believe in it. Even if we are

saying that there is a threat that we are expecting a disaster and there is a

warning, there is a possibility for a disaster tomorrow or day after tomorrow,

people do not listen. People trust religious leaders and trust whatever they say.

Similarly in other areas, priests are given utmost respect. They listen to the

chief priest of the temple. They are also doing lot of work” (Interview #17)





case studies.

Table 5.5.Surrogates (higher-order themes) mapping for indicator#5

Case study

locations/area

Surrogates to measure cultural and behavioural

norms in disaster management (Emerging

themes)

S51 S52 S53 S54 S55 S56

Case study #1 (KMT) X X X X


Case study #3 (KMM) X X X

Case study #4

(sub-national level) X X X X

S51: Faith-based organisations/practices/activities

S52: Culture of women in the society

S53: Involvement of religious institutions in disaster preparedness/response activities

S54: Early warning dissemination trust on religious institutions

S55: Religious places as evacuation centres with facilities

S56: Government disaster management plan integrating religious institutions/ faith-

based organisations

For each of the potential surrogates, a brief description of the surrogate, its

relationship with the target indicator, and assessment protocols from the synthesis of

interview responses are summarised below.

1. Faith-based organisations/practices/activities (S51): The communities with strong

religious beliefs have faith-based organisations. Past evidence shows that faith-based

organisations have played a major role in disaster response and recovery activities.

Although faith-based organisations are not trained professionally to work in disaster

management, their activities such as leading an evacuation work, most importantly in

critical situations cannot be underestimated. However, they lack strategies in disaster

preparedness work and are often influenced by faith based values such as helping each

other in difficult times and place the wellbeing of the community as a priority,

particularly in hardship. This surrogate was highlighted in all four case studies.


The surrogate (S51) measurement protocol can include the measure of faith-based

organisations registered and active in the division, the faith-based organisations which

have disaster management in their portfolio, percentage of faith-based organisations

that have the volunteer base for disaster relief work, and percentage of population

attached to the faith-based organisations in the division.

2. Culture of women in society (S52): In certain cultures and faith based communities,

the restrictions for women in many aspects of social life can be seen. For example, in

Muslim villages, lack of mobility of women alone in public have been a social norm

and people believe that it has many advantages in terms of reduction in harassment

and crimes against women. However, the limitations of women being restricted

culturally, such as the need for the same dress code during an emergency, has

negatively impacted on the ability of rescuing them during disasters. The existing

cultural norms as to how the women are treated in a society is a good indicator of

existent cultural and behavioural norms that can impede or assist in disaster

management activities. This surrogate was highlighted in three case studies except case

study #1.

The surrogate (S52) can be measured by the number of women volunteers trained

in first aid, swimming and other critical disaster management skills, extent of women

volunteer networks/organisations working in disaster management projects, and the

percentage of the women population employed against the women population

employed in critical jobs such as police and emergency health services.

3. Involvement of religious institutions in disaster preparedness, relief and response

activities (S53): This surrogate was highlighted as a key facet of the indicator –‘cultural

and behavioural norms in disaster management’ in case studies #1, #3, and #4. When

there is a stronger trust on religious institutions and faith-based engagement by a

community, preparing religious places as evacuation centres to accommodate people

in times of disasters is not sufficient and effective. The involvement of religious

leaders and faith based organisations proactively in all phases of disaster management

cycle will help productive and effective plans and projects. However, the degree of

belief on such measures and willingness to support preparedness and resilience

building activities can be a good surrogate to measure the existent local cultural and

religious beliefs of a community.


The surrogate (S53) can be measured by the religious leaders trained on disaster

management skills such as search and rescue/other volunteer basic skills such as first

aid, disaster preparedness activities engaged by religious institutions, and the

participation of committee members of the religious institutions in disaster

management committee activities or other government projects.

4. Early warning dissemination trust on religious institutions (S54): The trust on

religious institutions such as Mosques, Temples, and Churches can be indicated by the

issuance of early warning and disseminations. The trust on religious institutions is

higher, when the government systems lack good systems and facilities to educate the

public on early warning. People consider religious leaders as trustworthy and

therefore, local authorities utilise the religious institutions and leaders to disseminate

the disaster early warning and to raise awareness on disaster risk reduction. The trust

is already built with religious institutions in strong faith-based communities through

participation in religious activities. This surrogate was highlighted in two case studies

(case studies #2 and #4).

The surrogate (S54) can be measured by the facilities available for early warning

dissemination by religious institutions, the percentage of population who attend

religious institutions regularly (some minimum attachment), the coverage of early

warning dissemination through religious places, and the ability and available

mechanism or organisational structure for early warning dissemination.

5. Religious places as evacuation centres with facilities (S55): This surrogate was

highlighted only in case study #1. The religious places, schools, and community

buildings are demarcated as evacuation centres during disasters. However, better

facilities for accommodating people particularly taking care of people with specific

needs in those evacuation centres become the priority for any disaster management

agency and for the local authorities. People from faith oriented backgrounds seek

shelter in religious places than other infrastructure facilities. Hence, the ability of the

religious places with sufficient facilities to accommodate people evacuated during

disasters can be a surrogate to measure the existent cultural and religious beliefs that

can help in managing disaster situations.

Further, the measurement of this surrogate (S55) should include the measure of

religious places demarcated as evacuation centres, the past occupancy of the religious

places vs other places demarcated as evacuation centres, the number of religious places


available with and without evacuation centre facilities, and the disaster awareness

programs conducted at the religious based evacuation centres.

6. Government disaster management plan integrating religious institutions/ faith-

based organisations (S56): This surrogate was selected as the fifth surrogate from case

study #1. The role of faith based organisations in a strongly religious community

cannot be overlooked due to long term trust built by them with the community. The

local authorities leading disaster management planning process and implementing

projects or actions have invited and included faith-based organisations and their

leaders/members to be part of the disaster management committee. The active

participation and proactive initiatives by the faith-based organisations will help to take

forward the implementation of disaster management activities proposed in the disaster

management plan.

The surrogate (S56) can be assessed by the measure of number of local cultural

and faith based organisations in the disaster management committee,

activities/projects by the cultural and faith based organisations with the government in

disaster preparedness work, and the capacity of cultural and faith based committees to

engage in disaster preparedness work (volunteer base, financial capacity, and

infrastructure or equipment etc.).


The study initially identified 74 concepts and 33 themes from four case studies

for indicator #5. Similar themes were then aggregated and six higher-order themes

were identified through cross-case synthesis which can be considered as potential

surrogates to measure cultural and behavioural norms and practices in disaster

Figure 5.26. Summary of synthesis for fifth social resilience indicator –

social beliefs


management. The summary of synthesis for the fifth social resilience indicator – social

beliefs is shown in Figure 5.26.

The following three surrogates, S51: Faith based

organisations/practices/activities, S52: Culture of women in the society, and S53:

Involvement of religious institutions in disaster preparedness, relief and response

activities have high validity, since they were identified in three of the four case studies.

Another set of relevant higher-order themes (S54: Early warning dissemination

trust on religious institutions, S55: Religious places as evacuation centres with

facilities, and S56: Government disaster management plan integrating religious

institutions/ faith-based organisations) were found in at least one of the case studies,

can also be potential surrogates to measure the cultural and behavioural norms and

practices in disaster management. Hence, six potential surrogates to measure the

cultural and behavioural norms and practices in disaster management are highly

reliable and practically applicable in similar contexts, since they were identified in

consultation with practitioners and policy makers who are highly experienced in

implementing disaster management activities at community level.


SUMMARY OF POTENTIAL SURROGATES AND SELECTION OF

SET OF SURROGATES FOR EVALUATION IN PHASE II

The potential surrogates were obtained by analysing the higher-order themes

generated from the case study data analysis. All potential surrogates for the five social

resilience indicators identified from the case studies analysis, as shown in Sections 5.1

to 5.5, were compared and consolidated in Table 5.6 below.

Table 5.6. Cross-case tabulation for potential surrogates

Potential surrogates

Case

Study

#1

Case

Study

#2

Case

Study

#3

Case

Study

#4

Indicator 1: mobility and access to transport facilities

S11: Available transport facilities

targeting PwSN X X X X

S12: Evacuation places and centres X X X X

S13: Awareness programs/Early

warning systems X X X X

S14: Evacuation routes and plans X X

S15: Social support systems X

S16: Emergency information

dissemination and sources X

Indicator 2: social trust in pre and post disaster phases

S21: Effectiveness of community

based organisations/social service X X X X

S22: Level of services and resources of

local authorities/state support

for people

X X X X

S23: Functioning and effectiveness of

disaster management systems and

complain mechanisms

X X X

S24: Trust on information sharing/

early warning dissemination X

S25: Level of public support for

awareness and public programs X

S26: Public-government officer

relationship X

Indicator 3: learning from the past disaster experience

S31: Reaction to disaster early warning X X X X

S32: Awareness and disaster

knowledge level X X X


S33: New DRR programs such as

innovative construction methods X X X

S34: Functional level of the disaster

management committee/members X X

S35: Disaster Risk Reduction (DRR)

and Disaster Preparedness

strategy

X

S36: Evacuation drills & disaster

preparedness kits X X

Indicator 4: involvement of People with Specific Needs (PwSN)

S41: Social safety programs for PwSN X X X X

S42: PwSN committees/groups or

representation of PwSN in

committees

X X X

S43: Organisations/projects for PwSN X X X X

S44: Participation of PwSN in training

programs X

S45: Disaster facilities with

access/priority to PwSN X X

S46: Planning and resources for PwSN

(organisational and government

support)

X X

Indicator 5: existent religious, cultural and behavioural practices

S51: Faith-based

organisations/practices and

activities

X X X X

S52: Culture of women in the society

(Gender norms) X X X

S53: Involvement of religious

institutions in disaster

preparedness, relief and response

activities

X X X

S54: Early warning dissemination trust

on religious institutions X X

S55: Religious places as evacuation

centres with facilities X

S56: Government disaster management

plan integrating religious

institutions/faith-based

organisations

X


There are six potential surrogates for each of the five social resilience indicators

as shown in Table 5.6. However, some of the surrogates were only identified in one or

two of the case studies among the four case studies. In order to keep the number of

surrogates manageable for an online survey, it was decided to limit it to surrogates that

have a higher validity, which were identified in multiple case studies. Therefore,

potential surrogates that were identified in minimum three of the four case studies were

selected for evaluation in phase II. This made a total of 15 potential surrogates for

ranking by experts.

Three potential surrogates for each of the five key social resilience indicators

that were identified at least in three of the four case study areas that were selected for

evaluation and ranking in the next phase of this research is listed in Table 5.7

Table 5.7. Potential surrogates selected for evaluation in online survey (phase II)

Resilience indicator Potential surrogate measures

I1: Measuring ‘social

mobility and access to

transport facilities’

S11: Available transport facilities targeting PwSN

S12: Evacuation places and centres

S13: Awareness programs/Early warning systems

I2: Measuring ‘social

trust’

S21: Effectiveness of community based

organisations/social service

S22: Level of services and resources of local

authorities/state support for people

S23: Functioning and effectiveness of disaster

management systems and complain mechanisms

I3: Measuring the

‘learnings from the

past’

S31: Reaction to disaster early warning

S32: Awareness and disaster knowledge level

S33: New DRR programs such as innovative

construction methods

I4:Measuring

‘involvement/equity for

persons with specific

needs (PWSN)’

S41: Social safety programs for PwSN

S42: PwSN committees/groups or representation of

PwSN in committees

S43: Organisations/projects for PwSN

I5: Measuring

‘cultural/religious

norms and practices’

S51: Faith-based organisations/practices and activities

S52: Culture of women in the society (Gender norms)

S53: Involvement of religious institutions in disaster

preparedness, relief and response activities


Chapter 6: Survey analysis - Evaluation and

ranking of potential surrogates

As shown in Figure 6.1, this chapter has five key sections including the last

section (Section 6.5) as a summary of findings from phase II online survey.

Section 6.1 presents the key findings from the phase II of this research, which

aimed to evaluate and rank the potential surrogates identified in phase I of the study.

Those potential surrogates to assess social resilience were evaluated by international

and national disaster management experts. The results of the survey were analysed

using the multi-criteria decision making with equal criteria weights.

Section 6.2 presents the calculation of weights for each criterion from the survey

inputs by disaster management experts.

Section 6.3 provides a summary of the PROMETHEE ranking done with the

weights for the surrogate evaluation criteria obtained from the experts’ judgement

through the phase II survey (as shown in Section 6.2).

Section 6.4 presents the list of first ranked social resilience surrogates that was

selected for application as a final output from the evaluation and ranking of potential

surrogates.

Finally, Section 6.5 provides a brief summary of the phase II online survey

results.


6.1. Chapter 6 with key sections in the thesis structure


6.1. PROMETHEE RANKING AND ANALYSIS OF POTENTIAL

SURROGATES WITH EQUAL CRITERIA WEIGHTS

The results of PROMETHEE, which was used as a Multi-Expert Multi Criteria

Decision Analysis (ME-MCDA) to rank potential surrogates are presented in Sections

6.1.1 to 6.1.5. Two types of analysis were carried out for each social resilience

indicator:

1. Ranking of potential surrogates: Overall PROMETHEE ranking for each

social resilience indicator (Table 6.1) was based on the analysis of alternatives against

all five criteria as an overall group decision making using the inputs from multi-

experts. The Phi value is the net flow value as a result of the pairwise comparisons of

surrogates.

2. Analysis of opinions about potential surrogates by different cohorts of experts:

It is possible that different types of disaster management experts can have different

preferences for surrogates. Since the experts, who participated in the evaluation of

surrogates, have different degree of experience and also belong to different categories

such as practitioners, researchers, and policy makers, it is important to analyse

similarities and differences between their preferences. GAIA representations are

shown in Figures 6.2a-6.6a for different years of experience ranging from >10 years,

5-10 years, 3-5 years, to <3 years, while GAIA representations are shown in Figures

6.2b-6.6b for different types of experts ranging from practitioners, researchers, and

policy makers.

Table 6.1. Overall PROMETHEE rankings for five social resilience indicators

Resilience

indicator Rank Potential surrogate measures Net Phi

I1: Measuring

‘social

mobility and

access to

transport

facility’ using

surrogates

1 S12: Availability of evacuation places

and centres 0.0318

2 S13: Awareness raising programs/plans

and early warning systems -0.0080

3

S11: Transport facilities available

(emphasis to access transport for

persons with special needs)

-0.0239

I2: Measuring

‘social trust’

using

surrogates

1 S21: Effectiveness of CBO’s

activities/social service 0.0500

2

S22: Level of services and resources of

local authorities/Support for people

from state institutions

-0.0125


3

S23:Functioning and effectiveness of

disaster relief/management system and

complain mechanisms

-0.0375

I3: Measure the

‘learnings

from the past’

using

surrogates

1 S31: Reaction to disaster early warning 0.0750

2 S32: Awareness and disaster

knowledge level 0.0136

3

S33: New DRR programs including

new construction methods (e.g.

Houses)

-0.0886

I4:Measuring

‘involvement/e

quity for

persons with

specific needs

(PWSN)’

using

surrogates

1 S41: Social safety programs for PwSN 0.0466

2 S43: Organizations/projects for PwSN 0.0080

3

S42: PwSN Committees/registered

groups or representation of PwSN in

committees

-0.0545

I5: Measuring

‘cultural/religi

ous norms and

practices’

using

surrogates

1 S52: Culture of women in the society 0.0318

2 S51: Faith-based

organizations/practices/activities 0.0045

3


institutions in disaster preparedness,

relief and response activities.

-0.0364

6.1.1 Indicator #1: Measuring ‘social mobility’

In this study, the surrogate ‘S12 - measure of evacuation places and centres’ was

preferred over the other two surrogates (S13 and S11) by the experts. This could be due

to the higher influence of evacuation potential of the population at-risk for emerging

disasters on social mobility. The identification and demarcation of evacuation places

and centres may have been done by authorities. However, the level of awareness of

evacuation places and centres among the population who are vulnerable to disasters

may be lacking to enable effective social mobility due to the absence of required

facilities for people with specific needs. On the other hand, the evacuation in times of

disasters using vehicles is most often a challenge in urban context, where streets can

be narrow that result in traffic congestion. Hence, the availability of transport facilities

sometimes may not be given first priority for the measurement of social mobility

during disasters.

Opinions of cohorts of experts on surrogates to measure indicator #1

Experts over five years of experience largely preferred the surrogate S12 over S13

and S11. However, the experts with average experience between 3 – 5 years opted


Figure 6.2 (b). GAIA representation of surrogates to measure social

mobility for different types of experts

Figure 6.2 (a). GAIA representation of surrogates to measure social

mobility for experts with varied years of experience

Legends for Figures 6.2 (a) and (b)


S13 as the surrogate measure of first choice. However, the experts with less than three

years of experience out-ranked S13 and S12 over S11. Overall, the results are skewed

towards S12, because the experts who have more than five years of experience form

the majority surveyed (68%). As shown in Figure 6.2a, the orientation of the decision

axis (red thick axis) indicate which cohorts of experts are in agreement with the

PROMETHEE rankings and who are not. Although it is obvious that the availability

of transport facilities and effective early warning messages are key to timely mobility

of people at-risk, the availability of evacuation centres and people’s awareness can be

the most important factor in determining the effectiveness of social mobility according

to the PROMETHEE ranking analysis. This could be the reason, the most experienced

cohort of experts aligned with the overall PROMETHEE ranking of the surrogate

(Figure 6.2a).

Based on occupation, all three groups of experts - practitioners, policy makers,

and researchers - ranked S12 as the most preferred surrogate. The surrogate S13 was

ranked second by policy makers and researchers, however it was the less preferred

option for practitioners. Similarly, Surrogate S11 performed better for practitioners,

compared to researchers and policy makers (Figure 6.2b). Practitioners had a strong

preference for surrogate S12 and S11. However, policy makers had greater preference

for surrogate S13 compared to S11. The surrogate S11 did not perform well among

researchers and policy makers compared to surrogates S12 and S13. Hence, this study

found, practitioners preferred the measure of evacuation places and centres (S12) and

available transport facilities (S11) to predict the mobility and transport accessibility

compared to the measure of awareness programs/early warning (S13). The policy

community prioritised the awareness of the availability of evacuation places/centres

(S13).

6.1.2 Indicator #2: Measuring ‘social trust’

In this study, the surrogate S21 - effectiveness of Community Based

Organizations’ (CBO) activities/social service was ranked first in PROMETHEE

ranking. It was also evident from this study that surrogate S22 – the existing level of

services and resources from the local authorities/state institutions got priority ranking

than surrogate S23 – the functioning and effectiveness of disaster management

mechanisms to measure social trust. In a resource limited local governance, the

functioning and effectiveness of disaster management mechanisms may become weak,


where the trust is built through routine work and development activities. Therefore,

the effectiveness and satisfactory level of routine services and resource allocation to

the community by local authorities (S22) can also be a priority measure of social trust

compared to the surrogate - functioning of disaster management mechanisms (S23).


Figure 6.3(a) shows that the results are skewed towards S21 where the decision

of experts with over 10 years of experience is closer. However, experts with 3-5 years

of experience preferred S22 over S21 and S23. The experts with 5-10 years of experience

were not conclusive about their preference between S21 and S23, while experts with

less than 3 years of experience were not conclusive in preferences between S21 and S22.

The less experienced cohort (less than 5 years of experience) preferred the measure of

service and resources by the local authorities and government departments as a priority

surrogate than the effectiveness of CBOs to assess social trust. In some cases, it is

possible that the measure of services and resource allocation with state departments

and local authorities can be easily done rather than obtaining accurate information for

measuring effectiveness of CBOs in measuring social trust in a disaster context.

The policy and practice groups aligned with the overall PROMETHEE rankings.

However, the highest preference of the research group inclined towards surrogate S22

as shown in Figure 6.3(b), which is similar to the preferences of less experienced

cohort of experts.

6.1.3 Indicator #3: Measuring ‘learnings from the past’

Surrogate S31 – the reaction by the community to disaster early warning

messages was preferred compared to other two surrogates (S32 and S33) to measure

learnings from the past disasters in the PROMETHEE ranking. Although, S32 – the

level of awareness and disaster management knowledge and S33 – new DRR programs

implemented such as new methods of housing construction - can provide a priority

measure of learnings from the past disasters as a key community competency to

disaster resilience. Experts in this study prioritized the reaction to disaster early

warning messages as a good surrogate measure to assess learning from past disasters.


