discontinuity in organizations: impacts of knowledge flows … · maintaining product feasibility...
TRANSCRIPT
DISCONTINUITY IN ORGANIZATIONS: IMPACTS OF KNOWLEDGE FLOWS ON
ORGANIZATIONAL PERFORMANCE
A DISSERTATION SUBMITTED TO THE
DEPARTMENT OF CIVIL AND ENVIRONMENTAL ENGINEERING
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
Rahinah Ibrahim
June 2005
© Copyright by Rahinah Ibrahim 2005 All Rights Reserved
ii
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy.
_____________________________________________ Boyd C. Paulson, Jr., Principal Adviser
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy.
_____________________________________________ Raymond E. Levitt
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy.
_____________________________________________ Mark E. Nissen
Approved for the University Committee on Graduate Studies.
iii
ABSTRACT
Maintaining product feasibility and managing knowledge flows are difficult if an
organization has to perform a complex process while operating in an equivocal
environment. My dissertation seeks to answer how knowledge flows impact the
organizational performance of an enterprise with discontinuous membership. Subject to
the process’s skills requirement, the position of the discontinuous member is added or
deleted during a workflow process. Utilizing data from the affordable housing domain, I
developed a mixed-method case-study research methodology combining research
methodologies from the field of anthropology (archival ethnography), sociology
(knowledge network analysis), and computer science and engineering (computational
organizational theory—COT).
The ethnographic study identified four operating environment constructs for
facility development: multiple sequential and concurrent phases, discontinuous
membership, interdependency of tasks, and different knowledge form dominating in
different phases. The COT modeling using SimVision® further developed and proposed
knowledge as another contingency factor in the organizational design fit. Additional
contingency fit parameters include reach and discontinuous. The knowledge network
analysis using social network analysis affirmed that knowledge flows (i.e., the
communications to retrieve and allocate information) between team members depend on
knowledge type and whether the agent is present in a workflow phase. Utilizing the
results of these three studies, I proposed and Marc Ramsey developed an extension of the
Virtual Design Team (VDT) tool to simulate the impacts of knowledge networks on
organizational performance in an enterprise with discontinuous membership.
iv
The cross-disciplinary mixed-method case-study research concludes that
discontinuity in organizations is a factor in the K-loss phenomenon in facility
development, and knowledge type and non-hierarchical information-processing of the
enterprise affects its organizational performance. The proof-of-concept COT model
demonstrates that inaccurate knowledge flow (i.e., communication to retrieve or allocate
information to enable action) in a discontinuous organization affects its organizational
performance. A discontinuous member’s inaccurate knowledge cognition could cause a
functional error at personal level, which is not obvious at the enterprise’s overall
performance level. The finding supports facility owners’ concern why knowledge loss
continues to occur despite their expensive investments in IT and knowledge management
for project management. It also affirms the existence of a non-hierarchical information-
processing system in an enterprise that led this dissertation to propose knowledge as
another contingency factor.
The dissertation advances the merging of two fields—knowledge flow dynamics
and organization—in the design of organizations based on knowledge flow
characteristics. Broader implications include the development of theories and applications
in the field of knowledge management; establishing foundations for mitigating the
knowledge loss phenomenon in dynamic environment; and refining the mixed-method
case-study research methodology for theoretical and application development in multi-
disciplinary research. My dissertation claims five contributions. First, it establishes
discontinuous membership in enterprises as a contributing factor to knowledge loss in
complex product development processes. Second, it merges knowledge flow dynamics
theory with organization theory in proposing knowledge as another contingency factor,
v
discontinuity as another structural configuration measure, and reach as another design
parameter property measure in the design of organizational fit. Third, it applies
established social network analysis methodology to study knowledge flow in an
engineering problem. Fourth, the dissertation extends Galbraith’s information-processing
theory to include Wegner’s transactive memory theory in the Virtual Design Team COT
tool. Finally, it develops a cross-disciplinary research methodology, which uses
established tools in other domains to solve an engineering problem.
vi
ACKNOWLEDGEMENTS
Praise is to God for all knowledge is from Him, and all the mistakes are humanely mine.