Figure 6.3(a). GAIA representation of surrogates to measure social

trust for experts with varied years of experience

Figure 6.3 (b). GAIA representation of surrogates to measure social

trust for different types of experts

Legend for Figures 6.3(a) and (b)



Experts with experience of over 10 years had similar overall preferences towards

indicator#3 surrogate measures. As shown in Figure 6.4(a), the first preference of

experts with over 10 years of experience and experts with less than 5 years of

experience was S31. Figure 6.4 (b) also indicates that the first ranked surrogate among

researchers and practitioners is S31. Surrogate S32 – the level of awareness and

knowledge about disasters can also be a good measure for learnings from the past

disasters, similar to surrogate S31 – reaction to disaster early warning messages, which

is preferred by policy experts and experts with 5-10 years of experience. More

awareness and better knowledge about disaster risks will lead to better reaction to

disaster early warning messages. However, the measurement complexity of surrogate

S31 – reaction to disaster early warnings can be higher compared to surrogate S32 –

measure of awareness and disaster knowledge level.

6.1.4 Indicator #4: Measuring ‘involvement/equity for persons with specific

needs (PWSN)’

In this study, surrogate S41 - Social safety programs for Persons with Specific

Needs (PwSN) was preferred compared to the other two surrogates (S42 and S43) to

assess the involvement and equity measures. The other two surrogates are S43:

Organizations/projects for PwSN and S42: PwSN Committees/registered groups or

representation of PwSN in committees. The social safety programs focus on vulnerable

groups in a community in a disaster situation. The inclusion of PwSN such as disabled,

elderly, and women headed households in the existing social safety programs helps to

improve their resilience to disasters. Surrogate S43 - availability of projects or specific

organizations to work on PwSN can be another potential surrogate which was ranked

second by the experts. This study found that the surrogate S42 - the representation of

PwSN in committees/the existence of such committees/registered groups as the least

preferred surrogate to indicate the involvement and equity for PwSN. The participation

in committees most often do not completely reflect the active implementation of

projects which can bring real impact. However, the social safety programs can indicate

the tangible involvement of people in building resilience.


Figure 6.4 (b). GAIA representation of surrogates to measure

learnings from the past disasters of different types of experts

Figure 6.4 (a). GAIA representation of surrogates to measure learnings

from the past disasters of experts with varied years of experience

Legend for Figures 6.4 (a) and (b)



The preference of all cohorts of experts retained S41 as the first ranked surrogate,

except the cohort with 5-10 years of experience, who ranked S43 as their first

preference. However, all cohorts ranked S42 as the least preferred surrogate as shown

in Figures 6.5(a) and 6.5(b). Similarly, preferences of practitioners and policy makers

were aligned with the overall ranking of surrogates, whereas the first preference of

researchers was the surrogate S43 – ‘the availability of organizations and projects’

targeting PwSN. This preference was similar to the cohort of experts with 5-10 years

of experience. In conclusion, preference of more experienced practitioners and policy

makers were same as the overall preference of surrogates to measure indicator #4.

Further, all cohorts of experts ranked surrogate S42- ‘representation of PwSN in

committees or in registered groups as a weak surrogate to measure the involvement

and equity for PwSN.

6.1.5 Indicator #5: Measuring ‘cultural/religious norms and practices’

In this study, the first ranked surrogate for measuring cultural/religious norms

and practices was S52: culture of women in society. Among the three surrogates

evaluated by experts, surrogate S51: faith-based organizations/practices/activities was

ranked second and surrogate S53: involvement of religious institutions in disaster

preparedness, relief and response activities was the least preferred surrogate to

measure cultural/religious norms and practices. Hence, the measure of gender based

practices was prioritised as an important surrogate in this study compared to faith-

based practices and involvements. The cultural practices among women, such as their

engagement in public forums and participation in awareness programs may be critical

in determining the social resilience to disasters in many contexts. In the evaluation of

surrogates by experts, it is likely that most of the experts preferred the most critical

factor of the respective resilience indicator as the most preferred surrogate.


As shown in Figures 6.6 (a), highly experienced cohort of experts with more than

5 years differed from the overall ranking and opted surrogate S51 as their first

preference to measure cultural/religious norms and practices. However, experts with

less than 5 years of experience have shown a similar preference to overall ranking.



‘involvement/equity for PWSN’ of different type of experts

Figure 6.5 (a). GAIA representation of surrogates to measure

‘involvement for PwSN’ of experts with varied years of experience

Legend for Figures 6.5(a) and (b)


Similarly, the practitioners and policy makers followed the same preference of overall

ranking, the cohort of researchers differed from overall ranking, selecting S51 as their

first preferred surrogate (Figure 6.6 (b)). From the analysis, researchers viewed the

role of faith based practices as an important element in measuring social beliefs

compared to cultural practices specific to women, since some specific cultural

practices are at times influenced by faith orientations. The measurement of faith based

practices (surrogate S51) in a community can also provide a broader measure including

the gender specific cultural practices that can enhance or deteriorate resilience. The

different cohort of experts have contrasting opinions between S52 and S51 as the priority

surrogate to assess cultural/religious practices and norms.


Figure 6.6 (a). GAIA representation of surrogates to measure

‘cultural/religious practices’ of experts with varied years of experience


‘cultural/religious norms and practices’ of different type of experts

Legend for Figures 6.6 (a) and (b)


CALCULATION OF SURROGATE EVALUATION CRITERIA

WEIGHT

The consolidated Analytic Hierarchy Process resulted in a weightage for each of

the surrogate evaluation criteria (Table 6.2). The results show that time-sensitivity,

communicability and accuracy are the three criteria that gained weight more than 20%.

The other two criteria - measurement complexity (13.8%) and cost-effectiveness

(16%) gained less weightage among the five surrogate evaluation criteria. The results

further showed that the respondents have more concerns on the ability of the surrogates

to measure resilience in different time periods/phases of a disaster and the ability of

surrogates to predict the target indicators accurately.

Table 6.2. Criteria weight obtained from survey responses using AHP

Criterion Criteria name Weight

1 Crit-1 Accuracy 23.9%

2 Crit-2 Cost-effectiveness 16.0%

3 Crit-3 Measurement complexity 13.8%

4 Crit-4 Communicability 22.3%

5 Crit-5 Time-sensitivity 24.0%

The weight for the criteria - cost-effectiveness and measurement complexity are

less than the average 20% (20% is the weight if the equal weightage is applied as it

was done in section 6.1). The weights for other three criteria (accuracy,

communicability, and time-sensitivity), have got weightage more than average 20%.

The calculation of weightage show that accuracy, communicability, and time-

sensitivity are three more important criteria than cost-effectiveness and measurement

complexity for evaluating surrogates, according to the results in this study.

OVERALL PROMETHEE RANKING WITH WEIGHTED CRITERIA

In section 6.1, PROMETHEE ranking of potential surrogates obtained with equal

weights for all five surrogate evaluation criteria was presented. This section presents


PROMETHEE rankings of same potential surrogates obtained with the weighted

criteria (Table 6.3).

Table 6.3. Overall PROMETHEE rankings for five social resilience indicators with

weighted criteria

Resilience

indicator

Ran

k

Alternatives (Potential surrogate

measures) Net Phi

I1:Measuring

‘social mobility

and access to

transport

facility’

1 S12: Availability of evacuation places

and centres 0.0339

2

S13: Awareness raising

programs/plans and early warning

systems

-0.0095

3

S11: Transport facilities available

(emphasis to access transport for

persons with special needs)

-0.0244

I2:Measuring

‘social trust’

1 S21: Effectiveness of CBO’s

activities/social service 0.0641

2

S22: Level of services and resources

of local authorities/Support for

people from state institutions

-0.0253

3

S23:Functioning and effectiveness of

disaster relief/management system

and complain mechanisms

-0.0388

I3:Measure the

‘learnings from

the past’

1 S31: Reaction to disaster early

warning 0.0803

2 S32: Awareness and disaster

knowledge level 0.0322

3

S33: New DRR programs including

new construction methods (e.g.

Houses)

-0.1125

I4:Measuring

‘involvement/eq

uity for persons

with specific

needs (PWSN)’

1 S41: Social safety programs for

PwSN 0.0512

2 S43: Organizations/projects for PwSN 0.0124

3

S42: PwSN Committees/registered

groups or representation of PwSN in

committees

-0.0636

I5: Measuring

‘cultural/religiou

s norms and

practices’

1 S52: Culture of women in the society 0.0347

2 S51: Faith-based

organizations/practices/activities 0.0066

3


institutions in disaster preparedness,

relief and response activities.

-0.0413


The PROMETHEE rankings between equally weighted criteria and consolidated

AHP weighted criteria were compared. There is no difference in the overall ranking,

although net phi values are different. Figures 6.7 to 6.11 show the net flow value

difference between each of the surrogate for equal criteria weights and experts’

consolidated criteria weights. Further, the differences in net flow values between

surrogates for both weights are also very minimal as shown in Figures 6.7 to 6.11.

It could be because the experts’ evaluation of surrogates was mostly consistent

and within a small margin of variation. The small deviation in the Likert scale will not

influence a larger deviation in the net phi value. This means that there is no influence

in ranking as a result of an extreme opinion in the criteria with higher or lower weight.

It can be concluded from this analysis that equal weight for five surrogate evaluation

criteria can be used for evaluating social resilience surrogates to rank and select

surrogates in disaster management

Figure 6.7. Comparison of net flow values (ranking) of surrogates for

indicator #1 [CW – Criteria Weight Equal and Experts]


FIRST RANKED SOCIAL RESILIENCE SURROGATE IN DISASTER

MANAGEMENT

The five first ranked surrogates include one surrogate for each of the five

resilience indicator, which were ranked first as the optimum surrogate among all

surrogate evaluation criteria. These are as follows:

S*1: Availability/access of evacuation centres for measuring I1: ‘social mobility and

access to transport facility’

S*2: Effectiveness/performance of CBO activities/social service for measuring I2:

‘social trust’

S*3: Reaction to disaster early warning for measuring I3: ‘learnings from the past’

S*4: Social safety programs for PwSN for measuring I4: ‘involvement/equity for

persons with specific needs (PWSN)’

S*5: Gender norms/culture of women in the society for measuring I5:

‘cultural/religious norms and practices’.


Figure 6.12. First ranked surrogates for priority assessment of social resilience



SUMMARY OF FINDINGS FROM THE SURVEY RESEARCH

Three potential surrogates for each of the five selected social resilience

indicators were identified in an exploratory case study in a disaster context (in phase

I) and evaluated independently against five criteria by multiple experts consisting of

practitioners, researchers, and policy makers through an online survey. Potential

surrogates were then ranked using multi-criteria group decision support system in

PROMETHEE.

The most-preferred surrogate (first ranked) can be the utmost critical facet of the

respective resilience indicator in a disaster context. Hence, the first ranked surrogate

can provide a fairly good representation of overall resilience of the respective indicator

as it is also mostly very relevant in practice. However, divergent opinions exist among

the different cohort of experts on the overall ranking of surrogates. For example, the

comparison between overall ranking of surrogates and the ranking of different cohort

of experts showed that the preferences of experts with more than five years of

experience from practitioners and policy makers have mostly aligned with overall

ranking of surrogates. Results further revealed that experienced practitioners tend to

opt for surrogates that can be easily measured with the existing data and communicated

without much complexities for effective policy decisions. The results from this study

will also have greater practical applicability in the field and policy decisions, since

more than two-third of the experts are highly experienced practitioners and policy

makers in disaster management.

The third and final key step in the social resilience surrogate framework is the

selection of optimum surrogates for application. This includes three sub-steps: (1)

evaluation of surrogates against five surrogate evaluation criteria, (2) ranking of

potential surrogates based on the evaluation results, and (3) selection of optimum

surrogates for application. In phase II of this study, the first sub-step was done through

an online survey and the second sub-step was executed through the analysis of the data

obtained from the survey. The third sub-step was done by selecting first ranked

surrogates for application.

Chapter 7: Synthesis of key findings 228

Chapter 7: Synthesis of key findings

This chapter has four key sections, as shown in Figure 7.1.

Sections 7.1 to 7.3 brings together key findings that emerged from the analysis

of data in Chapters 4, 5 and 6 in relation to each of the three research objectives (RO1

to RO3) as defined in Section 1.5. The research objectives are aligned with the three

key steps (A to C) of the conceptual surrogate development framework (See Figure

3.6) tested in this study to assess social resilience in a disaster context. These sections

also provide an interpretation of key findings with reference to the literature so as to

highlight the contribution to knowledge and demonstrate achievement of the three

research objectives.

The Section 7.4 details the revisions to the conceptual surrogate development

framework based on the preceding discussions. This revised framework for future

application, is presented as an integrated framework of all three key steps, which

answers the key research question of this study.

7.1. Chapter 7 with key sections in the thesis

structure


KEY FINDINGS IN RELATION TO RESEARCH OBJECTIVE ONE

(RO1)

RO1: To select key social resilience indicators that require surrogate

approach by developing an adaptive and inclusive social resilience

framework

RO1 of this study was tested in key step A of the conceptual surrogate

development framework (See Figure 3.6) through the selection of social resilience

indicators for applying the surrogate approach. The key outcomes in this step are the

development of an inclusive and adaptive ‘5S’ social resilience framework, the

surrogate decision criteria, and the selection of key social resilience indicators for

applying the surrogate approach as shown in Figure 7.2.

A.1: Identify key social resilience indicators:

The identification of social resilience indicators was done by developing an

inclusive and adaptive ‘5S’ social resilience framework, through a critical review of

existent social resilience frameworks, in the literature review and research design

phase. This study found that, a generic, but adaptable framework such as ‘5S’

framework (see Figure 2.5) proposed in this research help to select key social resilience

indicators in a consistent manner to develop the surrogate approach. There has been a

research gap to develop a comprehensive framework that is scientifically grounded,

but can be practically applicable to select resilience indicators for assessment (Keating

et al., 2017). Hence, the ‘5S’ framework in social resilience assessment that was

developed in this study, was used to select indicators methodically to apply the

surrogate approach. This study further found that, most of the social resilience

7.2. Key step A and related outcomes to achieve RO1


indicators in the ‘5S’ framework are either not feasible to measure through direct

methods or the existing census-based measures for them are not adequate, hence

require new methods for measurement, such as a surrogate approach.

The ‘5S’ framework is structured in 5 key dimensions of social resilience by

incorporating 16 characteristics and 46 indicators, mostly process-oriented indicators

(See Figure 2.5). Many existent frameworks in the literature have used outcome-

oriented indicators (Cutter et al., 2008) for resilience assessment rather than process-

oriented indicators (Sharifi, 2016). Many of the existing resilience frameworks were

developed to a specific context, hence, they tend to select specific indicators relevant

to the context, which limits its applicability to a wider context (Woolf et al., 2016).

For example, Kwok et al. (2016) and Khalili et al. (2015) developed context specific

social resilience indicators, which leave out some key indicators relevant in other

context, which make their framework not very adaptable to different contexts.

The existent frameworks mostly considered social resilience characteristics such

as social demography, social networks, and community engagement (Saja et al., 2018)

and have not given importance to social resilience dimensions such as social trust,

social competence, social equity and diversity/inclusiveness, and social beliefs. None

of the existing social resilience frameworks have succeeded in integrating all key

indicators in a generic and adaptable structure for social resilience measurement to-

date, makes the ‘5S’ framework innovative, which is adaptable in any context. These

findings extend the research of Cutter (2016), Sharifi (2016), and Kwok et al. (2016),

and Khalili et al. (2015) on the need for an inclusive and adaptable framework, by

identifying key social resilience indicators and structuring it in a ‘5S’ framework,

which will assist in selecting social resilience indicators to apply surrogate approach

in any context.

A.2: Decision for surrogate approach/define objective:

Once the key social resilience indicators are identified, a decision for surrogate

approach needs to be made, if the identified social resilience indicators are not easily

measurable. There has been a tendency to prioritise easily measurable outcome-

oriented resilience characteristics. The existing approaches used mostly from the

publicly available census data are useful as a tool for initial resource allocation and

investment decisions at high level (Jülich, 2017) (See Appendix A for existing

measures using census data for five selected social resilience indicators). However,


they are often contested for their practical applicability at the community level for

effective resilience building interventions, as they often fail to capture process-

oriented and dynamic resilience features. All the indicators except six indicators in the

two social resilience characteristics (social demography and household structure) of

the ‘5S’ framework require surrogate approach for their assessment.

Since there are no previous studies on the surrogate approach in disaster

management, a set of surrogate decision criteria were drawn from the surrogate

approach literature in environmental science. The most important decision factor

which led to the adoption of a surrogate approach in ecology is the difficulty to directly

assess ecological systems (Lindenmayer et al., 2015b). However, this study found that

a set of surrogate decision criteria are needed to select social resilience indicators to

develop a surrogate approach. Hence, three key surrogate decision criteria used in this

study to select the target indicator for applying the surrogate approach are: (1) The

indicators should be process-oriented resilience indicators that are not a very static

indicator (dynamic, which frequently changes over time) (Cai et al., 2018; Cutter,

2016). (2) Indicators are complex for conceptualisation and cannot be easily measured

quantitatively (Sharifi & Yamagata, 2016). (3) The existing measures using indirect

methods do not provide an adequate measure of the target indicator (for example, the

existing measures mainly used data from publicly available census data sources)

(Jülich, 2017). These findings extend Lindenmayer et al. (2015b) research on the use

of key factors to decide on a surrogate approach in environmental science, by

formulating three surrogate decision criteria for systematically selecting resilience

indicators for developing a surrogate approach in a disaster management context.

For the purpose of this research to test the surrogate approach, five social

resilience indicators were selected from each of the five dimensions of the ‘5S’ social

resilience framework. This was done through review of literature on existing

measurements, based on the above key selection criteria for applying the surrogate

approach. In addition to the above three criteria to apply surrogate approach, this study

also considered indicators that are relevant to multiple disasters and different

geographical and socio-economic contexts, to make the findings widely applicable in

different contexts.


KEY FINDINGS IN RELATION TO RESEARCH OBJECTIVE TWO

(RO2)

RO2: To identify potential surrogates to measure key social resilience


In order to achieve RO2 of this study, the key step B of the conceptual surrogate

development framework (See Figure 3.6) was implemented to identify potential

surrogates to measure the five key social resilience indicators. The key outcomes of

this process as shown in Figure 7.3, include: six potential surrogates for each of the

five social resilience indicators; a set of measurement protocols for each of the

potential surrogates; and finally, the selection of surrogates identified in at least three

of the four case studies for further evaluation against surrogate evaluation criteria in

Phase II.

B.1. Explore all potential surrogates

This study found six potential surrogates for each of the five social resilience

indicators selected for testing the surrogate approach, as listed in Table 5.6. However,

three of the potential surrogates identified for each of the five social resilience

indicators have high validity, since they were found in at least three of the four case

studies in this research. Each indicator is discussed below in relation to their

contribution to the resilience assessment literature.

7.3. Key step B and related outcomes to achieve RO2


Indicator #1: Assessing social mobility and access to transport facilities:

The six potential surrogates and corresponding measurement protocol for

indicator #1, are summarised in Section 5.1.5. Three potential surrogates that were

found in multiple case studies include: S11: Available transport facilities targeting

People with Specific Needs (PwSN), S12: Evacuation places and centres, and S13:

Awareness programs/Early warning systems as shown in Table 5.6.

In the literature, social mobility and access to transport is most commonly

measured by vehicle ownership and available transport means (Burton, 2015; Kotzee

& Reyers, 2016) using census data (Kusumastuti et al., 2014; Mayunga, 2007; Qasim

et al., 2016b). This study highlighted that the use of a vehicle by every household ends

up in chaotic traffic congestion, given the narrow roads and bridges. Hence, the priority

for transport assistance should be given to people with specific needs (Peacock et al.,

2010). This study also identified the measure of vehicle availability and accessibility

for the transportation of people with specific needs (S11), as a potential surrogate to

assess social mobility. This finding extends the work of Peacock et al. (2010) on

assessing transport assistance for people with specific needs, by proposing a new

surrogate - vehicle availability and accessibility for transportation of people with

specific needs (S11), to assess social mobility in times of disasters, which has not been

used in a disaster context.

This study findings also highlighted that at times, people in highly congested

urban areas need to evacuate by foot directly to the nearest evacuation centre. When

the evacuation centre is located far away, people may need to go to the nearest

highway, where there is access to mass transportation. For example, Guadagno (2016)

emphasised that the identification of evacuation centres and awareness programs for

effective evacuation are important measures of resilience to disasters. These findings

extends the research by Guadagno (2016) on the identification of evacuation centres

and awareness programs for effective evacuation as measures of social mobility by the

use of two potential surrogates - evacuation centres/places with accessibility for PwSN

(S12) and awareness programs to effectively respond to early warning messages (S13),

for the assessment of social mobility.


Indicator #2: Assessing social trust in a disaster context:

This study found six potential surrogates and corresponding measurement

protocol for indicator #2, as summarised in Section 5.2.5. Three potential surrogates

that were found in multiple case studies include: S21: Effectiveness of community

based organisations/social service; S22: Level of services and resources of local

authorities/state support for people and S23: Functioning and effectiveness of disaster

management systems and complaint mechanisms as shown in Table 5.6.

In the literature, social trust was mostly assessed through direct methods using

community surveys and observations (Uslaner, 2016) or through publicly available

census-based measures such as voter turn-out in elections, level of ethnic segregation,

and crime rate (Asadzadeh et al., 2017; Joerin et al., 2014; Sherrieb et al., 2010).

Although, the critical role played by CBOs to help communities to enhance their

resilience is well-established in the literature, Drennan and Morrissey (2019)

highlighted that how effectively CBOs perform disaster management activities need

to be assessed. This study found that the effectiveness and performance of CBOs and

their level of active social service (S21) is a potential surrogate measure of social trust

in a disaster context.

Further, the need for high-quality disaster resilience building programs to build

trust demand proper resource allocation and functioning of disaster management

systems by local authorities, since they play an important role in disaster management

(Henstra, 2010). This study also found two more potential surrogates - level of services

and resource allocation from local authorities in disaster management activities (S22)

and regular functioning of disaster management systems (S23) to assess social trust in

a disaster context. All three surrogates (S21- S23) found in this study can capture the

important processes related to social trust with different social entities as opposed to

census data. These findings contradict that of Asadzadeh et al. (2017), Joerin et al.

(2014), and Sherrieb et al. (2010), on effectively assessing social trust in a disaster

context through census-based measures, by providing new measures such as

effectiveness of community based organisations/social service (S21), level of services

and resources of local authorities/state support for people (S22), and functioning and

effectiveness of disaster management systems and complaint mechanisms (S23) for the

assessment of social trust in a disaster context.


Indicator #3: Assessing learnings from the past as social competence:



that were found in multiple case studies include: S31: Reaction to disaster early

warning, S32: Awareness and disaster knowledge level, and S33: New DRR programs

such as innovative construction methods as shown in Table 5.6.

Many frameworks fail to incorporate the learnings from past disaster experience

as an indicator due to the difficulty in measuring it through census based datasets.

Effective early warning systems have saved many fatalities in the past and every time

when there is a disaster, people learn new ways to react to early warning (Keating et

al., 2016). Further, the perceptions of emerging disaster risk and hazard severity has

been cited in the literature as a potential measure for community competence and

learnings from the past experience (Cutter, 2016; Leykin et al., 2016). This study

finding supports the research by Cutter (2016) and Leykin et al. (2016) on assessing

learnings from the past as a social competence through perceptions of disaster risk and

hazard severity, by the use of surrogate - awareness and disaster knowledge level (S32)

as a measure of learnings from the past disaster experience.

Further, there has been increasing implementation of innovative community

based disaster risk reduction (DRR) initiatives based on past disaster experience

(Izumi et al., 2019), which provide an important measure of learnings from past

disasters. However, the existent frameworks failed to include the measure of

innovative initiatives as a measure of learnings from the past disasters. These findings

extend the research of Izumi et al. (2019) on the identification of innovative

community DRR initiatives as potential measures for learnings from the past disasters

by two potential surrogates - the reaction to disaster early warning (S31) and

implementation of innovative Disaster Risk Reduction (DRR) programs (S33), for the

assessment of learnings from the past disasters as a community competence.

Indicator #4: Assessing involvement of people with specific needs as social equity:



that were found in multiple case studies include: S41: Social safety programs for PwSN,


S42: PwSN committees/groups or representation of PwSN in committees, and S43:

Organisations/projects for PwSN as shown in Table 5.6.

In a disaster context, the highest priority should always be given to People with

Specific Needs (PwSN) such as children, older people, pregnant and lactating mothers,

and people with disabilities, who are the most vulnerable segment of the population be

easily and severely affected by a disaster (Sphere, 2018). However, most of the

resilience frameworks operationalised in a disaster context used census based social

demographic measures such as percentages of elderly population and children,

population in productive age, race and ethnicity, female population, people with

disability and age dependency ratio (Sherrieb et al., 2010; Woolf et al., 2016; Yoon et

al., 2016), to assess involvement of PwSN.

The effective design and use of safety net programs to increase resilience to

disasters and to respond to post-disaster needs are increasingly emphasised in disaster

management (Pelham et al., 2011). Further, the inclusion of people with specific needs

in the decision making structure such as in disaster management committees or in

organisations dealing with disasters, play an important role in enhancing their

resilience (Twigg et al., 2018). The existent frameworks fail to include key surrogate

measures such as social safety programs for PwSN (S41), PwSN committees/groups or

representation of PwSN in committees (S42), and Organisations/projects for PwSN

(S43) as found in this study. These findings are contrary to research findings of Sherrieb

et al. (2010), Woolf et al. (2016), and Yoon et al. (2016) on assessing the involvement

of PwSN as social equity through measures such as percentage of elderly population,

productive age population, and female population, by providing more important

measures in a disaster context such as social safety programs for PwSN (S41), PwSN

committees/groups or representation of PwSN in committees (S42), and

Organisations/projects for PwSN (S43) for assessing the involvement of PwSN.

Indicator #5: Assessing cultural norms/behaviours as social belief:



that were found in multiple case studies include: S51: Faith-based

organisations/practices and activities, S52: Culture of women/Gender norms that hinder

disaster management activities, and S53: Involvement of faith-based organisations and


religious institutions in disaster management. These were found as potential surrogates

across many case study locations as shown in Table 5.6.

In some communities which are faith-oriented, social beliefs/norms are largely

influenced by religious beliefs and practices. Religious belief was measured through a

survey of households who ‘believe that disasters happen by god’s will and do nothing’

(Lovell & Le Masson, 2014). Some other frameworks have used census-based

measures (Cutter, 2016; Peacock et al., 2010; Sherrieb et al., 2010).

The gender norms and cultural traditions in certain cultures and faiths are crucial

factors to understand the diversity needs of resilience building in a disaster context

(Hazeleger, 2013). This study found that the cultural or religious barriers for women

to actively participate in social activities, more specifically the involvement of women

in disaster related projects as a key measure of resilience across all case study

locations. Similar measures can be considered as one of the priority surrogates in

similar contexts, such as in strong faith-oriented communities. Further, the utilisation

of resources, practices, and networks available with faith-based organisations are key

sources of building resilience to disasters (Alawiyah et al., 2011; Ostadtaghizadeh et

al., 2016), particularly in strong faith-oriented communities such as in this case study

locations. Hence, these findings extend the research by Peacock et al. (2010), Cutter

(2016) and Sherrieb et al. (2010) on assessing cultural norms using census-based

measures from public data bases, by three potential surrogates (S51, S52, and S53) from

locally available data sources, for assessing cultural practices as social beliefs in a

disaster context.

B.2. Establish the surrogacy relationship

In this study, the relationship of each potential surrogate with the target indicator

was established through qualitative interpretations using interview transcripts, as

detailed in Chapter 5. The surrogate relationship with the target indicator of measure

is established mostly through statistical inferences and measurement methods largely

using sampling approaches (Athey et al., 2016). For example, in a stormwater quality

study using surrogates, different quantitative techniques were used to establish the

surrogacy relationship (Singh et al., 2013; Ziyath et al., 2013), such as multivariate

data analysis (Settle et al., 2007) and single linear regression (Miguntanna et al., 2010).

Since, the social resilience indicators selected for surrogate development in a disaster

context were process-oriented and dynamic indictors, it is very difficult to determine


the strength and sensitivity of the surrogacy relationship quantitatively at the initial

stage of exploring potential surrogates. Even, in the ecological context, establishing

surrogacy relationship through understanding causal relationships between surrogates

and the target is difficult, since robust quantitative tests are needed (Lindenmayer &

Likens, 2011).

The interview participants explained, how the potential surrogate they proposed

are linked to the target and how it changes, through examples from their work in

disaster management. The understanding of the link between the surrogate and the

target indicator by disaster management stakeholders at the local level, who implement

disaster resilience projects is important to inform the literature (Keating & Hanger-

Kopp, 2019; Levine, 2014). This study found that the discussion with experts to

establish surrogacy relationship on the proposed potential surrogates was a robust

method at the initial stages of exploring potential surrogates for social resilience

assessment in disaster management. The methodology of this study extends the

research by Athey et al. (2016), Settle et al. (2007), and Miguntanna et al. (2010) on

establishing surrogacy relationship through quantitative methods. This is done in this

research by establishing surrogacy relationship for social resilience measures in a

disaster context using qualitative interpretations through stakeholder engagement and

practical examples locally.

B.3. Determine the protocols for surrogate measurement:

This study also identified a set of surrogate measurement protocols for all

potential surrogates as detailed in Chapter 5. In general, the surrogates in ecology were

measured by the population trends of species to monitor changes in biodiversity

(Gregory et al., 2007), and mapping of surrogate systems to understand surrogate

behaviour (Lindenmayer et al., 2015b). However, Cutter et al. (2010) highlighted that

determining standards and matrices for resilience assessment is a challenge. For

example, trend analysis in sustainability assessment (Singh et al., 2009), and

participatory mapping such as vulnerability and community resource maps in disaster

risk management projects (Cadag & Gaillard, 2012) have been used as assessment

tools in vulnerability and sustainability assessment.

This study found spatial and temporal types of measurement protocols proposed

for measuring potential surrogates using locally available and frequently updated

administrative data. These can be widely categorised as the following key types:


analysing trends of initiatives or projects such as social safety program subscribers

targeted for disaster related assistance, spatial mapping of accessibility and social

entities, such as access to evacuation centres by the most vulnerable social groups and

active CBOs spatial coverage; and coverage of community based disaster risk

reduction projects from the reports produced by the state departments or local

authorities. These findings extends the research by Singh et al. (2009) and Cadag and

Gaillard (2012) on spatial and temporal participatory mapping in sustainability and

vulnerability assessment by analysing trends for disaster related initiatives in the

community, mapping of accessibility of disaster related entities/social entities, and

mapping of the coverage of disaster related projects to measure the proposed

surrogates in social resilience assessment in disaster management.

Further, the qualitative measures found in this study included: the recording of

past disaster experiences, identification of positive or negative trends in the community

that hinder resilience, and changes or improvements happened for the target measures.

As opposed to the current tendency to quantify resilience for developing an index for

different geographic locations, the focus of resilience assessment should also be on

what interventions are needed that will help communities to build their resilience to

emerging disasters (Levine, 2014). The practitioner perspective of resilience

assessment demands more practically oriented methods to assess resilience (Keating

& Hanger-Kopp, 2019).

In the existent social resilience assessment frameworks, only a small number of

frameworks have used qualitative measures. This include: recording the improvements

and weaknesses in asset base, capacities, and external resources of community

interventions (Woolf et al., 2016); recording the ideas for resilience improvements and

provided case study examples for key resilience concepts such as social networks,

economy, and infrastructure (Hegney et al., 2008); and participatory tools such as

stakeholder analysis, vulnerability and capacity assessment, and community

conversations to assess social resilience (Pfefferbaum et al., 2013). This study found

recording of past disaster experiences and positive/negative trends in the community

that supports or hinder as key types of qualitative measurement protocols to assess

social resilience indicators such as the learnings from the past disaster experience as a

social competency and cultural norms and behaviours as social beliefs that support or

hinder resilience. Further, the changes and improvements for the target surrogate


measures need to be recorded qualitatively, so that the required resilience building

activities can be identified. These findings extend the research by Woolf et al. (2016),

Pfefferbaum et al. (2013), and Hegney et al. (2008) on qualitative resilience

assessment methods by recording past disaster experiences, identification of positive

or negative trends in the community that hinder resilience, and changes or

improvements of the target surrogate measures. The measurement protocols for the

surrogates that were ranked first in the next phase (RO3) and how they contribute to

the literature are discussed in Section 7.3.1.

KEY FINDINGS IN RELATION TO RESEARCH OBJECTIVE THREE

(RO3)

RO3: To evaluate and select optimum surrogates for application by ranking the

potential surrogates against surrogate evaluation criteria

In order to achieve RO3, the key step C of the surrogate development framework

was implemented to evaluate potential surrogates identified in the case study research

(phase I). The key outcome of this process as shown in Figure 7.4, is the selection of

first ranked surrogates for priority measurement, based on the ranking obtained from

the robust evaluation process against five surrogate evaluation criteria and analysis of

ranking by three cohorts of experts: practitioners, policy makers, and researchers.

C.1. Evaluation of potential surrogates against criteria

Three potential surrogates identified for each indicator that were found in at least

three of four case study locations (Table 5.7) were evaluated in this study. The

evaluation of surrogates implemented in this study is a robust process, since each of

the potential surrogates were evaluated against each of the surrogate evaluation criteria

using qualitative 1-5 Likert scale (From very good to very bad). In the literature,

resilience indicators are mostly evaluated using pair-wise comparisons of indicators,

without imposing set of evaluation criteria to evaluate each indicator (Alshehri et al.,

2015b; Orencio & Fujii, 2013) or to rank indicators across different case study areas

(Carone et al., 2018). Hence, the method of evaluation for each surrogate against each

criteria applied in this study extends the literature on evaluation methods used by

Alshehri et al. (2015b), Orencio and Fujii (2013), and Carone et al. (2018) in resilience

assessment by evaluating each surrogate against each surrogate evaluation criteria on

a qualitative scale.


C.2. Rank the potential surrogates

The experts’ evaluation of potential surrogates against five surrogate evaluation

criteria was analysed using a multi-criteria decision analysis tool to rank the potential

surrogates. The final outcome by implementing key step C of the surrogate

development framework is the selection of first ranked surrogates for application.

(Figure 6.12). This study further found that the overall ranking of three potential

surrogates were mostly aligned with the ranking of experts from the policy making

and practitioner cohorts, as discussed in Chapter 6 for each of the target indicators.

Hence, the application of priority surrogates, which ranked first have wide practical

applicability in the assessment of target social resilience indicators at the community

and sub-national levels.

Further, the application of PROMETHEE as a multi-criteria decision making

method to rank potential surrogates also extends Alshehri et al. (2015b), Carone et al.

(2018), and Orencio and Fujii (2013) research on applying AHP applied to rank

resilience indicators by the use of qualitative Likert scale for evaluation of surrogates

in PROMETHEE.

7.4. Key step C and related outcomes to achieve RO3


C.3. Select the optimum surrogates for application

The evaluation of surrogates in PROMETHEE resulted in a first ranked

surrogate (Figure 6.12) for priority assessment of each target social resilience

indicator. The five first ranked surrogates to assess social resilience are highly relevant

for practical application, since they were identified in at least three of four case study

locations in Phase I and ranked first in the evaluation in Phase II of the study. These

are listed below:

- I1-Social mobility: S*1 - Availability/accessibility of evacuation centres;

- I2-Social trust: S*2 - Effectiveness/performance of community-based

organisations;

- I3-Social competence: S*3 - Reaction to early warning as a learning from the

past;

- I4-Social equity for PwSN: S*4 - Effectiveness of social safety programs ; and

- I5-Social cultural beliefs: S*5 - Gender norms and culture of women.

The relevance and measurement protocols of the five first ranked surrogates to

advance the resilience assessment knowledge using surrogate approach are discussed

below.

7.3.1 The relevance and measurement protocols of first ranked surrogates for

assessing social resilience indicators

I1-Social mobility: S*1 - Availability/accessibility of evacuation centres

Two most important interconnected components of a disaster preparedness and

response system include evacuation centres and early warning systems (Sorensen &

Sorensen, 2007). A set of measurement protocols identified in the case study locations

(Phase I of this study) for surrogate - S*1 are provided in Section 5.1.5.

The availability of and access to evacuation centres are key decision making

factors for peoples’ mobility in a disaster situation (Bañgate et al., 2017), because the

degree of availability of evacuation centres and the level of awareness largely

influence the decision for evacuation and mobility in times of disasters. However, this

study found that the accessibility of the evacuation centres for PwSN as the priority

surrogate to assess social mobility. In most instances, the focus of measure is given for

demarcating evacuation centres in community maps as part of the disaster

preparedness plans (Wright & Johnston, 2010), but there has been lack of auditing of


the evacuation centres to assess their suitability for use by People with Specific Needs

(PwSN). Hence, the finding extends the research by Wright and Johnston (2010) on

existing measures such as demarcating evacuation centres by the measure of

availability/accessibility of evacuation centres with facilities for PwSN (S*1), since the

mobility assistance in a disaster context should be prioritised for PwSN.

This study highlighted that some parts of the population may be disconnected

from the evacuation centres in a disaster situation, which makes the identification of

alternative routes an important social mobility measurement protocol. The

identification of alternative routes was done in evacuation transportation planning

(Campos et al., 2012). This study found that the identification of alternative evacuation

routes should be one of the key measurement protocols for assessing the accessibility

of evacuation centres. In the existent resilience frameworks, alternative evacuation

routes were assessed through arterial roads/km2 and distance to the nearest highway

using census data (Kotzee & Reyers, 2016; Parsons et al., 2016). These findings

contradicts the research by Kotzee and Reyers (2016) and Parsons et al. (2016) on the

existing social mobility measures such as arterial roads/km2 and distance to the nearest

highway by the identification of alternative routes to evacuation centres as a

measurement protocol for the surrogate - availability/accessibility of evacuation

centres (S*1).

I2-Social trust: S*2 - Effectiveness/performance of community-based

organisations

Community Based Organisations (CBOs) are one of the key social systems that

play an important role in enhancing resilience to disasters, more importantly in the

disaster preparedness phase. The involvement of CBOs in resilience building projects

provide an important measure of trust built in the community (Gin et al., 2017). A set

of measurement protocols identified in case study locations (Phase I of this study) for

surrogate - S*2 are provided in Section 5.2.5.

The increase in the involvement of CBOs in local disaster management work

improves the effectiveness of disaster response and recovery activities (Drennan &

Morrissey, 2019). This study highlighted that CBOs were faster to respond to address

immediate relief and recovery needs in disaster situations, as compared to the state

mechanisms which take time to respond due to bureaucratic hierarchy of decision

making. Hence, this study identified the importance of the disaster management


activities by CBOs and their ability to mobilise resources to assess social trust in a

disaster context.

The assessment of social trust is more than just numbers of registered CBOs in

the community, as it should capture how those CBOs function in a community to build

trust and the extent to which they engage with the community in a disaster context

(Cutter, 2016). However, research by Cutter (2016) did not propose measurement

protocols for assessing the engagement of CBOs in disaster management work. Yoon

et al. (2016) measured social resilience using census data such as the number of

registered CBOs, their volunteer base, and past budget. The study findings extends the

research Yoon et al. (2016) and Cutter (2016) on assessing CBOs’ performance using

census-based measures by the new surrogate measures such as mapping of actively

functioning CBOs in the selected region, their past experiences in disaster management

work, resource allocation for disaster management activities, and the provision for

disaster management work in CBO constitutions.

I3-Social competence: S*3 - Reaction to early warning as a learning from the past

The reaction to disaster early warning provides an important and priority

measure of competencies of a community as a learning from past disaster experience.

Early warnings in disaster situations mostly fail due to the lack of effective social

processes in which early warning messages are poorly communicated (Kelman &

Glantz, 2014). A set of measurement protocols identified in case study locations (Phase

I of this study) for the surrogate - S*3 are provided in Section 5.3.5.