The Ministry of Science, Technology, and Innovation of Malaysia in affiliation with
Universiti Putra Malaysia, for sponsoring this doctoral study at Stanford University. In
addition, the UPS Foundation at Stanford University for partly sponsoring this study.
Professor Boyd C. Paulson, Jr., for his patience in allowing me to explore beyond the
humbleness of affordable housing; for establishing affordable housing as a rich source for
high-tech research; and for being a great friend and supporter in the Construction
Engineering and Management program.
Professor Raymond E. Levitt, for introducing me to organizations; and for embracing me
into the Virtual Design Team Research Group and the Collaboratory for Research on
Global Projects at Stanford University; and becoming a quality benchmark in my doctoral
work.
Professor Mark E. Nissen, for many hours of discussions that gradually changed my
belief that I could, indeed, achieve greater heights in knowledge contributions; for
developing my writing skills; and for being a cool intellect.
vii
My fellow colleagues at the Construction Engineering and Management Program,
especially those in the VDT and CRGP groups, for all the support and encouragement as
we walk the shaky doctoral study path together. My special thank you to Somroj
Vanichvatana, for his ‘virtual’ generosity and assistance.
My colleagues at the Faculty of Design and Architecture (especially Dean Dr. Mustafa
Kamal Mohd. Shariff), and the administration of Universiti Putra Malaysia, for their
support and encouragement.
The Housing Research Center, Universiti Putra Malaysia, for inculcating my research
interest.
Doha Hamza and Ahmed Sultan, Taqwa and Aldrin Aviananda, Sherifa and Atef
Ibrahim, Rahaf Choaib, Norhanani Muhammad and Amir Razelan; and everyone I came
to know at Stanford University; for keeping my feet on the ground, and for being my best
friends in times of happiness and in need.
Reverend John Hester, Reverend Dr. George Fitzgerald, Candace Mindigo, the Spiritual
Care Services’ volunteers; the Partners in Caring’s volunteers, and especially the 308
patients I visited at Stanford University Medical Center from July 2001 until December
2004, for giving me a reason to survive my doctoral pursuit, and becoming friends for
life.
viii
My Circle of Stanfordian Strangers, and everyone in the community for strengthening
my belief that humanity exists when the heart touches another. May we eventually meet
to know each another in person.
Anena Lanyom Otii, Atim Miriam Otii, and Abdul-Basit Ibin Whalid, for editing.
My parents, Bonda Kasmawati Abdullah and the late Ayahanda Ibrahim Haji Hamzah,
for instilling in their children the love for intellectual pursuits. Members of my family,
especially Kamsuri, Kamsuraini, Hazamri, and Iskazri, for extending their love and care
to my children while I was away.
Yelly Augustini, Kurais Abdul Karim, Hajjah Suadah, and Titin for holding the fort at
home.
My children, Raihan, Rafeah, Luqman Alhakeem, Muhammad Al-Ameen, Fatimah
Rabbaniyyah, Mujahid Rabbani, Umar Al-Farouq, Syaheeda Rabbaniyyah, and their
brothers and sisters for filling my life with joy, laughters, and lots of loving sibling
squabbles.
Foremost to my husband, Yg. Bhg. Dato’ Haji Mustafa Kamal Bin Haji Mohd. Zaini, for
his never-ending generosity of faith, love, support, and encouragement in ensuring my
successful pursuit towards intellectual excellence.