There were many instances in the past when people reacted to disaster early

warnings ineffectively and the real learning to face future disasters happens by false

reactions. In the case study areas, people have learnt lessons over the past years and

improved their understanding of early warning messages from their inappropriate

reactions to early warning. The improvements or lapses as to how the community

reacts to the disaster early warning messages can be measured to understand whether

the community has learned lessons from the past disaster experiences. Sharma and Patt

(2012) identified three different key elements related to early warning response: the

severity of the impact of past disaster experience; past experience with false early

warning alarms; and past evacuation experience including the quality of evacuation

centres. However, the effectiveness of early warning is mostly assessed by the


availability of public awareness programmes/disaster drills in the existing resilience

frameworks (Joerin et al., 2014; Lorenz, 2013).

This study found that the assessment of reactions in past disaster early warning

instances and during drills provide a priority measure of communities’ ability to learn

lessons from the past disasters. When people participate in disaster drills and how they

react to disaster early warning messages during drills can be measured by the

improvements in early warning/evacuation drills and the trends in the participation of

people from most vulnerable areas. Further, the past real early warning instances and

the reaction of the public by the trend in number of people who reacted/evacuated

appropriately and trend in confirmation calls to disaster management centre are also

proposed as measurement protocol to assess reaction to early warning. These findings

extend Joerin et al. (2014) and Lorenz (2013) measures highlighted above on the

availability of public awareness programs/disaster drills. The new measures found in

this study include: the identification of early warning systems built after a disaster or

the improvements in existing early warning systems, the past real early warning

instances and the reaction of the public by number of people reacted appropriately, the

improvements in early warning/evacuation drills and trends in peoples’ participation,

trends in number of confirmation calls to disaster management centre after early

warning messages, and the number of people evacuated with and without official early

warning messages from local authorities.

I4-Social equity for PwSN: S*4 - Effectiveness of social safety programs

Adaptive social safety schemes such as targeting transfer of essential resources

to prepare for disasters provide not only short-term benefits, but largely contribute to

strengthen the resilience of the most vulnerable segment of the population (Béné et al.,

2012; Wu & Drolet, 2016). A set of measurement protocols identified in case study

locations (Phase I of this study) for surrogate - S*4 are provided in Section 5.4.5.

Social safety programs are also key processes for enhancing social support to the

most needy to promote self-recovery after disasters (IFRC, 2016). This study also

emphasised that the social safety programs ensure the right to priority assistance and

access to necessary resources for people who need special and utmost care in a society,

in times of a disaster. In disaster situations, social and cultural networks also serve as

social safety net mechanisms for people affected by disasters through the provision of

emergency food, shelter, and health assistance (USIOTWSP, 2007). In the existing


social resilience frameworks, the diversity of available social assistance programs

targeting various groups was used as a measure for social support systems (Burton,

2015).

Although social safety programs are still centred on poverty reduction and social

wellbeing strategies, there has been an increasing importance given to social safety as

an integrated set of programs to address poverty, vulnerability, risk, and inequality

(Fiszbein et al., 2013). This study highlighted that the integration of disaster

preparedness and risk reduction assistance within longer-term safety programs is being

increasingly introduced, particularly for People with Specific Needs (PwSN). These

findings extend the research by Burton (2015) on assessing social support systems

through the measurement of availability of diverse social assistance programs targeting

various groups. The new measures to assess social support systems found in this study

include: the measure of social safety programs effectiveness through trend in the

inclusion of most vulnerable people in social safety program and the degree of disaster

preparedness integration within the social safety programs to assess the involvement

of people with specific needs as a social equity in a disaster context.

I5-Social cultural beliefs: S*5 - Gender norms and culture of women

The gender norms are important to understand local or cultural beliefs that may

hinder building resilience to disasters among other barriers such as race and social

class (Kapucu et al., 2013). For example, this study highlighted the mobility limitations

in public for women headed households due to cultural barriers in some communities,

which can reduce their resilience to disasters. A set of measurement protocols

identified in case study locations (Phase I of this study) for surrogate - S*5 are provided

in Section 5.5.5.

The disaster management strategy and programs should ensure gender equality

and cultural diversity, which is also an important measure of community inclusiveness.

The role of women in disaster related work is neglected due to cultural norms in some

communities (Lovell & Le Masson, 2014). The lack of women trained on critical

disaster management skills such as first aid was highlighted in this study as an

important concern in communities with certain cultural or faith practices that has a

very restricted role for women in public places. Many of the existing social resilience

frameworks limit the measure of gender to socio-demographic indicators such as the

percentage of female population (Parsons et al., 2016; Yoon et al., 2016) and Kotzee


and Reyers (2016) used percentage of female labour as a measure of employment

equity in a disaster context.

The interaction between culture, social networks, and personal attributes of

women play a key role in determining social resilience in communities that prioritize

the cultural and religious norms and practices (Cottrell, 2006). For example, women,

who play a complementary role rather than an independent role in some contexts, are

one of the social segments that are highly exposed to disasters due to many gender

specific vulnerabilities such as cultural restrictions on mobility and decision making

powers (Alam & Rahman, 2014). Further, this study found that the absence of women

volunteer networks/organisations working in disaster management projects and the

lack of women population employed in critical jobs such as police and emergency

services are key challenges in disaster management work. These findings contradict

the research by Kotzee and Reyers (2016), Yoon et al. (2016), and Parsons et al. (2016)

on assessing gender related indicators through socio-demographic indicators such as

the percentage of female population and percentage of female labour, by measuring

gender norms in a disaster context through more specific social and cultural processes

that support or hinder resilience. The new surrogate measures proposed in this study

include, the number of women volunteers trained on first aid, swimming and other

critical disaster management skills, extent of women volunteer networks/organisations

working in disaster management projects, and the percentage of the female population

employed against the female population employed in critical jobs such as police and

emergency health services.

REVISIONS TO THE CONCEPTUAL SURROGATE DEVELOPMENT

FRAMEWORK

The revisions to the conceptual surrogate development framework (See Figure

3.6) are presented in this section, based on the findings and learnings from testing of

three key steps of conceptual surrogate development framework as explained in the

preceding sections. The integrated surrogate development framework (Figure 7.5) is

the main contribution to knowledge from this study, since it is a refined version based

on a robust testing of the conceptual framework developed through the review of

literature. The development of this framework answers the key research question of

this study, “How can key social resilience indicators in a disaster context be measured

using surrogate approach?”


Figure 7.5. An integrated (revised) surrogate development framework for resilience assessment in disaster management


7.4.1 An integrated surrogate development framework in disaster management

for future application

All sub-steps and key elements remain largely the same with some minor

changes in the integrated framework (Figure 7.5), compared to the conceptual

framework (Figure 3.6). The logical order of one sub-step and some key elements

within sub-steps are revised in the integrated framework, as discussed below.

(1) Key step A: Selection of social resilience indicators for surrogate

development:

There are no changes in the sub-steps of ‘key step A’. However, the key elements

within both sub-steps are included in the integrated surrogate development framework.

A.1. Identify key social resilience indicators:

Select the context: The context of social resilience measurement in a disaster context

needs to be decided as boundary conditions within which the social resilience

measurement becomes valid. In the tested framework, setting the context was part of

key step B in this research, since the data collection was done after selecting the key

indicators from the literature review. However, the context needs to be decided if the

surrogates are explored in a particular context, to make them applicable in the same

context.

The boundary conditions include: geography, disaster type, and target group. Although

some resilience indicators are more generic, many of the social resilience measures are

context-specific phenomena (Mitchell & Harris, 2012),. Hence, setting the context can

be two types: a) if there is a need for a specific measure to be applied in the local

context, the boundary can be set as the type of disaster exposure in the region, and

specific target group; or b) if the surrogate approach is applied to generic indicators,

multi-hazard context and common target group can be used, which can have a wide

applicability.

Select key indicators relevant to the context to develop surrogates: The critical

review of social resilience literature identified many process-oriented indicators,

which were often neglected in the existing resilience assessment frameworks. They

were structured in five social resilience dimensions in a ‘5S’ social resilience

framework (Saja et al., 2018). This framework or any other social resilience

framework that can guide to methodically identify key social resilience indicators most

relevant in the context of measurement to apply the surrogate approach can be used.


A.2. Decision for surrogate approach:

Once the social resilience indicators are selected, the decision for surrogate

approach needs to be made. The decision for the surrogate approach is based on three

surrogate decision criteria applied in this study:

1. Indicators should be process-oriented resilience indicators that are not static

(dynamic, which frequently change overtime).

2. Indicators are complex for conceptualisation and cannot be easily measured

quantitatively.

3. The existing measures using indirect methods do not provide an adequate measure

of the target indicator.

The inclusion of three surrogate decision criteria provides a uniform approach

for the users to make a methodical decision regarding the surrogate approach. This

could be either done through literature search or from reports or through consultation

with key disaster management experts in the selected study area.

(2) Key step B: Identification of potential surrogates for the selected key

indicators:

Key step B was revised from three to two sub-steps. The sub-step B.3 –

‘determining the protocols for surrogate measurement’ in the original framework is

moved to sub-component C.3 in the revised framework. Since the protocols for

surrogate measurement are only needed for the final selected surrogates, it was not

necessary to determine surrogate measurement protocols for all the potential

surrogates at the time of identification of potential surrogates.

There is no change in sub-step B.1. However, for sub-step B.2 in Figure 7.5, two

key elements in the tested conceptual framework were the strength and sensitivity of

the surrogacy relationship. At the initial stage of identifying potential surrogates for

measuring social resilience indicators, establishing surrogacy relationship (strength

and sensitivity) through qualitative explanation can be done, as tested in this study.

The qualitative interpretation of surrogacy relationship helps to elaborate the different

dynamics of its relationship, which is difficult to do using quantitative methods.

However, the evaluation of strength and sensitivity of the surrogacy relationship are

considered, when the potential surrogates are evaluated (Key step C) against the

criteria – ‘accuracy’ and ‘time-sensitivity’. The ‘accuracy’ criteria assesses the degree


of accuracy of the surrogate for the target measure, which considers how strongly the

surrogate is connected to the target. The ‘time-sensitivity’ criteria assesses the change

of surrogate over time (in disaster preparedness, response, and recovery phases).

The robustness of the strength and sensitivity of the surrogacy relationship can

be increased using quantitative approaches such as statistical and correlation analysis

methods (Ziyath et al., 2013). The surrogate relationship with the target indicator of

measure in environmental science is established mostly through statistical inferences

and measurement methods largely used in sampling approaches, as discussed in

Section 7.2 (Athey et al., 2016). Similar quantitative approaches to establish surrogacy

relationship can also be applied in resilience measurement in a disaster context.

However, in the context of social resilience assessment, quantitative findings also need

to be interpreted qualitatively, since qualitative explanation can increase the richness

of understanding the surrogacy relationship.

(3) Key step C: Selection of optimum surrogates for application in a disaster

context:

In this key step, the important addition is sub-step C.3 – protocols for surrogate

measurement, which was moved from the key step B as explained in the above section.

There is no change in sub-step C.1 from the tested framework for selecting optimum

surrogates. However, the sub-step C.3 in the tested framework is merged as part of

sub-step C.2 in the integrated (revised) framework. Since the selection of optimum

surrogates is an end result of ranking surrogates, it is more suitable to combine the

selection process with the sub-step - ranking of surrogates (C.3) (Figure 7.5).

Further to the application of the integrated surrogate development framework to

advance the surrogate approach research in disaster management, it can be applied by

practitioners at the local and sub-national levels with appropriate adaptations using

participatory methods. For example, the potential surrogates can be identified using

focus groups or participatory workshops with disaster management

stakeholders/committees. The evaluation of surrogates to select the final set of

surrogates for application can be done by participatory multi-criteria decision making

tools such as pair-wise comparisons. This is a one-off process and final set of

surrogates can be regularly updated to continuously monitor the progress of social

resilience indicators.

Chapter 8: Conclusions and recommendations for future research 252

Chapter 8: Conclusions and

recommendations for future research

This chapter contains four key sections: Section 8.1 presents the achievement of

three study objectives aligned with the three key steps of the surrogate development

framework (A to C). The next section (Section 8.2) outlines how this study addresses

the key research question. The following sections present the study contribution to

theory, policy and practice (Section 8.3) and study limitations (Section 8.4). The final

section provides key recommendations for future research (Section 8.5).

ACHIEVEMENT OF RESEARCH OBJECTIVES ONE TO THREE

The schematic diagram in Figure 8.1 shows the achievement of research

objectives in each key step of surrogate development framework (A to C), aligned with

each Research Objective (RO1 to RO3). The achievement of each research objective

as shown in Figure 8.1 is discussed below.

Research Objective one (RO1): To select key social resilience indicators that

require surrogate approach by developing an adaptive and inclusive social

resilience framework

The key step A of the surrogate development framework for selecting resilience

indicators to apply the surrogate approach was operationalised in the literature review

and research design phase of this study to achieve RO1. A critical review of social

resilience frameworks found that there was no systematic approach to identify key

indicators to apply the surrogate approach. This review resulted in a set of key social

resilience indicators, which were then structured in an inclusive and adaptive ‘5S’

social resilience framework (Figure 2.5) to guide the selection of social resilience

indicators to develop the surrogate approach. Further, three surrogate decision criteria

were found to select social resilience indicators to apply surrogate approach: the

indicators should be process-oriented resilience indicators that are not static indicators;

indicators are complex for conceptualisation and cannot be easily measured

quantitatively; and the existing measures using indirect methods do not provide an

adequate measure of the target indicator. Based on the surrogate decision criteria, five

indicators were selected from the ‘5S’ framework to develop the surrogate approach.


Figure 8.1. Overall schematic showing achievement of research objectives in each key step of surrogate

development to respond to the key RQ

RO – Research Objective RQ – Research Question


Research objective two (RO2): To identify potential surrogates to measure key

social resilience indicators in disaster management

The key step B of the surrogate development framework was operationalised to

identify potential surrogates (RO2) using a case study approach in phase I of this study.

The analysis of case study interview data from disaster management experts

(Practitioners and policy makers) revealed six potential surrogates for each of the

social resilience indicators (Table 5.6) and a set of measurement protocols for each

surrogate (Sections 5.1.5 - 5.5.5). The potential surrogates found in this study provide

a significant contribution to knowledge on resilience assessment (as explained in

Section 7.2), since many of the existing measures were based on census data which do

not provide an adequate assessment of resilience. However, some of the potential

surrogates were only identified in one or two of the four case study units. Among six

potential surrogates, three surrogates were found across multiple case studies for each

of the five social resilience indicators. Most of these surrogates and their measurement

protocols contribute as new measures to extend the existing resilience measures in the

literature. Further, some surrogates for indicators such as social trust and equity for

people with specific needs contradict existing measures proposed or applied in the

social resilience assessment literature in disaster management. Three potential

surrogates found in multiple case study units which have high validity were chosen for

further evaluation and ranking in phase II to select the optimum surrogate for

application.

Research objective three (RO3): To evaluate and select optimum surrogates for

application by ranking the potential surrogates against surrogate evaluation

criteria

The final key step C of the surrogate framework to select optimum surrogates

for application in a disaster context (RO3), was operationalised in phase II of this

research using an online survey with disaster management experts ranging from

research, policy and practice cohorts. The five first ranked surrogates (Figure 6.12)

were found as the most robust surrogates for application, as explained in Section 7.3,

include: availability/accessibility of evacuation centres to assess social mobility;

effectiveness/performance of community-based organisations to assess social trust;

reaction to early warning as a learning from the past to assess social competence;

effectiveness of social safety programs to assess social equity for PwSN; and gender

norms and culture of women to assess social cultural beliefs. The analysis also showed


that most of the first ranked surrogates were also the preference of highly experienced

experts from the cohort of practitioners and policy makers, and their ranking mostly

aligned with overall ranking of surrogates (Section 6.4).

The application of three key steps of surrogate development framework yielded

priority surrogates for application to assess social resilience to disasters, which is the

final outcome by synthesising the findings from three research objectives, achieved

through a robust multi-phase mixed method sequential research strategy. The findings

from this synthesis addressed the key knowledge gap on the need for a method to

identify and select priority surrogates using a set of measurement protocols for

assessing social resilience indicators (See Sections 7.2 and 7.3), which is a new

contribution to the resilience assessment knowledge in disaster management.

ADDRESSING THE KEY RESEARCH QUESTION AND AIM

The key Research Question (RQ) of this study is - ‘how can key social resilience

indicators in a disaster context be measured using surrogate approach?’ as described

in Section 1.5. To investigate the above key RQ, the main aim of this research was to

develop an approach for conceptualising, identifying, and evaluating surrogates to

assess social resilience indicators in a disaster context.

A conceptual surrogate development framework for assessing social resilience

in disaster management (See Figure 3.6) was created from the review of surrogate

approach literature in environmental science and tested through a multi-phase, mixed-

method research in this study. An integrated surrogate development framework that

was created from the lessons of testing the conceptual framework for future

application, is the main contribution to the knowledge in resilience assessment in

disaster management. The surrogate approach developed in this research can address

the challenges in resilience assessment by identifying a set of potential surrogates. The

proposed potential surrogates can be measured by accessing the most updated

administrative data at the community and sub-national levels and participatory

methods with key disaster management stakeholders as explained in Chapter 5. The

results from the assessment of social resilience indicators using the surrogate approach

will help to devise comprehensive action-oriented resilience decision making. Further,

the resilience measurement using surrogates will assist to regularly update and monitor


social resilience status, since they will use regularly updated administrative data

available at the local or sub-national levels.

The surrogate development framework provides a new method to assess

resilience indicators. This allows continuous monitoring of resilience and to devise

appropriate strategies and plans for resilience enhancement of communities to

emerging disasters. The creation of the surrogate development framework and testing

it through a rigourous research method in this study achieved the research aim, which

was to ‘develop an approach for conceptualising, identifying, and evaluating

surrogates to assess social resilience indicators in a disaster context’.

STUDY CONTRIBUTIONS TO KNOWLEDGE AND PRACTICE

8.3.1 Contribution to knowledge

This research makes three innovative contributions to knowledge with respect to

resilience assessment in a disaster context, by developing and operationalising the

surrogate development framework to assess social resilience.

Firstly, the integrated (revised) surrogate development framework (Figure 7.5)

to conceptualise, identify, and select surrogates to assess social resilience indicators

created in this study is innovative. It is the key contribution to resilience assessment

knowledge, since a framework to apply surrogate approach does not currently exist in

disaster management literature. This framework can be used in any context by disaster

management stakeholders to select surrogates for real world application. The

integrated surrogate development framework will address the knowledge gap in

resilience measurement methods in disaster management.

Secondly, a set of potential surrogates and their measurement protocols

(Sections 5.1.5 - 5.5.5) that were developed through an exploratory multiple case study

research at the local level has contributed to resilience assessment knowledge in

disaster management. These surrogates have wide applicability with sufficient

contextualisation in any urban context, since the selected social resilience indicators

such as social mobility, social trust, community competence, social equity, and social

cultural beliefs are important and common in many disaster and geographical contexts.

Thirdly, the first ranked surrogates were selected through a robust multi-phase

process of identification and evaluation of potential surrogates as part of a sequential

exploratory research design process in this study contribute to the priority resilience


assessment knowledge. The broader applicability of first ranked surrogates to the

disaster management literature was ensured by consulting disaster management

experts beyond Sri Lanka in the second phase of this research through an online survey.

8.3.2 Contribution to policy and practice

The findings from this research contributes to the disaster management policy

and practice in three ways.

Firstly, for policy makers and practitioners, the first ranked surrogates (Figure

6.12) can be a starting point for further consultation with disaster management

stakeholders at the local level for practical application to assess social resilience

indicators.

Secondly, a method to measure potential surrogates (measurement protocols)

proposed in this study has real-world application to undertake social resilience

assessment in disaster management. The practitioners and policy makers at the

community and sub-national levels can utilise the proposed surrogate measurement

protocols (Sections 5.1.5 - 5.5.5) to assess current status of resilience and plan

activities needed for resilience enhancement, based on the assessment results. The

surrogate approach will assist practitioners to assess resilience at the local level, since

they use locally available, frequently updated, and easily accessible administrative

data.

Thirdly, the overall contribution is the introduction of the surrogate approach in

disaster management with the aim to assess social resilience locally. The integrated

surrogate development framework was proposed by testing it through consultation

with practitioners and policy makers working in different levels from community to

national levels. Hence, it can be applied in any local context with appropriate

participatory tools by engaging key disaster management stakeholders to select the

required surrogates for application.


STUDY LIMITATIONS

The findings presented in this study can be applicable to any context, since all

the social resilience indicators selected are generic to different contexts. However, a

number of study limitations need to be considered.

1. This study was conducted in a specific socio-economic and cultural context in a

developing country. For example, the study area was largely dominated by Muslims

and Hindu population (minority populations in Sri Lanka), which have influenced

some findings such as in the case of social beliefs (religious and cultural beliefs).

Hence, the application of these findings in a different cultural setting require proper

contextualisation of socio-economic and cultural characteristics.

2. This study was conducted in the Eastern coastal region of Sri Lanka, which is prone

to floods, cyclone, and Tsunami. Although many findings are common and can be

applicable to any disaster type, the application of some of findings to other disaster

types need to consider specific characteristics of the disaster.


KEY RECOMMENDATIONS FOR FUTURE RESEARCH

The following key recommendations are proposed for future research studies in

assessing social resilience to disasters.

1. Advancing disaster resilience measurement research through the surrogate

approach in different contexts:

The surrogate approach in this study was applied in an urban context in a developing

country. Future research can test the surrogate framework to assess social resilience in

rural contexts, and in developed countries in order to develop a generic approach to

any context. Future research contexts also can include different socio-cultural settings

such as multi-cultural and marginalised communities and different disaster types such

as bush fires and volcanos.

2. Application of first ranked surrogates to assess social resilience:

Future research needs to apply first ranked surrogates found in this study and improve

it based on lessons from applying it in disaster management practice.

3. Application of surrogate measurement protocols:

Future research need to test the application of a set of surrogate measurement protocols

found in this study for each of the surrogates in order to measure the target resilience

indicator in different contexts to further advance the resilience assessment.

4. Developing a composite social resilience surrogate model:

A composite surrogate model is a conceptual mapping of surrogates with multiple

indicators and final outcomes. The potential surrogates identified in this study can be

applied in future research programs to develop composite surrogate models that will

better reflect the complex inter-linkages between surrogates.

5. Quantitative methods to establish surrogate and target indicator relationship and

surrogate measurement protocols

This research established the relationship of surrogates with target indicators and

surrogate measurement protocols qualitatively. Future research can test quantitative

methods to establish surrogate relationships and measurement protocols.

Bibliography 260

Bibliography

Abenayake, C., Yoshiki, M., Marasinghe, A., Takashi, Y., & Masahiro, I. (2016).

Applicability of extra-local methods for assessing community resilience to

disasters: A case of Sri Lanka. Journal of Environmental Assessment Policy

and Management, 18(02), 1650010.

Adger, W. N. (2000). Social and ecological resilience: are they related? PROGRESS

IN HUMAN GEOGRAPHY, 24(3), 347-364.

Adger, W. N. (2003). Social capital, collective action, and adaptation to climate

change. Economic geography, 79(4), 387-404.

Adger, W. N., Hughes, T. P., Folke, C., Carpenter, S. R., & Rockström, J. (2005).

Social-ecological resilience to coastal disasters. Science, 309(5737), 1036-

1039.

Ahern, J. (2011). From fail-safe to safe-to-fail: Sustainability and resilience in the

new urban world. Landscape and Urban Planning, 100(4), 341-343.

Ainuddin, S., & Routray, J. K. (2012). Earthquake hazards and community resilience

in Baluchistan. Natural Hazards, 63(2), 909-937. doi:10.1007/s11069-012-

0201-x

Ainuddin, S., Routray, J. K., & Ainuddin, S. (2015). Operational indicators for

assessing vulnerability and resilience in the context of natural hazards and

disasters. International Journal of Risk Assessment and Management, 18(1),

66-88. doi:10.1504/IJRAM.2015.068135

Alam, K., & Rahman, M. H. (2014). Women in natural disasters: a case study from

southern coastal region of Bangladesh. International Journal of Disaster Risk

Reduction, 8, 68-82.

Alasuutari, P., Bickman, L., & Brannen, J. (2008). The SAGE handbook of social

research methods: Sage.

Alasuutari, P., Bickman, L. & Brannen, J. (2008). Focus groups. In J. Smithson

(Ed.), The SAGE handbook of social research methods (pp. 357-370).

London: SAGE Publications Ltd.

Alawiyah, T., Bell, H., Pyles, L., & Runnels, R. C. (2011). Spirituality and faith-

based interventions: Pathways to disaster resilience for African American

Hurricane Katrina survivors. Journal of Religion & Spirituality in Social

Work: Social Thought, 30(3), 294-319.

Aldrich, D. P., & Meyer, M. A. (2014a). Social Capital and Community Resilience.

American Behavioral Scientist, 59(2), 254-269.

doi:10.1177/0002764214550299

Aldrich, D. P., & Meyer, M. A. (2014b). Social capital and community resilience.

American Behavioral Scientist, 0002764214550299.

Aldrich, D. P., & Meyer, M. A. (2015). Social Capital and Community Resilience.

American Behavioral Scientist, 59(2), 254-269.

doi:10.1177/0002764214550299

Aldunce, P., Beilin, R., Howden, M., & Handmer, J. (2015). Resilience for disaster

risk management in a changing climate: Practitioners’ frames and practices.

Global Environmental Change, 30, 1-11.

doi:http://dx.doi.org/10.1016/j.gloenvcha.2014.10.010

http://dx.doi.org/10.1016/j.gloenvcha.2014.10.010

Bibliography 261

Alexander, D. E. (2013). Resilience and disaster risk reduction: an etymological

journey. Natural Hazards and Earth System Sciences, 13(11), 2707-2716.

Alshehri, S. A., Rezgui, Y., & Li, H. (2015a). Delphi-based consensus study into a

framework of community resilience to disaster. Natural Hazards, 75(3),

2221-2245. doi:10.1007/s11069-014-1423-x

Alshehri, S. A., Rezgui, Y., & Li, H. (2015b). Disaster community resilience

assessment method: a consensus-based Delphi and AHP approach. Natural

Hazards, 78(1), 395-416. doi:10.1007/s11069-015-1719-5

Angus, D., Rintel, S., & Wiles, J. (2013). Making sense of big text: a visual-first

approach for analysing text data using Leximancer and Discursis.

International Journal of Social Research Methodology, 16(3), 261-267.

Arbon, P., Steenkamp, M., Cornell, V., Cusack, L., & Gebbie, K. (2016). Measuring

disaster resilience in communities and households. International Journal of

Disaster Resilience in the Built Environment, 7(2), 201-215.

doi:10.1108/ijdrbe-03-2015-0008

Aronson, J. K. (2005). Biomarkers and surrogate endpoints. British journal of

clinical pharmacology, 59(5), 491-494.

Asadabadi, M. R., Chang, E., & Saberi, M. (2019). Are MCDM Methods Useful? A

Critical Review of Analytic Hierarchy Process (AHP) and Analytic Network

Process (ANP). Cogent Engineering(just-accepted), 1623153.

Asadzadeh, A., Kötter, T., Salehi, P., & Birkmann, J. (2017). Operationalizing a

concept: The systematic review of composite indicator building for

measuring community disaster resilience. International Journal of Disaster

Risk Reduction.

Asher, M. J., Croke, B. F., Jakeman, A. J., & Peeters, L. J. (2015). A review of

surrogate models and their application to groundwater modeling. Water

Resources Research, 51(8), 5957-5973.

Athey, S., Chetty, R., Imbens, G., & Kang, H. (2016). Estimating treatment effects

using multiple surrogates: The role of the surrogate score and the surrogate

index. arXiv preprint arXiv:1603.09326.

Axinn, W. G., & Pearce, L. D. (2006). Mixed method data collection strategies:

Cambridge University Press.

Baker, S. G. (2005). A simple meta-analytic approach for using a binary surrogate

endpoint to predict the effect of intervention on true endpoint. Biostatistics,

7(1), 58-70. doi:10.1093/biostatistics/kxi040

Bañgate, J., Dugdale, J., Adam, C., & Beck, E. (2017). A review on the influence of

social attachment on human mobility during crises. Paper presented at the

T2-Analytical Modelling and Simulation Proceedings of the 14th ISCRAM

Conference, Albi, France.

Barton, P. S., Pierson, J. C., Westgate, M. J., Lane, P. W., & Lindenmayer, D. B.

(2015). Learning from clinical medicine to improve the use of surrogates in

ecology. Oikos, 124(4), 391-398.

Beccari, B. (2016). A Comparative Analysis of Disaster Risk, Vulnerability and

Resilience Composite Indicators. PLoS Currents, 8. Retrieved from

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4807925/

Becker, D., Schneiderbauer, S., Forrester, J. M., & Pedoth, L. (2015). Guidelines for

development of indicators, indicator systems and provide challenges: CRED,

Louvain.

Begum, D. (2008). Disaster Management in Bangladesh: Addressing women's needs.

BRAC University.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4807925/

Bibliography 262

Bell, E., Bryman, A., & Harley, B. (2018). Business research methods: Oxford

university press.

Béné, C., Wood, R. G., Newsham, A., & Davies, M. (2012). Resilience: New Utopia

or New Tyranny? Reflection about the Potentials and Limits of the Concept

of Resilience in Relation to Vulnerability Reduction Programmes. IDS

Working Papers, 2012(405), 1-61. doi:10.1111/j.2040-0209.2012.00405.x

Bennett, E., Cumming, G., & Peterson, G. (2005). A systems model approach to

determining resilience surrogates for case studies. Ecosystems, 8(8), 945-957.

doi:10.1007/s10021-005-0141-3

Berkes, F., & Seixas, C. S. (2005). Building resilience in lagoon social–ecological

systems: a local-level perspective. Ecosystems, 8(8), 967-974.

doi:10.1007/s10021-005-0140-4

Bion, W. R., Farr, R. M., Glaser, B. G., & Strauss, A. L. (2000). Qualitative

Researching with Text, Image and Sound. London: SAGE Publications Ltd.

Retrieved from http://methods.sagepub.com/book/qualitative-researching-

with-text-image-and-sound. doi:10.4135/9781849209731

Birkmann, J. (2006). Measuring vulnerability to natural hazards: towards disaster

resilient societies.

Birkmann, J. (2013). Data, indicators and criteria for measuring vulnerability:

Theoretical bases and requirements. In J. Birkmann (Ed.), Measuring

vulnerability to natural hazards: Towards disaster resilient societies (second

edition). United Nations University

Birnbaum, M. L., Daily, E. K., O'Rourke, A. P., & Loretti, A. (2016). Research and

Evaluations of the Health Aspects of Disasters, Part IX: Risk-Reduction

Framework. Prehospital and disaster medicine, 31(3), 309-325.

doi:10.1017/S1049023X16000352

Blau, P. M. (1977). A Macrosociological Theory of Social Structure. American

Journal of Sociology, 83(1), 26-54. Retrieved from

http://www.jstor.org.ezp01.library.qut.edu.au/stable/2777762

Boeije, H. (2009). Analysis in qualitative research: Sage publications.

Boin, A., Comfort, L. K., & Demchak, C. C. (2010). The rise of resilience.

Designing resilience: Preparing for extreme events, 1-12.

Brans, J.-P., & De Smet, Y. (2016). PROMETHEE Methods. In S. Greco, M.

Ehrgott, & J. R. Figueira (Eds.), Multiple Criteria Decision Analysis: State of

the Art Surveys (pp. 187-219). New York, NY: Springer New York.

Brinkhoff, P. (2011). Multi-criteria analysis for assessing sustainability of remedial

actions-applications in contaminated land development. Retrieved from

Brown, C., Shaker, R. R., & Das, R. (2018). A review of approaches for monitoring

and evaluation of urban climate resilience initiatives. Environment,

Development and Sustainability, 20(1), 23-40. doi:10.1007/s10668-016-9891-

7

Bruneau, M., Chang, S. E., Eguchi, R. T., Lee, G. C., O'Rourke, T. D., Reinhorn, A.

M., Shinozuka, M., Tierney, K., Wallace, W. A., & von Winterfeldt, D.

(2003). A framework to quantitatively assess and enhance the seismic

resilience of communities. Earthquake spectra, 19(4), 733-752.

doi:10.1193/1.1623497

Bryman, A. (2015). Social research methods: Oxford university press.

Burton, C. G. (2015). A Validation of Metrics for Community Resilience to Natural

Hazards and Disasters Using the Recovery from Hurricane Katrina as a Case

http://methods.sagepub.com/book/qualitative-researching-with-text-image-and-sound

http://methods.sagepub.com/book/qualitative-researching-with-text-image-and-sound

http://www.jstor.org.ezp01.library.qut.edu.au/stable/2777762

Bibliography 263

Study. Annals of the Association of American Geographers, 105(1), 67-86.

doi:10.1080/00045608.2014.960039

Cadag, J. R. D., & Gaillard, J.-C. (2012). Integrating knowledge and actions in

disaster risk reduction: the contribution of participatory mapping. Area,

44(1), 100-109.

Cai, H., Lam, N. S., Qiang, Y., Zou, L., Correll, R. M., & Mihunov, V. (2018). A

synthesis of disaster resilience measurement methods and indices.

International Journal of Disaster Risk Reduction, 31, 844-855.

Campos, V., Bandeira, R., & Bandeira, A. (2012). A method for evacuation route

planning in disaster situations. Procedia-Social and Behavioral Sciences, 54,

503-512.

Carone, M. T., Marincioni, F., & Romagnoli, F. (2018). Use of multi-criteria

decision analysis to define social resilience to disaster: the case of the EU

LIFE PRIMES project. Energy Procedia, 147, 166-174.

Carpenter, S. R., Westley, F., & Turner, M. G. (2005). Surrogates for resilience of

social–ecological systems. Ecosystems, 8(8), 941-944. doi:10.1007/s10021-

005-0170-y

Chandra, A., Williams, M., Plough, A., Stayton, A., Wells, K. B., Horta, M., & Tang,

J. (2013). Getting actionable about community resilience: The Los Angeles

county community disaster resilience project. American Journal of Public

Health, 103(7), 1181-1189. doi:10.2105/AJPH.2013.301270

Cinelli, M., Coles, S. R., & Kirwan, K. (2014). Analysis of the potentials of multi

criteria decision analysis methods to conduct sustainability assessment.

Ecological Indicators, 46, 138-148.

Cottrell, A. (2006). Weathering the strom: Women’s prepareness as a form of

resilience to weather-related hazards in Northern Australia Disaster

resilience: An integrated approach (pp. 128-142): Charles C. Thomas

Publisher Ltd.

Cox, R. S., & Hamlen, M. (2015). Community Disaster Resilience and the Rural

Resilience Index. American Behavioral Scientist, 59(2), 220-237.

doi:10.1177/0002764214550297

CRED. (2018). Economic Losses, Poverty & Disasters 1998-2017 Retrieved from

www.cred.be,. www.cred.be,

CRED. (2019a). 2018 REVIEW OF DISASTER EVENTS Retrieved from

https://www.cred.be/2018-review-disaster-events. from CRED and UNISDR

https://www.cred.be/2018-review-disaster-events

CRED. (2019b). Natural Disasters 2018. Retrieved from Brussels:

https://emdat.be/sites/default/files/adsr_2018.pdf

Creswell, J. W., & Clark, V. L. P. (2007). Designing and conducting mixed methods

research (2nd edition ed.): SAGE.

Creswell, J. W., & Creswell, J. D. (2017). Research design: Qualitative, quantitative,

and mixed methods approaches: Sage publications.

Cruzes, D. S., & Dyba, T. (2011). Recommended steps for thematic synthesis in

software engineering. Paper presented at the Empirical Software Engineering

and Measurement (ESEM), 2011 International Symposium on.

Cruzes, D. S., Dybå, T., Runeson, P., & Höst, M. (2015). Case studies synthesis: a

thematic, cross-case, and narrative synthesis worked example. Empirical

Software Engineering, 20(6), 1634-1665.

http://www.cred.be,/

http://www.cred.be/



https://emdat.be/sites/default/files/adsr_2018.pdf

Bibliography 264

Custance, J., & Hillier, H. (1998). Statistical issues in developing indicators of

sustainable development. Journal of the Royal Statistical Society: Series A

(Statistics in Society), 161(3), 281-290.

Cutter, S. L. (2016). The landscape of disaster resilience indicators in the USA.

Natural Hazards, 80(2), 741-758. doi:10.1007/s11069-015-1993-2

Cutter, S. L., Ash, K. D., & Emrich, C. T. (2014). The geographies of community

disaster resilience. Global Environmental Change, 29, 65-77.


Cutter, S. L., Barnes, L., Berry, M., Burton, C., Evans, E., Tate, E., & Webb, J.

(2008). A place-based model for understanding community resilience to

natural disasters. Global Environmental Change, 18(4), 598-606.


Cutter, S. L., Boruff, B. J., & Shirley, W. L. (2003). Social Vulnerability to

Environmental Hazards*. Social Science Quarterly, 84(2), 242-261.

doi:10.1111/1540-6237.8402002

Cutter, S. L., Burton, C. G., & Emrich, C. T. (2010). Disaster resilience indicators for

benchmarking baseline conditions. Journal of Homeland Security and

Emergency Management, 7(1), 51. doi:10.2202/1547-7355.1732

Davies, J. (2010). Preparation and process of qualitative interviews and focus groups.

Practical research and evaluation: A start-to-finish guide for practitioners,

126-144.

Davis, C. (2016). Focus groups: Applying communication theory through design,

facilitation, and analysis: Routledge.

De Ruyter, K. (1996). Focus versus nominal group interviews: a comparative

analysis. Marketing Intelligence & Planning, 14(6), 44-50.

DiCicco‐Bloom, B., & Crabtree, B. F. (2006). The qualitative research interview.

Medical education, 40(4), 314-321.

Djalante, R., & Thomalla, F. (2011). Community Resilience to Natural Hazards and

Climate Change: A Review of Definitions and Operational Frameworks.

Asian Journal of Environment and Disaster Management, 3(3).

doi:10.3850/S1793924011000952

Donnelly, A., Jones, M., O'Mahony, T., & Byrne, G. (2007). Selecting

environmental indicator for use in strategic environmental assessment.

Environmental Impact Assessment Review, 27(2), 161-175.

Doorn, N. (2017). Resilience indicators: opportunities for including distributive

justice concerns in disaster management. Journal of risk research, 20(6), 711-

731.

Drennan, L., & Morrissey, L. (2019). Resilience policy in practice—Surveying the

role of community based organisations in local disaster management. Local

Government Studies, 45(3), 328-349.

EC. (2015). Indicators for sustainable cities. In-depth Report 12. . Retrieved from

Produced for the European Commission DG Environment by the Science

Communication Unit, UWE, Bristol. : Available at:

http://ec.europa.eu/science-environment-policy

Eckstein, D. (2018). Global climate risk index 2018. A briefing paper. Retrieved

from

https://germanwatch.org/sites/germanwatch.org/files/Global%20Climate%20

Risk%20Index%202019_2.pdf

Eiser, J. R., Bostrom, A., Burton, I., Johnston, D. M., McClure, J., Paton, D., Van

Der Pligt, J., & White, M. P. (2012). Risk interpretation and action: A



http://ec.europa.eu/science-environment-policy

https://germanwatch.org/sites/germanwatch.org/files/Global%20Climate%20Risk%20Index%202019_2.pdf

https://germanwatch.org/sites/germanwatch.org/files/Global%20Climate%20Risk%20Index%202019_2.pdf

Bibliography 265

conceptual framework for responses to natural hazards. International Journal

of Disaster Risk Reduction, 1, 5-16.

Enarson, E. (1998). Through women’s eyes: A gendered research agenda for disaster

social science. Disasters, 22(2), 157-173.

Endress, M. (2015). The social constructedness of resilience. Social Sciences, 4(3),

533-545.

Faifua, D. (2014). The key informant technique in qualitative research: SAGE

Publications, Ltd.

FAO. (2008). Food Security Information for Action , Vulnerability, Lesson 3 -

Vulnerability Indicators

Fernández, N. G. (2013). The management of missing values in PROMETHEE

methods.

Fiszbein, A., Kanbur, R., & Yemtsov, R. (2013). Social protection, poverty and the

post-2015 agenda: The World Bank.