ix
TABLE OF CONTENTS
ABSTRACT …………………………………………….….……………………… iv
ACKNOWLEDGEMENTS …………………………………………………………. vii
TABLE OF CONTENTS …………………………………………………………….. x
LIST OF TABLES …..……………………………………………………………….. xv
LIST OF ILLUSTRATIONS ……………………………………………………….. xvii
Chapter
1 INTRODUCTION ……………………………….….……………………… 1
PART I: BACKGROUND INTRODUCTION
1.1 Overview ………………………………….……………….…………. 1
1.2 Statement of Problem ……………………..……………….…………. 2
1.3 Motivating Background Problem ………….……………….…………. 9
1.4 Point of Departure ………………………….……………….…………. 13
PART 2: MIXED-METHOD CASE-STUDY RESEARCH
1.5 Method …………………………………….……………….…………. 16
1.6 Research Questions ……………………….……………….…………. 16
1.7 Propositions ……………………………….……………….…………. 20
1.8 Unit of Analysis …………………………..……………….…………. 21
1.9 Logic Linking Data to Proposition ……………………….…………. 22
1.10 Criteria for Interpreting Findings …………………………………….. 23
1.11 Validation ………………..………………………………….…………. 24
1.12 Limitations ………………………………………………….…………. 26
PART 3: CONTRIBUTIONS, IMPLICATIONS, AND READER’S GUIDE
1.13 Claimed Contributions ……………………………………………….. 27
1.14 Implications …………………………………………………………… 30
x
1.15 Practical Implications …………………………………………………. 31
1.15.1 Knowledge Management and Knowledge Flows ……………… 31
1.15.2 Organization ……………………………………………………. 32
1.15.3 Construction Industry ………………………………………….. 33
1.15.4 Cross-Discipline Research …………………………………….. 34
1.16 Suggested Future Research ………………………………………….. 36
1.17 Reader’s Guide ………………………………………………………… 37
1.18 References …………………………………………………………….. 39
2 DISCONTINUITY IN ORGANIZATIONS: HOW ENVIRONMENTAL
CHARACTERISTICS CONTRIBUTE TO THE PROJECT’S
KNOWLEDGE LOSS PHENOMENON …………………………………… 43
2.1 Abstract ……………………………………………..………………… 43
2.2 Introduction ………………………………………………………….. 44
2.3 Point of Departure ……………………………………………………. 47
2.4 Ethnography Research Methodology …………………………………. 49
2.4.1 Overview ……………………………………………………… 49
2.4.2 Data Collection ……………………………………………….. 49
2.4.3 Data Analysis …………………………………………………. 50
2.4.4 Limitations of Study ………………………………………….. 51
2.4.5 Validation …………………………………………………….. 52
2.5 Determining Affordable Housing Development Milestones ………….. 52
2.6 Determining Affordable Housing Financing Milestones ……………… 57
2.7 Facility Development Phases ……………………………………….. 60
2.7.1 Sequential Facility Development Phases …………………….. 61
2.7.2 Concurrent Facility Development Phases …………………….. 66
2.8 The Evolving Organizations …………………………………………. 69
2.9 Operational Constructs ……………………………………………..... 71
2.10 Validation …………………………………………………………….. 74
2.11 Conclusions and Future Studies ……………………………………….. 78
xi
2.12 Acknowledgements …………………………………………………… 82
2.13 References ……………………………………………………………. 83
3 DISCONTINUITY IN ORGANIZATIONS: KNOWLEDGE FLOW
BEHAVIORS IN SEQUENTIAL PHASES ...……….……………………… 86
3.1 Abstract …………………………………………….………………… 86
3.2 Introduction …………………………………………………………… 87
3.3 Literature Review …………………………………………………….. 90
3.3.1 Organization Theory ………………………………………….. 90
3.3.2 Knowledge Management and Dynamic Flows Theories ……… 91
3.3.3 Transactive Memory Theory …………………………………. 95
3.3.4 Results from Ethnographic Study ……………………………... 98
3.4 Research Method …………………………………………………….. 100
3.5 Results ………………………………………………………………… 105
3.6 Discussion ……………………………………………………………. 110
3.7 Conclusions …………………………………………………………… 118
3.8 Acknowledgements …………………………………………………… 119
3.9 References …………………………………………………………….. 119
4 DISCONTINUITY IN ORGANIZATIONS: IMPACTS OF
KNOWLEDGE FLOWS ON ORGANIZATIONAL PERFORMANCE …… 124
4.1 Abstract ……………………………………………..………………… 124
4.2 Introduction ………………………………………………………….. 125
4.3 Literature Review ……………………………………………………. 128
4.3.1 Discontinuity in Organizations ……………………………… 128
4.3.2 Knowledge Flow Dynamics ……………………………..…… 129
4.3.3 Organization ………………………………………………….. 132
4.3.4 Transactive Memory …………………………………………. 133
4.3.5 Knowledge Flow and Computational Organization Theory ….. 138
xii
4.4 Research Method …………………………………………………….. 140
4.5 Functional Exceptions Parameter Extension for Knowledge
Network ……………………………………………………………..... 144
4.6 Results and Analysis …………………………………………………. 148
4.7 Validation …………………………………………………………….. 152
4.8 Discussion and Recommendations ……………………………………. 153
4.9 Conclusions …………………………………………………………… 157
4.10 References …………………………………………………………….. 158
5 DISCONTINUITY IN ORGANIZATIONS: DEVELOPING A
KNOWLEDGE-BASED ORGANIZATIONAL PERFORMANCE
MODEL FOR DISCONTINUOUS MEMBERSHIP ……….……………… 163
5.1 Abstract ……………………………………………..………………… 163
5.2 Introduction ………………………………………………………….. 164
5.3 Literature Review ……………………………………………………. 168
5.3.1 Information Processing in Organizations ……..….…………… 168
5.3.2 Knowledge-flow Dynamics …………………………………… 172
5.4 Modeling Knowledge Flows Computationally …..…………………… 176
5.4.1 Facility Development Characteristics ………………………... 177
5.4.2 Integrating Knowledge Flow Theory in COT ………………….. 182
5.4.3 A COT Tool …………………………………………………… 188
5.5 A COT Case-Study Model ……………………………………………. 189
5.6 Results, Analysis, and COT Validation …………………………..…..... 198
5.7 Discussion ………….…………………………………………………. 204
5.8 Conclusions ………………………………………………………….. 211
5.9 Acknowledgements …………………………………………………… 213
5.10 References …………………………………………………………….. 213
xiii
Appendix 1: Riverwood Case-Study ………………………………………………….. 219
1-1. Riverwood Fact Sheet …………………………………………………….… 220
1-2 Riverwood Workflows and Organizations for Test Cases ………………… 224
1-3 Project Summary Tasks List ……………………………………………….. 234
1-4 Property Development Documents and Schedule Tracer Form ……………. 236
1-5 Operation and Warranty Manual Content List …………………………….. 243
Appendix 2: Virtual Design Team – Knowledge Network ………………………… 245
Appendix 3: Knowledge Asset Mapping Exercise (KAME) …………………………. 248
3-1 Pre-Knowledge Asset Mapping Interview Protocol ………………………… 249
3-2 Riverwood Code ……………………………………………………………. 253
BIBLIOGRAPHY….…………………………………………………………………… 281
xiv
LIST OF TABLES
Table Page 1-1. Comparison of a Typical Residential Development Process between
Landis (2001), Ibrahim (2001), and the American Institute of Architects (1997) …………………………...……………………………………… 12
1-2 Mixed-method Case-study Steps for Multi-disciplinary Research ……… 18 1-3 Validation Tests for Mixed-method Case-study for Multi-disciplinary
Research ………………………………………………………………… 25 1-4 Summary of Results from Mixed-method Case-study for Multi-
disciplinary Research …………………………………………………… 35 2-1 List of Affordable Housing Development Cases ……………………….. 51 2-2 Examples of Equity Investment Programs, Permanent Soft and Hard
Loans for Affordable Housing Development (CRA, 1998) ……………. 59 2-3 Staff’s Position and Contributing Fulltime Equivalent (FTE) Allocations
for Different Facility Development Life Cycle Phase in an Affordable Housing Case-Study ……………………………………………………. 70
3-1 Comparison of Standard Coefficients of Being Continuous and Having
Perceived Expertise on Betweeness, Indegree, and Outdegree Centrality for Feasibility-Entitlements and Building Permit Phases ………………. 106
3-2 Comparison of MRQAP Coefficients Predicting Knowledge Networks
in the Feasibility-Entitlements Phase (Phase 1) …………………………. 108 3-3 Comparison of MRQAP Coefficients Predicting Knowledge Networks
in the Building Permit Phase (Phase 2) …………………………………. 110 4-1 Baseline Parameters for Discontinuous Membership Organization ……. 142 4-2 Environmental Consultant’s Knowledge Cognition of Other Team
Members in Baseline and X-Baseline Cases ………………………….. 142 4-3 Comparison of Selected Organizational Performance Measures for
Continuous Versus Discontinuous Membership ………………………… 149 4-4 Comparison of Organizational Performance Measures for Selected
Members ………………………………………………………………… 151
xv