Fleming, T. R., & Powers, J. H. (2012). Biomarkers and surrogate endpoints in

clinical trials. Statistics in medicine, 31(25), 2973-2984.

Flick, U. (2014). An introduction to qualitative research (5th ed.). London: Sage.

Fontana, A., & Prokos, A. H. (2016). The interview: From formal to postmodern:

Routledge.

Fothergill, A., & Peek, L. A. (2004). Poverty and disasters in the United States: A

review of recent sociological findings. Natural Hazards, 32(1), 89-110.

Frailing, K., & Harper, D. W. (2017). The Resilience of Communities Toward a

Criminology of Disaster: What We Know and What We Need to Find Out (pp.

167-193). New York: Palgrave Macmillan US.

Frazier, T. G., Thompson, C. M., Dezzani, R. J., & Butsick, D. (2013). Spatial and

temporal quantification of resilience at the community scale. Applied

Geography, 42, 95-107. doi:10.1016/j.apgeog.2013.05.004

Gin, J. L., Eisner, R. K., Der-Martirosian, C., Kranke, D., & Dobalian, A. (2017).

Preparedness is a marathon, not a sprint: A tiered maturity model for

assessing preparedness in homeless residential organizations in Los Angeles.

Natural hazards review, 19(1), 04017027.

Goepel, K. D. (2013). Implementing the analytic hierarchy process as a standard

method for multi-criteria decision making in corporate enterprises–a new

AHP excel template with multiple inputs. Paper presented at the Proceedings

of the international symposium on the analytic hierarchy process.

GoSL. (2018). Population and Population Density by D.S. Division - 2018

Retrieved from

http://www.statistics.gov.lk/statistical%20Hbook/2019/Ampara/2.4.1.pdf.

Retrieved 20 February 2020, from District Secretariate, Ampara, Sri Lanka

http://www.statistics.gov.lk/statistical%20Hbook/2019/Ampara/2.4.1.pdf

Grantham, H. S., Pressey, R. L., Wells, J. A., & Beattie, A. J. (2010). Effectiveness

of biodiversity surrogates for conservation planning: different measures of

effectiveness generate a kaleidoscope of variation. PLoS One, 5(7).

Grayson, R. B., Finlayson, B. L., Gippel, C. J., & Hart, B. T. (1996). The potential of

field turbidity measurements for the computation of total phosphorus and

suspended solids loads. Journal of Environmental Management, 47(3), 257-

267. doi:10.1006/jema.1996.0051

Greene, J. C., & Hall, J. N. (2010). Dialectics and pragmatism: Being of

consequence. Handbook of mixed methods in social and behavioral research,

119-144.



Bibliography 266

Gregory, R. D., Vorisek, P., Van Strien, A., GMELIG MEYLING, A. W., Jiguet, F.,

Fornasari, L., Reif, J., Chylarecki, P., & Burfield, I. J. (2007). Population

trends of widespread woodland birds in Europe. Ibis, 149, 78-97.

Guadagno, L. (2016). Human Mobility in the Sendai Framework for Disaster Risk

Reduction. International Journal of Disaster Risk Science, 7(1), 30-40.

doi:10.1007/s13753-016-0077-6

Gubrium, J., Holstein, J., Marvasti, A., & McKinney, K. (2012). The SAGE

Handbook of Interview Research: The Complexity of the Craft Retrieved

from http://methods.sagepub.com/book/handbook-of-interview-research-2e

doi:10.4135/9781452218403

Hancock, D. R., & Algozzine, B. (2016). Doing case study research: A practical

guide for beginning researchers: Teachers College Press.

Hazeleger, T. (2013). Gender and disaster recovery: Strategic issues and action in

Australia. Australian Journal of Emergency Management, The, 28(2), 40.

Hegney, D., Ross, H., & Baker, P. (2008). Building resilience in rural communities:

toolkit.

Henn, M., Weinstein, M., & Foard, N. (2009). A critical introduction to social

research (2nd Edition ed.): Sage Publications.

Henstra, D. (2010). Evaluating local government emergency management programs:

What framework should public managers adopt? Public Administration

Review, 70(2), 236-246.

Ho, D., Newell, G., & Walker, A. (2005). The importance of property-specific

attributes in assessing CBD office building quality. Journal of Property

Investment & Finance, 23(5), 424-444.

Hoffmann, R., & Muttarak, R. (2017). Learn from the Past, Prepare for the Future:

Impacts of Education and Experience on Disaster Preparedness in the

Philippines and Thailand. World Development, 96, 32-51.

doi:https://doi.org/10.1016/j.worlddev.2017.02.016

Holling, C. S. (1973). Resilience and stability of ecological systems. Annual Review

of Ecology and Systematics, 4(1), 1-23.

Hyde, K. M. (2006). Uncertainty analysis methods for multi-criteria decision

analysis.

IFRC. (2016). World Disasters Report 2016: Resilience: saving lives today, investing

for tomorrow. Retrieved from www.ifrc.org/wdr2016.

Ishizaka, A., & Nemery, P. (2013). A Multi-Criteria Group Decision Framework for

Partner Grouping When Sharing Facilities. Group Decision and Negotiation,

22(4), 773-799. doi:10.1007/s10726-012-9292-8

Izumi, T., Shaw, R., Djalante, R., Ishiwatari, M., & Komino, T. (2019). Disaster risk

reduction and innovations. Progress in Disaster Science, 2, 100033.

Jameel, M. I. M., Musthaffa, A.H.M. (2009). Disaster history of Sainthamaruthu

village History of Sainthamaruthu (pp. 449). Sainthamaruthu: Community

Welfare and Disaster Management Council, Grand Mosque, Sainthamaruthu-

Maligaikadu, Sri Lanka.

JICA. (2008). RECOVERY, REHABILITATION AND DEVELOPMENT PROJECT

FOR TSUNAMI AFFECTED AREA OF NORTHERN AND EASTERN

REGION - Chapter 6 Retrieved from

http://open_jicareport.jica.go.jp/pdf/11871076_01.pdf

Joerin, J., Shaw, R., Takeuchi, Y., & Krishnamurthy, R. (2014). The adoption of a

climate disaster resilience index in Chennai, India. Disasters, 38(3), 540-561.

http://methods.sagepub.com/book/handbook-of-interview-research-2e

https://doi.org/10.1016/j.worlddev.2017.02.016

http://www.ifrc.org/wdr2016

http://open_jicareport.jica.go.jp/pdf/11871076_01.pdf

Bibliography 267

Johnson, J. M., & Rowlands, T. (2012). The SAGE Handbook of Interview

Research: The Complexity of the Craft (2 ed.). Thousand Oaks, California:

SAGE Publications, Inc. Retrieved from

http://methods.sagepub.com/book/handbook-of-interview-research-2e.

doi:10.4135/9781452218403

Jones, L., & Tanner, T. (2017). ‘Subjective resilience’: using perceptions to quantify

household resilience to climate extremes and disasters. Regional

Environmental Change, 17(1), 229-243. doi:10.1007/s10113-016-0995-2

Jülich, S. (2017). Towards a Local-Level Resilience Composite Index: Introducing

Different Degrees of Indicator Quantification. International Journal of

Disaster Risk Science, 8(1), 91-99. doi:10.1007/s13753-017-0114-0

Kafle, S. K. (2012). Measuring disaster-resilient communities: a case study of coastal

communities in Indonesia. Journal of business continuity & emergency

planning, 5(4), 316-326. Retrieved from

http://www.ingentaconnect.com/content/hsp/jbcep/2012/00000005/00000004

/art00004

Kapucu, N., Hawkins, C. V., & Rivera, F. I. (2013). Disaster Preparedness and

Resilience for Rural Communities. Risk, Hazards & Crisis in Public Policy,

4(4), 215-233. doi:10.1002/rhc3.12043

Keating, A., Campbell, K., Szoenyi, M., McQuistan, C., Nash, D., & Burer, M.

(2017). Development and testing of a community flood resilience

measurement tool. Natural Hazards and Earth System Sciences, 17(1), 77.

Keating, A., & Hanger-Kopp, S. (2019). Practitioner perspectives of disaster

resilience in international development. International Journal of Disaster

Risk Reduction, 101355. doi:https://doi.org/10.1016/j.ijdrr.2019.101355

Keating, A., Venkateswaran, K., Szoenyi, M., MacClune, K., & Mechler, R. (2016).

From event analysis to global lessons: disaster forensics for building

resilience. Natural Hazards and Earth System Sciences, 16(7), 1603-1616.

Keck, M., & Sakdapolrak, P. (2013). What is social resilience? Lessons learned and

ways forward. Erdkunde, 5-19.

Kelman, I., & Glantz, M. H. (2014). Early Warning Systems Defined. In A. Singh &

Z. Zommers (Eds.), Reducing Disaster: Early Warning Systems For Climate

Change (pp. 89-108). Dordrecht: Springer Netherlands.

Khalili, S., Harre, M., & Morley, P. (2015). A temporal framework of social

resilience indicators of communities to flood, case studies: Wagga wagga and

Kempsey, NSW, Australia. International Journal of Disaster Risk Reduction,

13, 248-254. doi:10.1016/j.ijdrr.2015.06.009

Kilic, H. S., Zaim, S., & Delen, D. (2015). Selecting “The Best” ERP system for

SMEs using a combination of ANP and PROMETHEE methods. Expert

Systems with Applications, 42(5), 2343-2352.

doi:https://doi.org/10.1016/j.eswa.2014.10.034

Kimhi, S., & Shamai, M. (2004). Community resilience and the impact of stress:

Adult response to Israel's withdrawal from Lebanon. Journal of Community

Psychology, 32(4), 439-451.

Klein, R. J., Nicholls, R. J., & Thomalla, F. (2003). Resilience to natural hazards:

How useful is this concept? Global Environmental Change Part B:

Environmental Hazards, 5(1-2), 35-45.

Kotzee, I., & Reyers, B. (2016). Piloting a social-ecological index for measuring

flood resilience: A composite index approach. Ecological Indicators, 60, 45-

53.

http://methods.sagepub.com/book/handbook-of-interview-research-2e

http://www.ingentaconnect.com/content/hsp/jbcep/2012/00000005/00000004/art00004

http://www.ingentaconnect.com/content/hsp/jbcep/2012/00000005/00000004/art00004

https://doi.org/10.1016/j.ijdrr.2019.101355

https://doi.org/10.1016/j.eswa.2014.10.034

Bibliography 268

Kulig, J. C., Edge, D. S., Townshend, I., Lightfoot, N., & Reimer, W. (2013).

COMMUNITY RESILIENCY: EMERGING THEORETICAL INSIGHTS.

Journal of Community Psychology, 41(6), 758-775. doi:10.1002/jcop.21569

Kusumastuti, R. D., Viverita, Husodo, Z. A., Suardi, L., & Danarsari, D. N. (2014).

Developing a resilience index towards natural disasters in Indonesia.

International Journal of Disaster Risk Reduction, 10, Part A, 327-340.

doi:http://dx.doi.org/10.1016/j.ijdrr.2014.10.007

Kwok, A. H., Doyle, E. E. H., Becker, J., Johnston, D., & Paton, D. (2016). What is

‘social resilience’? Perspectives of disaster researchers, emergency

management practitioners, and policymakers in New Zealand. International

Journal of Disaster Risk Reduction, 19, 197-211.


Lane, P., Barton, P., & Lindenmayer, D. (2015). Use of surrogates in medicine: ideas

that may be useful for ecology. Indicators and Surrogates of Biodiversity and

Environmental Change, 1, 103.

Leeuw, E. d. (2008). The SAGE Handbook of Social Research Methods. London:

SAGE Publications Ltd. Retrieved from

http://methods.sagepub.com/book/the-sage-handbook-of-social-research-

methods. doi:10.4135/9781446212165

Levine, S. (2014). Assessing resilience: why quantification misses the point.

Humanitarian Policy Group (ODI) Working Paper.

Leximancer. (2017). Leximancer User Guide Release 4.5.

Leykin, D., Lahad, M., Cohen, R., Goldberg, A., & Aharonson-Daniel, L. (2016).

The dynamics of community resilience between routine and emergency

situations. International Journal of Disaster Risk Reduction, 15, 125-131.


Lindenmayer, D., Barton, P., & Pierson, J. (2015a). Indicators and Surrogates of

Biodiversity and Environmental Change: CSIRO PUBLISHING.

Lindenmayer, D., Pierson, J., Barton, P., Beger, M., Branquinho, C., Calhoun, A.,

Caro, T., Greig, H., Gross, J., & Heino, J. (2015b). A new framework for

selecting environmental surrogates. Science of the total Environment, 538,

1029-1038. doi:http://dx.doi.org/10.1016/j.scitotenv.2015.08.056

Lindenmayer, D. B., & Likens, G. E. (2011). Direct measurement versus surrogate

indicator species for evaluating environmental change and biodiversity loss.

Ecosystems, 14(1), 47-59. doi:10.1007/s10021-010-9394-6

Lorenz, D. F. (2013). The diversity of resilience: contributions from a social science

perspective. Natural Hazards, 67(1), 7-24.

Lovell, E., & Le Masson, V. (2014). Equity and inclusion in disaster risk reduction:

building resilience for all: Overseas Development Institute.

Maclaren, V. W. (1996). Urban sustainability reporting. Journal of the American

Planning Association, 62(2), 184-202.

Maduz, L., & Roth, F. (2017). The Urbanization of Disaster Management. CSS

Analyses in Security Policy, 204.

Maguire, B., & Hagan, P. (2007). Disasters and communities: understanding social

resilience. Australian Journal of Emergency Management, The, 22(2), 16.

Majumder, M. (2015). Multi Criteria Decision Making Impact of Urbanization on

Water Shortage in Face of Climatic Aberrations (pp. 35-47). Singapore:

Springer Singapore.

http://dx.doi.org/10.1016/j.ijdrr.2014.10.007


http://methods.sagepub.com/book/the-sage-handbook-of-social-research-methods

http://methods.sagepub.com/book/the-sage-handbook-of-social-research-methods


http://dx.doi.org/10.1016/j.scitotenv.2015.08.056

Bibliography 269

Manyena, B. (2016). After Sendai: Is Africa Bouncing Back or Bouncing Forward

from Disasters? International Journal of Disaster Risk Science, 7(1), 41-53.

doi:10.1007/s13753-016-0084-7

Markelj, J., Kitek Kuzman, M., Grošelj, P., & Zbašnik-Senegačnik, M. (2014). A

simplified method for evaluating building sustainability in the early design

phase for architects. Sustainability, 6(12), 8775-8795.

Marshall, C., & Rossman, G. B. (2014). Designing qualitative research (5th ed.):

Sage publications.

Matyas, D., & Pelling, M. (2015). Positioning resilience for 2015: the role of

resistance, incremental adjustment and transformation in disaster risk

management policy. Disasters, 39(s1), s1-s18. doi:10.1111/disa.12107

Mayunga, J. S. (2007). Understanding and applying the concept of community

disaster resilience: a capital-based approach. Summer academy for social

vulnerability and resilience building, 1, 16.

McMillen, H., Campbell, L., Svendsen, E., & Reynolds, R. (2016). Recognizing

Stewardship Practices as Indicators of Social Resilience: In Living Memorials

and in a Community Garden. Sustainability, 8(8), 775. Retrieved from

http://www.mdpi.com/2071-1050/8/8/775

Mellin, C., Delean, S., Caley, J., Edgar, G., Meekan, M., Pitcher, R., Przeslawski,

R.,Williams, A., Bradshaw, C. (2011). Effectiveness of biological surrogates

for predicting patterns of marine biodiversity: a global meta-analysis. PLoS

One 6, e20141.

Miguntanna, N. S., Egodawatta, P., Kokot, S., & Goonetilleke, A. (2010).

Determination of a set of surrogate parameters to assess urban stormwater

quality. Science of the total Environment, 408(24), 6251-6259.

doi:10.1016/j.scitotenv.2010.09.015

Miller, M., Paton, D., & Johnston, D. (1999). Community vulnerability to volcanic

hazard consequences. Disaster Prevention and Management: An

International Journal, 8(4), 255-260.

Mills, J., Harrison, H., Franklin, R., & Birks, M. (2017). Case study research:

Foundations and methodological orientations. Paper presented at the Forum

Qualitative Sozialforschung/Forum: Qualitative Social Research.

Mitchell, T., & Harris, K. (2012). Resilience: A risk management approach. ODI

background note, 1-7.

Mitchell, T., Jones, L., Lovell, E., Comba, E. (2013). Disaster Risk Management in

Post-2015 Development Goals - Potential targets and indicators. Overseas

Development Institute, London.

Moher D, L. A., Tetzlaff J, Altman DG. (2009). Preferred Reporting Items for

Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS

Med, 6(7). doi:10.1371/journal.pmed1000097

Morgan, D. L. (2007). Paradigms Lost and Pragmatism Regained:Methodological

Implications of Combining Qualitative and Quantitative Methods. Journal of

Mixed Methods Research, 1(1), 48-76. doi:10.1177/2345678906292462

Nadel, S. F. (2013). The theory of social structure: Routledge.

Neuman, L. W. (2014). Social Research Methods: Qualitative and Quantitative

Approaches (7E ed.): Pearson Education India.

Norris, F. H., Stevens, S. P., Pfefferbaum, B., Wyche, K. F., & Pfefferbaum, R. L.

(2008). Community resilience as a metaphor, theory, set of capacities, and

strategy for disaster readiness. American journal of community psychology,

41(1-2), 127-150. doi:DOI 10.1007/s10464-007-9156-6

http://www.mdpi.com/2071-1050/8/8/775

Bibliography 270

O’Loughlin, L. S., Lindenmayer, D. B., Smith, M. D., Willig, M. R., Knapp, A. K.,

Cuddington, K., Hastings, A., Foster, C. N., Sato, C. F., & Westgate, M. J.

(2018a). Surrogates underpin ecological understanding and practice.

Bioscience, 68(9), 640-642.

O’Loughlin, L. S., Lindenmayer, D. B., Smith, M. D., Willig, M. R., Knapp, A. K.,

Cuddington, K., Hastings, A., Foster, C. N., Sato, C. F., & Westgate, M. J.

(2018b). Surrogates Underpin Ecological Understanding and Practice.

BioScience.

Oppenheim, A. N. (2000). Questionnaire design, interviewing and attitude

measurement: Bloomsbury Publishing.

Orencio, P. M., & Fujii, M. (2013). A localized disaster-resilience index to assess

coastal communities based on an analytic hierarchy process (AHP).



Ostadtaghizadeh, A., Ardalan, A., Paton, D., Jabbari, H., & Khankeh, H. R. (2015).

Community disaster resilience: a systematic review on assessment models

and tools. PLoS Currents, 7.

doi:10.1371/currents.dis.f224ef8efbdfcf1d508dd0de4d8210ed

Ostadtaghizadeh, A., Ardalan, A., Paton, D., Khankeh, H., & Jabbari, H. (2016).

Community disaster resilience: a qualitative study on Iranian concepts and

indicators. Natural Hazards, 83(3), 1843-1861. doi:10.1007/s11069-016-

2377-y

Park, B., & Youngchul Kim, R. (2014). Making a decision about importance analysis

and prioritization of use cases through comparison the analytic hierarchy

process (AHP) with use case point (UCP) technique. International Journal of

Software Engineering and Its Applications, 8(3), 89-96.

Parsons, M., Glavac, S., Hastings, P., Marshall, G., McGregor, J., McNeill, J.,

Morley, P., Reeve, I., & Stayner, R. (2016). Top-down assessment of disaster

resilience: A conceptual framework using coping and adaptive capacities.



Paton, D. (2007). Preparing for natural hazards: The role of community trust.

DISASTER PREVENTION AND MANAGEMENT, 16(3), 370-379.

doi:10.1108/09653560710758323

Paton, D., Millar, M., & Johnston, D. (2001). Community resilience to volcanic

hazard consequences. Natural Hazards, 24(2), 157-169.

Patton, M. Q. (2002). Qualitative evaluation and research methods (3rd ed.): SAGE

Publications, inc.

Peacock, W. G., Brody, S. D., Seitz, W. A., Merrell, W., Vedlitz, A., Zahran, S.,

Harriss, R., & Stickney, R. (2010). Advancing Resilience of Coastal

Localities: Developing, Implementing, and Sustaining the Use of Coastal

Resilience Indicators: A Final Report. Hazard Reduction and Recovery

Center.