5-1 Variables and Parameter Settings for Baseline Model ………………….. 193 5-2 Total Staffing and Position FTE’s for Baseline Model ………………… 195 5-3 Distribution of FTE’s for Team Members in City, Building, and Owner
Matrices ………………………………………………………………….. 197 5-4 Statistics of Selected Simulated Values for Baseline and Alternate Models …………………………………………………………………. 199 A-1 Multiple Skill Sets of the Riverwood Apartments Enterprise …………... 230 A-2 Riverwood Actors’ Fulltime Equivalent (FTE) Distribution …………… 231 A-3 Riverwood Owner Personels’ Roles and Application Experiences ....… 232 A-4 Riverwood Consultants’ and Builder’s Roles and Application
Experiences ……………………………………………………………… 233
xvi
LIST OF ILLUSTRATIONS
Figure Page 1-1 Research contribution to dynamics knowledge-flow theory by
amplifying tacit knowledge from individual to inter-organization (Adapted from Nissen 2002) ……………………………………………. 14
2-1 Financing and architectural-engineering-construction interdependencies
during a facility development life cycle ………………………………… 56 2-2 Multiple concurrent and sequential phases in a typical facility
development life cycle with a different organization in each phase ……. 62 3-1 Multiple concurrent and sequential phases in a typical facility
development life cycle with a different organization in each phase (Adapted from Ibrahim and Paulson (2005), Fig. 2-2) …………………. 101
4-1 The organization and workflow for the hypothetical concept design
project …………………………………………………………………… 141 5-1 A Knowledge Group Set (KGS) unit consisting of two or more
horizontal workflow processes with task interdependency links during knowledge life cycle period (Adapted and revised from Nissen’s Horizontal and Vertical Processes Model (2002, Fig. 3)) ……………….. 184
5-2 The Knowledge Group Set (KGS) Flow Model during a facility
development process …………………………………………………….. 187 5-3 Network diagram of the concurrent finance phase in Baseline Model ….. 191 5-4 Network diagram of the sequential feasibility, entitlements, and building permit phases representing the pre-construction stage in
Baseline Model …………………………………………………………... 192 A-1 Riverwood Apartments affordable housing development program in
SimVision® ……………………………………………………………. 225 A-2 Test cases for sequential knowledge flows ……………………..………. 226 A-3 Workflow process and organization for feasibility-entitlements phase of
Riverwood Apartments …………………………………………………. 227
xvii
A-4 Workflow process and organization for building permit phase of Riverwood Apartments …………………………………………………. 228
A-5 Workflow process and organization for development project financing
phase of Riverwood Apartments.………………………………………… 229 A-6 Workflow process and organization without knowledge network ……… 246 A-7 Workflow process and organization with knowledge network …………. 247
xviii
CHAPTER 1
INTRODUCTION
1.1 OVERVIEW
Knowledge flows (Nissen and Levitt 2004) enable workflows, and hence are essential to
organizational performance wherever knowledge and information work are involved. My
dissertation seeks to understand how established organization theory and emerging
knowledge flow dynamics theory can be extended to inform the design of enterprises
with discontinuous membership. Its goal is amplifying knowledge flows to enhance the
enterprise’s performance. My main research question is:
RQ. How do knowledge flows impact the organizational performance of
enterprises with discontinuous membership?
Chapter 1 introduces my doctoral dissertation in general. Part 1 of this chapter
presents the problem statement, the background literature, and my point of departure. Part
2 describes the cross-disciplinary mixed-method case-study research design. Part 3
describes my claims of contributions, the predicted impacts of my dissertation, and
suggested future research. Part 4 includes a reader’s guide to the dissertation.
1
PART I:
BACKGROUND INTRODUCTION
1.2 STATEMENT OF PROBLEM
Knowledge flows (Nissen and Levitt 2004) enable workflows, and hence are essential to
organizational performance wherever knowledge and information work are involved.
Nissen and Levitt define knowledge flows as the movement of knowledge from one
person, organization, location or time of application to another. They describe how rapid
and efficient knowledge flows are critical to organizational performance. But tacit
knowledge does not flow well through the enterprise. Tacit knowledge is deeply rooted in
action, commitment, and involvement in a specific context (Polanyi 1967; Nonaka 1994).
In contrast, explicit knowledge refers to knowledge that is transmittable in formal,
systematic language. Flows of tacit knowledge attenuate particularly quickly in the
architectural-engineering-construction (AEC) context, in part because organizations
experience discontinuous participation. Although the topic of enterprise knowledge flow
is attracting increasing attention from scholars (Nissen and Levitt 2004; Grant 1996),
there is a dearth of literature on how an enterprise transfers its knowledge (Alavi and
Leidner 2001). Moreover, the literature also lacks studies about the interaction between
organizational environment and knowledge flow. Organization scholars who developed
and enhanced contingency theory—such as Lawrence and Lorsch (1967), Galbraith (1974
and 1977), and Burton and Obel (2003)—argued that the organizational environment is
likely to play a major role and is a key factor in design, but they have yet to link
2
knowledge flow as one of the environmental contingency factors that could influence the
organization’s performance.
Definitions of key terms
Before proceeding further, I would like to define several key terms I will use in this
dissertation. Knowledge is a set of commitments and beliefs of its holder that enables the
holder to undertake certain actions (Nonaka 1994). The criterion for using the term
‘knowledge’ in my dissertation is its enabling action entity that allows the holder of a
knowledge entity to undertake certain action. The following terminologies follow this
definition.
Amplify- the extension of reach (Nonaka 1994) by sharing knowledge among individuals,
among groups, within organization, and between organizations.
Data- facts that an individual or enterprise can use to analyze or make a decision.
Discontinuous membership- the operational situation of an organization where a position
in an organizational structure is added or deleted while the process is on-going.
For example, the mechanical engineer is not required during the feasibility-
entitlements phase because the design team has to prepare only conceptual design
that does not require mechanical feedback. However, the design development
contract requires the architect to integrate building systems in the building permit
submission. Hence, the architect would add a mechanical engineer to his design
team. This mechanical engineer’s position is in a state of discontinuous
3
membership because of the different needs and goals of the conceptual design
versus the detailed design phases of the facility development life cycle process.
Discontinuous organization- an entity comprising several agents, where one or more of
its members has discontinuous membership.
Enterprise- a group of organizations which is responsible for accomplishing a certain
work process.
Explicit knowledge- the selected and applicable group of facts that is transmittable in a
formal systematic language that enables its holder to take some actions to
complete a task. For example, the fact about a requirement to preserve the oak
grove on a building site is documented in writing in the development permit. The
mechanical engineer does not normally see a copy of the development permit. His
required term of reference becomes explicit knowledge when it is recorded on a
drawing or in meeting minutes that the mechanical engineer is likely to see.
Hence, the mechanical engineer can utilize this selective fact to ensure he or she
reroutes the water piping system around the oak grove on the site.
Information- the selective collection of facts that an agent can use to perform a task. For
instance, the information about the oak grove preservation is available in the
development permit specifically for the project, and is public to anyone who
requested that selective fact from the authority—something most engineers do not
routinely do. So potentially useful information turns into usable explicit
knowledge when the information is grouped with other relevant information, and
it is recorded and distributed in a form and at a time that enables other agents to
perform their tasks.
4
Knowledge flow- the process of moving knowledge by way of communicating to retrieve
or allocate working information of any knowledge type (either tacit or explicit)
that would enable an individual or enterprise to complete a workflow process.