Pelham, L., Clay, E., & Braunholz, T. (2011). Natural disasters: what is the role for

social safety nets? : World Bank.

Penn-Edwards, S. (2010). Computer aided phenomenography: The role of

Leximancer computer software in phenomenographic investigation. The

Qualitative Report, 15(2), 252.

Peters, K., Bahadur, A., Tanner, T., & Langston, L. (2016). ‘Resilience’across the

post-2015 frameworks: towards coherence?



Bibliography 271

Petrini, M. A., Rocha, J. V., Brown, J. C., & Bispo, R. C. (2016). Using an analytic

hierarchy process approach to prioritize public policies addressing family

farming in Brazil. LAND USE POLICY, 51, 85-94.

Pfefferbaum, R. L., Pfefferbaum, B., Van Horn, R. L., Klomp, R. W., Norris, F. H.,

& Reissman, D. B. (2013). The Communities Advancing Resilience Toolkit

(CART): An Intervention to Build Community Resilience to Disasters.

Journal of Public Health Management and Practice, 19(3), 250-258.

doi:10.1097/PHH.0b013e318268aed8

Pickard, A. J. (2013). Research methods in information: Facet publishing.

Plough, A., Fielding, J. E., Chandra, A., Williams, M., Eisenman, D., Wells, K. B.,

Law, G. Y., Fogleman, S., & Magaña, A. (2013). Building Community

Disaster Resilience: Perspectives From a Large Urban County Department of

Public Health. American Journal of Public Health, 103(7), 1190-1197.

doi:10.2105/ajph.2013.301268

Punch, K. F. (2013). Introduction to social research: Quantitative and qualitative

approaches: sage.

Qasim, S., Qasim, M., Shrestha, R. P., Khan, A. N., & Tun, K. (2016a). Community

resilience to flood hazards in Khyber Pukhthunkhwa province of Pakistan.

International Journal of Disaster Risk Reduction.

Qasim, S., Qasim, M., Shrestha, R. P., Khan, A. N., Tun, K., & Ashraf, M. (2016b).

Community resilience to flood hazards in Khyber Pukhthunkhwa province of

Pakistan. International Journal of Disaster Risk Reduction, 18, 100-106.


Quinlan, A. E., Berbés‐Blázquez, M., Haider, L. J., & Peterson, G. D. (2015).

Measuring and assessing resilience: broadening understanding through

multiple disciplinary perspectives. Journal of Applied Ecology.

Reed, M. S., Fraser, E. D. G., & Dougill, A. J. (2006). An adaptive learning process

for developing and applying sustainability indicators with local communities.

Ecological Economics, 59(4), 406-418.

doi:https://doi.org/10.1016/j.ecolecon.2005.11.008

Renschler, C. S., Frazier, A., Arendt, L., Cimellaro, G.-P., Reinhorn, A. M., &

Bruneau, M. (2010). A framework for defining and measuring resilience at

the community scale: the PEOPLES resilience framework. MCEER, Buffalo.

Rockström, J. (2003). Resilience building and water demand management for

drought mitigation. Physics and Chemistry of the Earth, Parts A/B/C, 28(20),

869-877.

Rodrigues, A. S., & Brooks, T. M. (2007). Shortcuts for biodiversity conservation

planning: the effectiveness of surrogates. Annual review of ecology,

evolution, and systematics, 713-737. doi:10.1

146/annurev.ecolsys.38.091206.095737

Saaty, T. L. (1990). How to make a decision: the analytic hierarchy process.

European journal of operational research, 48(1), 9-26.

Saaty, T. L. (2008). Decision making with the analytic hierarchy process.

International journal of services sciences, 1(1), 83-98.

Saja, A. M. A., Goonetilleke, A., Teo, M., & Ziyath, A. M. (2019). A critical review

of social resilience assessment frameworks in disaster management.

International Journal of Disaster Risk Reduction, 101096.

doi:https://doi.org/10.1016/j.ijdrr.2019.101096

Saja, A. M. A., Teo, M., Goonetilleke, A., & Ziyath, A. M. (2018). An inclusive and

adaptive framework for measuring social resilience to disasters. International


https://doi.org/10.1016/j.ecolecon.2005.11.008

https://doi.org/10.1016/j.ijdrr.2019.101096

Bibliography 272

Journal of Disaster Risk Reduction, 28, 862-873.

doi:https://doi.org/10.1016/j.ijdrr.2018.02.004

Saleh, A., & Bista, K. (2017). Examining factors impacting online survey response

rates in educational research: Perceptions of graduate students. Journal of

MultiDisciplinary Evaluation, 13(29), 63-74.

Sanyal, S., & Routray, J. K. (2016). Social capital for disaster risk reduction and

management with empirical evidences from Sundarbans of India.


Sapsford, R. (2006). Survey research (2nd edition ed.): Sage.

Saunders, M., Lewis, P., & Thornhill, A. (2016). Research Methods for Business

students.(ed. 7 th) Harlow: Pearson Education Limited.

Saunders, M. N., Lewis, P., Thornhill, A., & Bristow, A. (2015). Understanding

research philosophy and approaches to theory development.

Saunders, M. N. K., Lewis, P., & Thornhill, A. (2019). Research methods for

business students.

Schilling, K. E., Kim, S.-W., & Jones, C. S. (2017). Use of water quality surrogates

to estimate total phosphorus concentrations in Iowa rivers. Journal of

Hydrology: Regional Studies, 12, 111-121.

doi:https://doi.org/10.1016/j.ejrh.2017.04.006

Schipper, E. L. F., & Langston, L. (2015). A comparative overview of resilience

measurement frameworks.

Scopus. (2017). About Scopus Retrieved from

https://www.elsevier.com/solutions/scopus

Sempier, T., Swann, D., Emmer, R., Sempier, S., & Schneider, M. (2010). Coastal

community resilience index: a community self-assessment. online]

http://www. masgc. org/pdf/masgp/08-014. pdf (accessed 17 June 2013).

Serfilippi, E., & Ramnath, G. (2018). RESILIENCE MEASUREMENT AND

CONCEPTUAL FRAMEWORKS: A REVIEW OF THE LITERATURE.

Annals of Public and Cooperative Economics. doi:10.1111/apce.12202

Settle, S., Goonetilleke, A., & Ayoko, G. A. (2007). Determination of surrogate

indicators for phosphorus and solids in urban stormwater: application of

multivariate data analysis techniques. Water, air, and soil pollution, 182(1-4),

149-161.

Sharifi, A. (2016). A critical review of selected tools for assessing community

resilience. Ecological Indicators, 69, 629-647.

doi:10.1016/j.ecolind.2016.05.023

Sharifi, A., & Yamagata, Y. (2016). On the suitability of assessment tools for

guiding communities towards disaster resilience. International Journal of

Disaster Risk Reduction, 18, 115-124.

Sharma, U., & Patt, A. (2012). Disaster warning response: the effects of different

types of personal experience. Natural Hazards, 60(2), 409-423.

Sharp, J. L., Mobley, C., Hammond, C., Withington, C., Drew, S., Stringfield, S., &

Stipanovic, N. (2012). A Mixed Methods Sampling Methodology for a

Multisite Case Study. Journal of Mixed Methods Research, 6(1), 34-54.

doi:10.1177/1558689811417133

Shaw, D., Scully, J., & Hart, T. (2014). The paradox of social resilience: How

cognitive strategies and coping mechanisms attenuate and accentuate

resilience. Global Environmental Change, 25, 194-203.


https://doi.org/10.1016/j.ijdrr.2018.02.004

https://doi.org/10.1016/j.ejrh.2017.04.006

https://www.elsevier.com/solutions/scopus

http://www/


Bibliography 273

Sherrieb, K., Louis, C. A., Pfefferbaum, R. L., Betty Pfefferbaum, J. D., Diab, E., &

Norris, F. H. (2012). Assessing community resilience on the US coast using

school principals as key informants. International Journal of Disaster Risk

Reduction, 2, 6-15. doi:http://dx.doi.org/10.1016/j.ijdrr.2012.06.001

Sherrieb, K., Norris, F. H., & Galea, S. (2010). Measuring capacities for community

resilience. Social Indicators Research, 99(2), 227-247. doi:10.1007/s11205-

010-9576-9

Sinclair, R. G., Rose, J. B., Hashsham, S. A., Gerba, C. P., & Haas, C. N. (2012).

Criteria for selection of surrogates used to study the fate and control of

pathogens in the environment. Appl. Environ. Microbiol., 78(6), 1969-1977.

Singh, A., Ayoko, G. A., Herngren, L., & Goonetilleke, A. (2013). A Multivariate

Approach to the Identification of Surrogate Parameters for Heavy Metals in

Stormwater. Water, Air, & Soil Pollution, 1(224), 1-9.

Singh, K. (2007). Quantitative social research methods: Sage.

Singh, R. K., Murty, H. R., Gupta, S. K., & Dikshit, A. K. (2009). An overview of

sustainability assessment methodologies. Ecological Indicators, 9(2), 189-

212.

Sorensen, J. H., & Sorensen, B. V. (2007). Community Processes: Warning and

Evacuation Handbook of Disaster Research (pp. 183-199). New York, NY:

Springer New York.

Sotiriadou, P., Brouwers, J., & Le, T.-A. (2014). Choosing a qualitative data analysis

tool: a comparison of NVivo and Leximancer. Annals of Leisure Research,

17(2), 218-234. doi:10.1080/11745398.2014.902292

Sphere. (2018). Humanitarian Charter and Minimum Standards in Disaster

Response - 2018 Edition. Geneva, Switzerland.

Stough, L. M., & Kang, D. (2015). The Sendai Framework for Disaster Risk

Reduction and Persons with Disabilities. International Journal of Disaster

Risk Science, 6(2), 140-149. doi:10.1007/s13753-015-0051-8

Swanborn, P. (2010). Case study research: What, why and how? : Sage.

Tapsell, S. (2007). Social indicator set. T11-07-01.

Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed methods research:

Integrating quantitative and qualitative approaches in the social and

behavioral sciences: Sage.

Thoresen, S., Birkeland, M. S., Wentzel-Larsen, T., & Blix, I. (2018). Loss of trust

may never heal. Institutional trust in disaster victims in a long-term

perspective: associations with social support and mental health. Frontiers in

Psychology, 9, 1204.

Tierney, K. (2009). Disaster response: Research findings and their implications for

resilience measures. Retrieved from

Timmerman, P. (1981). Vulnerability resilience and collapse ofsociety. Rev. Models

Possible Clim. Appli-Cations.

Townshend, I., Awosoga, O., Kulig, J., & Fan, H. (2015). Social cohesion and

resilience across communities that have experienced a disaster. Natural

Hazards, 76(2), 913-938. doi:10.1007/s11069-014-1526-4

Tulloch, A. I., Chadès, I., Dujardin, Y., Westgate, M. J., Lane, P. W., &

Lindenmayer, D. (2016). Dynamic species co‐occurrence networks require

dynamic biodiversity surrogates. Ecography, 39(12), 1185-1196.

Twigg, J. (2009). Characteristics of a disaster-resilient community: a guidance note

(version 2).


Bibliography 274

Twigg, J., Kett, M., & Lovell, E. (2018). Disability inclusion and disaster risk

reduction.

Tyler, S., Nugraha, E., Nguyen, H. K., Nguyen, N. V., Sari, A. D., Thinpanga, P.,

Tran, T. T., & Verma, S. S. (2016). Indicators of urban climate resilience: A

contextual approach. Environmental Science & Policy, 66, 420-426.

doi:https://doi.org/10.1016/j.envsci.2016.08.004

Udie, J., Bhattacharyya, S., & Ozawa-Meida, L. (2018). A conceptual framework for

vulnerability assessment of climate change impact on critical oil and gas

infrastructure in the niger delta. Climate, 6(1), 11.

UN. (2016). Report of the open-ended intergovernmental expert working group on

indicators and terminology relating to disaster risk reduction. Retrieved from

https://www.preventionweb.net/files/50683_oiewgreportenglish.pdf

UNHABITAT. (2015). Kalmunai - Disaster Risk Reduction and Preparedness Plan -

towards a sustainable and resilient city. Retrieved from

http://unhabitat.lk/wp-content/uploads/2015/01/DRRPKmEng.pdf

UNISDR. (2009). 2009 UNISDR Terminology on Disaster Risk Reduction. Retrieved

from United Nations International Strategy for Disaster Reduction:

http://www.unisdr.org/files/7817_UNISDRTerminologyEnglish.pdf

UNISDR. (2015a). Proposed Updated Terminology on Disaster Risk Reduction: A

Technical Review. Retrieved from

http://www.preventionweb.net/files/45462_backgoundpaperonterminologyau

gust20.pdf

UNISDR. (2015b). Sendai framework for disaster risk reduction 2015–2030.: United

Nations International Strategy for Disaster Reduction Retrieved from

http://www.wcdrr.org/uploads/Sendai_Framework_for_Disaster_Risk_Reduc

tion_2015-2030.pdf.

UnitedWay. (1996). Measuring Program Outcomes: A Practical Approach.

USIOTWSP. (2007). How Resilient is Your Coastal Community? A Guide for

Evaluating Coastal Community Resilience to Tsunamis and Other Coastal

Hazards. Retrieved from Bangkok, Thailand:

Uslaner, E. M. (2016). Disasters, Trust, and Social Cohesion.

Vaismoradi, M., Jones, J., Turunen, H., & Snelgrove, S. (2016). Theme development

in qualitative content analysis and thematic analysis. Journal of Nursing

Education and Practice, 6(5).

Verheyden, T., & De Moor, L. (2016). Process-oriented social responsibility

indicator for mutual funds: A multi-criteria decision analysis approach.

International Journal of Multi-Criteria Decision Making, 6(1), 66-99.

VP. (2013). PROMETHEE Methods - Visual PROMETHEE 1.4 Manual. In V.

Solutions (Ed.).

Wedley, W. C. (1993). Consistency prediction for incomplete AHP matrices.

Mathematical and Computer Modelling, 17(4-5), 151-161.

Weeraratne, B. (2016). Re-defining urban areas in Sri Lanka. Colombo: Institute of

Policy Studies of Sri Lanka.

Wilkinson, O. (2015). Faith and Resilience after Disaster - The Case of Typhoon

Haiyan. Retrieved from

Woolf, S., Twigg, J., Parikh, P., Karaoglou, A., & Cheab, T. (2016). Towards

measurable resilience: A novel framework tool for the assessment of

resilience levels in slums. International Journal of Disaster Risk Reduction,

19, 280-302. doi:http://dx.doi.org/10.1016/j.ijdrr.2016.08.003

https://doi.org/10.1016/j.envsci.2016.08.004

https://www.preventionweb.net/files/50683_oiewgreportenglish.pdf

http://unhabitat.lk/wp-content/uploads/2015/01/DRRPKmEng.pdf

http://www.unisdr.org/files/7817_UNISDRTerminologyEnglish.pdf

http://www.preventionweb.net/files/45462_backgoundpaperonterminologyaugust20.pdf

http://www.preventionweb.net/files/45462_backgoundpaperonterminologyaugust20.pdf

http://www.wcdrr.org/uploads/Sendai_Framework_for_Disaster_Risk_Reduction_2015-2030.pdf

http://www.wcdrr.org/uploads/Sendai_Framework_for_Disaster_Risk_Reduction_2015-2030.pdf


Bibliography 275

WorldBank. (2014). 3.12 World Development Indicators: Urbanization. Retrieved

from: http://wdi.worldbank.org/table/3.12

Wright, K., & Johnston, D. (2010). Post-earthquake sheltering needs: how loss of

structures and services affects decision making for evacuation. Paper

presented at the 2010 New Zealand Society for Earthquake Engineering

Conference Proceedings.

Wu, H., & Drolet, J. (2016). Adaptive social protection: Climate change adaptation

and disaster risk reduction Social development and social work perspectives

on social protection (pp. 116-139): Routledge.

Yamamoto, L., Serraglio, D. A., & Cavedon-Capdeville, F. d. S. (2018). Human

mobility in the context of climate change and disasters: a South American

approach. International Journal of Climate Change Strategies and

Management, 10(1), 65-85.

Yin, R. K. (2014). Case study research: Design and methods (5th ed.): Sage

publications.

Yoon, D. K., Kang, J. E., & Brody, S. D. (2016). A measurement of community

disaster resilience in Korea. Journal of Environmental Planning and

Management, 59(3), 436-460. doi:10.1080/09640568.2015.1016142

Ziyath, A. M., Teo, M., & Goonetilleke, A. (2013, 17-19 September 2013 ).

Surrogate indicators for assessing community resilience. Paper presented at

the Inter-national Conference on Building Resilience, Ahungalla, Sri Lanka.

Zubair, L., Ralapanawe, V., Yahiya, Z., Perera, R., Tennakoon, U., Chandimala, J.,

Razick, S., & Lyon, B. (2005). Fine Scale Natural Hazard Risk and

Vulnerability Identification Informed by Climate in Sri Lanka. Project

Report: International Research Institute for Climate and Society. New York.

http://wdi.worldbank.org/table/3.12

Appendices 277

Appendices

Appendix A Existing social resilience measures identified in social

resilience frameworks analysed in this study

Appendix B QUT Human Research Ethics Committee Approval

Appendix C Letter of recruitment for interview participants

Appendix D Participant interview consent form

Appendix E Participant information sheet for interviews

Appendix F Interview guide

Appendix G Participant recruitment email for the survey

Appendix H Participant information for the survey

Appendix I Survey questionnaire in key survey

Appendix J Abstracts of published and under review manuscripts

Appendices 278

Appendix A

Existing social resilience measures identified in 31 social resilience frameworks

analysed in this study

The social resilience measure for other indicators of the ‘5S’ framework (Figure 2.5)

identified from the review of 31 existent social resilience frameworks are provided in

Saja et al. (2018).

A Sub-dimension: Social structure

References from the 31 social

resilience frameworks analysed Existing measures for access to transport

1 % population with vehicle access /Transport

dependence ratio

Kusumastuti, et al. (2014), Burton

(2015), Kotzee and Reyers (2016)

2 % Households with at least one

vehicle/vehicle facility

Cutter, et al. (2010), Qasim,

Qasim, Shrestha, Khan, Tun, et al.

(2016)

3 Access/evacuation potential Kotzee and Reyers (2016)

4 Available transportation means Mayunga (2007)

B Sub-dimension: Social capital References from the 31 social

resilience frameworks analysed Existing measures for social trust

5 Level of ethnic segregation Joerin, et al. (2014)

C Sub-dimension: Social mechanisms References from the 31 social

resilience frameworks analysed

Existing measures for past experience with disaster recovery/Learning from

the past

6 Hazard severity Cutter (2016)

D Sub-dimension: Social equity/diversity References from the 31 social

resilience frameworks analysed Existing measures for Involvement and equality for people with special needs

None N/A

E Sub-dimension: Social beliefs/culture/faith References from the 31 social

resilience frameworks analysed Existing indicators for existing cultural and behavioural norms

None N/A

Appendices 279

Appendix B

QUT Human Research Ethics Committee Approval

Appendices 280

Appendix C

Letter of recruitment for interview participants

10.11.2017

………………………………………

Subject Title:

Participation in a research study on surrogate indicators for measuring social resilience to

disasters (Interviews)

Dear colleagues

My name is Aslam Saja from the Science and Engineering Faculty, Queensland University of

Technology (QUT). I am undertaking a PhD on the use of surrogate indicators for measuring

social resilience to disasters.

The aim of this research is to identify, evaluate, and select surrogate indicators to measure

social resilience to disasters.

I’m looking for participants with minimum three years of experience in disaster management

at policy/ implementation level to participate in an interview which is being undertaken as a

key component of my doctoral study. Interview will last for approximately 75 to 90 minutes.

Interviews will be scheduled on a date and time that is convenient for you and at the place of

your choice.

More details of this project and about the interviews can be found in the attached Participant

Information Sheet.

If you are interested in participating or have any questions, please contact me via phone or

email.

Please note that this study has been approved by the QUT Human Research Ethics Committee

(approval number 1700000832). Any personal information that is collected, will be treated as

confidential. Participation in the interviews is strictly voluntary and your written consent will be

sought prior to conducting interviews.

Many thanks for your consideration of this request.

Aslam Saja

PhD Student (+94) 77 395 8387 [email protected]

Melissa Teo

Principal Supervisor (+617) 3138 9953 [email protected]

Science and Engineering Faculty, Queensland University of Technology

mailto:[email protected]


Appendices 281

Appendix D

Participant interview consent form

CONSENT FORM FOR QUT RESEARCH PROJECT

– Individual Interview –

Developing Surrogate Indicators to Measure

Social Resilience to Disasters QUT Ethics Approval Number 1700000832

RESEARCH TEAM

Mr Aslam Saja +94 773 958 387 [email protected]

Dr Melissa Teo +61 731 389 953 [email protected]

Prof Ashantha Goonetilleke +61 731 381 539 [email protected]

STATEMENT OF CONSENT

By signing below, you are indicating that you:

Have read and understood the information document regarding this project.

Have had any questions answered to your satisfaction.

Understand that if you have any additional questions you can contact the

research team.

Understand that you are free to withdraw at any time, without comment or

penalty.

Understand that if you have concerns about the ethical conduct of the project

you can contact the Research Ethics Advisory Team on +61 7 3138 5123 or

email [email protected].

Understand that the interview will include an audio recording.

Understand that non-identifiable data from this project may be used as

comparative data in future projects.

Agree to participate in the interview.

Name

Signature

Date

PLEASE RETURN THIS SIGNED CONSENT FORM TO THE RESEARCHER.





Appendices 282

Appendix E

Participant information sheet for interviews

PARTICIPANT INFORMATION FOR QUT RESEARCH PROJECT

– Individual Interview –

Developing Surrogate Indicators to Measure

Social Resilience to Disasters QUT Ethics Approval Number 1700000832

RESEARCH TEAM

Principal Researcher: Mr Aslam Saja PhD Student

Associate Researchers: Dr Melissa Teo Principal Supervisor

Prof Ashantha Goonetilleke Associate Supervisor


Queensland University of Technology (QUT)

DESCRIPTION

This research project is being undertaken as part of the PhD study by Aslam Saja.

The purpose of this research project is to develop surrogate indicators to measure social

resilience to disasters.

Disasters continue to cause enormous loss of lives and livelihoods, and severe disruptions

to social structures. Building resilient communities is crucial to reduce damages and to

recover quickly from disaster disruptions. Strengthening social resilience will help in

minimising damages and enable faster recovery from disasters. Hence, measuring social

resilience to disasters becomes essential. This research will develop a framework to identify,

evaluate, and select key surrogate indicators to measure social resilience to disasters.

You are invited to participate in this research project because you are a representative of a

government organization/local authority or a local/international non-governmental agency

working in Ampara district, Sri Lanka, who is involved in disaster management/disaster

resilience activities in your organization.

PARTICIPATION

Your participation will involve an audio recorded interview in your office or at an agreed

location at your convenience that will take approximately 90 minutes of your time.

Questions will be asked, such as:

What are the elements that indicate the ability of families to help their neighbours

during disasters?

How can we assess the level of trust among people to help them in disasters?

What are the sources of information that tell us your community’s trust in local

organizations/authorities helping them in disaster recovery?

Appendices 283

Your participation in this project is entirely voluntary. Even if you do agree to participate

you can withdraw from the project without comment or penalty. If you withdraw, on

request any identifiable information already obtained from you will be destroyed. Your

decision to participate or not participate will in no way impact on your current or future

relationship with QUT, Australia.

EXPECTED BENEFITS

It is expected that this project may not benefit you directly. However, it is expected to

benefit disaster management stakeholders engaged in planning and implementing

projects aimed to build social resilience. This research will contribute to measure social

resilience characteristics to disasters using surrogate indicators that are easily accessible.

Other expected benefits include:

Better understanding of the complex interrelationship of social resilience to disasters;

Addressing the practical and methodological challenges in measuring social resilience

indicators; and

Contribute to address the current knowledge gap in identifying and selecting

surrogate indicators to measure social resilience.

To recognise your contribution should you choose to participate, the research team is

offering a small souvenir.

RISKS

There are minimal risks associated with your participation in this project.

These include:

A low likelihood of unintentional disclosure of confidential information pertaining to

disaster management practices. Your responses and personal information will be treated

as confidential and aliases will be used instead of your real name in all transcripts and

future publications.

You may also have to give up some of your personal time to participate in the research.

The risks will be minimized through communication with you before the interview takes

place by obtaining your permission and willingness to be interviewed at your

convenience.

You may be suspicious of the researcher’s action and about the research. To reduce this

risk, you will be provided with a detailed explanation about the research and its purposes

prior to the interview.

To manage these risks, you are also reminded to consider the following before

participating in the interview or when responding to interview questions:

Seek the permission from your organization/superior (if needed) before participating

and abide by confidentiality and privacy rules as per your organisation’s requirements.

Consider your responses in relation to your experience/knowledge on social resilience

in disaster context.

Inform the interviewer if there are questions that they are uncomfortable with. Response

to questions are voluntary and you can withdraw your consent at any time. You also

have the option to verify the information provided prior to final inclusion in the research.

PRIVACY AND CONFIDENTIALITY

All comments and responses will be treated confidentially unless required by law. The

names of individual persons are not required in any of the responses and no names or

identifying information will be included in any reporting of the research.

Appendices 284

As the interview involves an audio recording:

You will have the opportunity to verify your comments and responses prior to final

inclusion.

The recording will be destroyed 5 years after the last publication.

The recording will not be used for any other purpose.

Only the named researchers will have access to the recording.

It is not possible to participate in the research without being recorded as all

information will be translated to English after the interviews are completed.

No specific individual can be identified by the researchers.

All data collected will be stored in secure severs at QUT, Australia, as per QUT’s

Management of Research Data policy. Please note that non-identifiable data from this

project may be used as comparative data in future projects or stored on an open access

database for secondary analysis.

CONSENT TO PARTICIPATE

We would like to ask you to sign a written consent form (enclosed) to confirm your

agreement to participate.

QUESTIONS / FURTHER INFORMATION ABOUT THE PROJECT

If have any questions or require further information please contact one of the listed

researcher:

Aslam Saja +94 773 958 387 [email protected]

Melissa Teo +61 731 389 953 [email protected]

CONCERNS / COMPLAINTS REGARDING THE CONDUCT OF THE PROJECT

QUT is committed to research integrity and the ethical conduct of research projects.

However, if you do have any concerns or complaints about the ethical conduct of the

project you may contact the QUT Research Ethics Advisory Team on +61 7 3138 5123 or

email [email protected]. The QUT Research Ethics Advisory Team is not connected

with the research project and can facilitate a resolution to your concern in an impartial

manner.

THANK YOU FOR HELPING WITH THIS RESEARCH PROJECT.

PLEASE KEEP THIS SHEET FOR YOUR INFORMATION.




Appendices 285

Appendix F

Interview guide

Interview structure Questions in each interview will be asked for following five social resilience indicators.

Interview questions will follow the same structure for each social resilience indicator:

- What are the potential surrogates to measure <social resilience indicator 1/2/3/4/5>

(To identify potential surrogates)

- Why it is the suitable surrogate indicator to measure <social resilience indicator

1/2/3/4/5>

(To establish surrogacy relationship to the target indicator)

- How can we measure those surrogates proposed for <social resilience indicator

1/2/3/4/5> (To identify protocols for surrogates to measure the target indicator)

- Probe more detail on each theme proposed

Interview questions 1. Social Mobility:

A key social resilience indicator to measure ability of people to have greater mobility during

and after disasters is having access to adequate transport facility.

Following questions will be asked to measure “social mobility and access to transport

facilities” for evacuation during/after a disaster.

Key target

indicator

Questions

Social

mobility

and access

to a

transport

facility

1. How people in your area evacuate when disaster early warning is issued?

2. What are the transport modes and facilities available for evacuation

during disasters in your area?

3. How did you assess the accessibility to a transport facility for evacuation

of people?

4. What are the other elements that can indicate the availability of transport

facilities and access to those transport facilities when needed during

disasters?

5. How can we know that there is adequate transport facility available for

evacuation during disasters?

Indicator No Social resilience indicator

Indicator 1 Social mobility and access to transport facility for evacuation

Indicator 2 Social trust in disaster preparedness/response/recovery

Indicator 3 Involvement/equity for people with special needs in disasters

Indicator 4 Learnings from the past disaster experience

Indicator 5 Cultural and behavioural norms that help/impede in disasters

Appendices 286

Probing more as to why it is a good indicator based on the potential themes

that participant proposed and what are the sources from which they can be

measured.

2. Social cohesion:

A key social resilience indicator to measure social cohesion in disaster management

activities is social trust (among people/with neighbours/with local organisations/local

authorities).

Following questions will be asked to measure “social trust” in disaster management

activities.

Key target

indicator

Questions

Social

trust

6. What do you think is the level of trust among people in your area?

7. During a disaster, will people in the neighbourhood help each other? If

so, to what extent?

8. How can we assess trust in the neighbourhood for help in the event of a

future disaster?

9. How can we assess trust of people in community based organizations and

non-government organisations in your area?

10. How can we assess trust of people in government institutions and local

government authorities?

Probing more as to why it is a good indicator based on the themes that

participant proposed and what are the sources from which they can be

measured.

3. Community inclusiveness and equity:

A key social resilience indicator to measure community inclusiveness and equity is

involvement/equity for people with special needs in disaster situations.

Following questions will be asked to measure “involvement/equity for people with

special needs” in disaster situations.

Key

target

indicator

Questions

Equity

for

people

with

special

needs

11. What type of assistance people with special needs (such as disable, elders,

marginalized people) have access to prepare for disasters?

12. What are the indicators that help to measure the assistance to people with

specific needs in general?

13. How can we assess the access to different types of assistance for people

with specific needs for preparing for a disaster?

14. How the equity of assistance and services to people with special needs

are ensured in your area?

15. How can we identify whether people with specific needs have access to

adequate assistance and services they require to manage disasters?



measured.

Appendices 287

4. Community competence:

A key social resilience indicator to measure community competence is learnings from

the past disaster experience.

Following questions will be asked to measure “learning from the past disaster

experience”.

Key target

indicator

Questions

Learnings

from the

past

disaster

experience

16. What are some initiatives of the community to mitigate and prepare for

disasters?

17. Can you provide some examples of the community initiatives that have

been done since the occurrence of the last disaster, which can help them

to minimize the impact of the future disasters?

18. What indications do you have that the community has done something

new since the last disaster, to mitigate the risk associated with the disaster

or to prepare to face the next disaster?

19. What are the indicators that help us to assess the community’s learning

from the past disaster experience?

20. How can we know that the community has taken some learnings from

the past disaster experiences and taken new initiatives/innovations to

increase its resilience?



measured.

5. Local culture/beliefs/faith:

A key social resilience indicator to measure local culture/beliefs/faith influence in

disaster is to measure cultural and behavioural norms that help/impede in disasters.

Following questions will be asked to measure “cultural and behavioural norms that

help/impede in disasters”.

Key target

indicator

Questions

Cultural

and

behavioural

norms that

help/impede

in disasters

21. What are some cultural norms in your area that is very specific only to

your area?

22. How do those cultural norms help in preparing for and responding to

disaster events? Provide some examples.

23. What are some indications of cultural norms that is specific in your

area? /what indicators help you to assess that certain cultural norms

exists that help/do not help in preparing for disasters?

24. How can we assess those cultural norms that help in preparing for

disasters/ that makes peoples’ less resilient to face disasters?

25. If you have some cultural norms specific to the area that help people

in preparing for disasters, how you make use of them in your work in

preparing communities for disasters?



measured.

Appendices 288

Appendix G

Participant recruitment email for the survey

Subject Title: Participation in a research study on surrogate indicators for measuring

social resilience to disasters (Online survey questionnaire)

Dear colleagues

My name is Aslam Saja from the Science and Engineering Faculty, Queensland

University of Technology (QUT). I am undertaking a PhD on the use of surrogate

indicators for measuring social resilience to disasters.

The aim of this research is to identify, evaluate, and select surrogate indicators to

measure social resilience to disasters.

I’m looking for participants with a minimum three years of experience in disaster

management at policy/ implementation/research level to complete a 15 to 20

minute online survey which is being undertaken as a key component of my doctoral

study.

More details of this project and about the online survey can be found in the

attached Participant Information Sheet.

Further details on the study and how to participate can be found by clicking on the

following link: xxxxxxxxxxxxxxxxxxx

If you are interested in participating or have any questions, please contact me via

email.

Please note that this study has been approved by the QUT Human Research Ethics

Committee (approval number 1700000832). The online survey is anonymous and

personal identifiable information such as your name or your contact details are NOT

required.

Many thanks for your consideration of this request.

Aslam Saja

PhD Student (+94) 77 395 8387 [email protected]

Melissa Teo

Principal Supervisor (+617) 3138 9953 [email protected]

Science and Engineering Faculty, Queensland University of Technology



Appendices 289

Appendix H

Participant information for the survey

PARTICIPANT INFORMATION FOR QUT RESEARCH PROJECT

Developing Surrogate Indicators to Measure Social

Resilience to Disasters QUT Ethics Approval Number 1700000832

RESEARCH TEAM

Principal Researcher: Mr Aslam Saja PhD Student

Associate Researchers: Dr Melissa Teo Principal Supervisor

Prof Ashantha Goonetilleke Associate Supervisor


Queensland University of Technology (QUT)

DESCRIPTION

This project is being undertaken as part of a PhD by Aslam Saja.

The purpose of this project is to develop surrogate indicators to measure social resilience to

disasters.

You are invited to participate in this project because you are a disaster management

professional or researcher with at least 3 years’ direct experience in disaster management in a

government institution or a non-government organization (NGO) or research/academic

institution and are aged above 18 years old.

PARTICIPATION

Your participation will involve the completion of an online survey and will take approximately

15 to 20 minutes of your time.

Questions will include:

In your opinion, how accurately <Surrogate indicator 1> can measure social trust

(Accuracy)

In your opinion, how easily the data can be accessible for <Surrogate indicator 1> to

measure social trust (Measurement complexity)

In your opinion, can <Surrogate indicator 1> be used to measure social trust in different

phases of a disaster (Time-sensitivity)

In your opinion, how easily can social trust be communicated using <Surrogate indicator

1> (Communicability)

In your opinion, how cost-effective <Surrogate indicator 1> is, to measure social trust

(Cost-effectiveness)

Your participation in this project is entirely voluntary. If you agree to participate, we ask that

you avoid skipping any questions where possible. Your decision to participate or not

participate will in no way impact upon your current or future relationship with QUT or

Appendices 290

University of Peradeniya. If you do agree to participate you can withdraw from the project

until the point of submission. Once the online survey is submitted, it will not be possible to

withdraw from the study as all responses are anonymous.

EXPECTED BENEFITS

It is expected that this project may not directly benefit you. However, it is expected to benefit

disaster management stakeholders engaged in planning and implementing projects aimed to

build social resilience. This research will contribute to measure social resilience characteristics

to disasters using surrogate indicators that are easily accessible.

Other expected benefits include:

• Better understanding of the complex interrelationship of social resilience to disasters;

• Addressing the practical and methodical challenges in measuring social resilience

indicators; and

• Contribute to address the current knowledge gap in identifying and selecting surrogate

indicators to measure social resilience. RISKS

There are no risks beyond your valuable time of 15-20 minutes associated with your

participation in online survey. There is a low likelihood of unintentional disclosure of

confidential information pertaining to disaster management practices.

PRIVACY AND CONFIDENTIALITY

All comments and responses are anonymous and will be treated confidentially unless

required by law. The names of individual persons are not required in any of the responses.

Any data collected as part of this project will be stored securely as per QUT’s Management of

research data policy. Please note that non-identifiable data from this project may be used as

comparative data in future projects or stored in an open access database for secondary

analysis.

CONSENT TO PARTICIPATE

Submitting the completed survey is accepted as an indication of your consent to participate

in this project.

QUESTIONS / FURTHER INFORMATION ABOUT THE PROJECT

If have any questions or require further information please contact one of the listed

researcher:

Aslam Saja +94 773 958 387 [email protected]

Melissa Teo +61 731 389 953 [email protected]

CONCERNS / COMPLAINTS REGARDING THE CONDUCT OF THE PROJECT

QUT is committed to research integrity and the ethical conduct of research projects.

However, if you do have any concerns or complaints about the ethical conduct of the project

you may contact the QUT Research Ethics Advisory Team on +61 7 3138 5123 or email

[email protected]. The QUT Research Ethics Advisory Team is not connected with the

research project and can facilitate a resolution to your concern in an impartial manner.

THANK YOU FOR HELPING WITH THIS RESEARCH PROJECT.

PLEASE KEEP THIS SHEET FOR YOUR INFORMATION.




Appendices 291

Appendix I

Survey questionnaire in KEY SURVEY

I.1: Introductory questions

1- Have you been involved in a disaster resilience project during the past 5 – 10

years? Yes b. No

2- If yes, in the disaster resilience project, what best describes your role in the

project?

a. Project Implementing Staff b. Project Manager c. Volunteer

d. Researcher e. Academic f. Policy maker g. other, please specify

3- Where do you currently work?

a. Government department b. Local NGO c. International NGO

d. UN agency e. Donor agency f. Private sector

g. Civil Society Organisation (CSO) h. Community Based

Organisation (CBO)

4. In which country do you currently work?

5. What are the countries you have undertaken/worked on disaster projects? List

all that applies

6. What is your gender? Male Female

7. How many years of experience you have in community resilience/disaster

resilience projects?

a. Less than 3 years b. 3-5 years c. 5-10 years d. More than 10 years

8. What is your highest level of education?

a. PhD b. Master degree c. Bachelor degree d. Advance diploma

e. Professional certificate f. High school education

Appendices 292

I.2: Evaluation of surrogates against five evaluation criteria:

A. Measuring access to transport facilities during and after disasters

Rate each surrogate proposed to measure “access to transport facility during and after

disasters” in the scale of 1 – 5 (Very high to very low)

I. How accurate are the following surrogate indicators proposed to measure “access

to transport facilities during and after disasters”?

Accuracy of the potential surrogate indicators

1 (Very high) 2 (High) 3 (Average) 4 (Low) 5 (Very low)

Surrogate 1

Surrogate 2

Surrogate 3

II. Rate the cost-effectiveness of the following surrogate indicators proposed to

measure “access to transport facilities during and after disasters”?

Cost-effectiveness of the potential surrogate indicators


Surrogate 1

Surrogate 2

Surrogate 3

III. In your opinion, how much the following surrogate indicators are difficult to

measure?

Measurement complexity of the potential surrogate indicators


Surrogate 1

Surrogate 2

Surrogate 3

IV. In your opinion, how easy the following surrogate indicators are to

communicate with the disaster management stakeholders?

Communicability of the potential surrogate indicators


Surrogate 1

Surrogate 2

Surrogate 3

V. In your opinion, can the following surrogate indicators be used to measure “access

to transport facility” in different time periods (Before/During/After)?

Time sensitivity of the potential surrogate indicators


Surrogate 1

Surrogate 2

Surrogate 3

Appendices 293

I.3. Importance of surrogate indicator evaluation criteria:

Please complete the following pair-wise comparison matrix of criteria based on the

importance of each criterion against another criterion list below on the scale of 1 to

5.

Appendices 294

Appendix J

Paper 1 Title: A critical review of social resilience assessment frameworks in

disaster management

Status: Published

Year: 2019

Journal: International Journal of Disaster Risk Reduction (Q1)

Abstract: In published format below

Appendices 295

Paper 2 Title: An inclusive and adaptive framework for measuring social

resilience to disasters

Status: Published

Year: 2018

Journal: International Journal of Disaster Risk Reduction (Q1)

Abstract: In published format below

Appendices 296

Paper 3 Title: Surrogate measures to assess mobility of people as a resilience

indicator in disaster management: An exploratory study in

Southeastern Sri Lanka

Status: Published

Year: 2020

Journal: International Journal of Disaster Risk Science (Q1)

Abstract: In the published format

Appendices 297

Paper 4 Title: Selection of surrogates to assess social resilience in disaster

management using Multi-Criteria Decision making

Status: Published

Year: 2020

Journal: International Journal of Disaster Resilience in the Built Environment

(Q1)

Abstract: In the published format

Appendices 298

surrogate approach to ssess social resilience in … majeed... · figure 1.1. total reported...

Documents