Organization- an entity that comprises a team of agents who are responsible for
accomplishing a certain work process.
Tacit knowledge- the entity of “knowing how” that an individual or an enterprise
possesses in selecting and applying a group of facts, which enables action to
complete a task (Polanyi 1967; Nonaka 1994). Professionals in the construction
industry ‘know’ the need to countercheck the current regulatory requirements
prior to any design or construction work. A mechanical engineer ‘knows’ to check
on the location of the closest water main connections from the building
department, which enables him to design the water piping routes accordingly.
Unfortunately, in one case that I observed, the information about the oak grove
preservation requirement was not included and the error was costly.
Turnover- is an operational situation where an agent of a position in an organizational
structure is replaced with another agent to fulfill the same position’s role while
the process is on-going. For example, a facility development project requires a
project manager in its organizational structure. Due to the long development
period, the agent who is performing the project manager’s role decides to leave
the facility developer’s firm for a better paying job in another firm. The project
manager’s position is fulfilled by a new agent without changing the overall
organizational structure of the development team.
5
Workflow- the sequence of completing a set of tasks, either concurrently or sequentially,
that an individual or enterprise adopts to complete a product development process.
The construction industry understands very well that incomplete knowledge flow
can cause unnecessary rework and delay (Paulson 1976; Jin and Levitt 1996). Knowledge
loss (K-loss) matters when it impacts a project’s schedule and cost. In facility
development, emerging information could force facility developers to consider
abandoning the development project when it causes irreconcilable schedule and cost
increase. I used two instances to illustrate this phenomenon. The first is about an oak
grove conservation requirement that was ‘missed’ by the mechanical engineer (mentioned
in detail in Chapters 2 and 3), and the second is about omitting a play structure required by
a funding program (mentioned in detail in Chapters 2 and 4).
My research is motivated by the need to extend the movement of tacit knowledge
from individuals to other members of a facility development team with discontinuous
membership. It is my intent to develop a flexible database system for the enterprise’s
knowledge management, which could capture both tacit and explicit knowledge during a
facility development life-cycle process to mitigate this K-loss phenomenon. The
challenge is how to develop such a user-friendly knowledge management system that
captures the inherent tacit knowledge of individuals or the enterprise. However, in order
for me to reach this higher goal, I must first understand why it is that K-loss is still
recurring despite measures by facility owners to invest in information technology to help
curb the loss.
6
This dissertation is my attempt to determine the environmental contextual causes
that could explain the K-loss phenomenon, and the extent of how knowledge flows affect
the organizational performance of an enterprise. My intuition was that K-loss led to many
potential projects being abandoned before they were constructed. Hence, I chose Yin’s
(2003) case-study research methodology because it allows me to design a research
methodology that includes a contextual understanding surrounding the unit of analysis I
studied. The research methodologies I used in my case-study are 1) ethnographic study in
the field of anthropology; 2) social network analysis in the field of sociology; and 3)
computational organization theory (COT) from engineering and computer science. I
finalized the main research question after conducting the ethnographic study, and
developed the other two sub-research questions below to guide me towards answering my
main research question afterwards. The two sub-research questions are:
Sub-RQ1. What are the operating environmental constructs that are representative
of how knowledge flows in a complex process with discontinuous membership?
Sub-RQ2. How different are the knowledge flows within a life-cycle phase of a
complex process?
The choice of tool, the objectives of why I chose the research methods, their
limitations, and the unit of analysis are presented in Part 2 of this chapter. I applied Yin’s
(2003) four-step validation process (i.e., construct, internal, external, and reliability
validations) for all the research methodologies used in the research.
7
The ethnographic study (detailed in Chapter 2) identified four operating
environment constructs for facility development. They are multiple sequential and
concurrent phases, discontinuous membership, interdependency of tasks, and different
knowledge form dominating in different phases. Results from the COT modeling—using
SimVision®—further developed a proposal for knowledge as another contingency factor
in the design of organizational fit (detailed in Chapter 5). The knowledge network
analysis using social network analysis (detailed in Chapter 3) affirmed that knowledge
flows (i.e., the communications to retrieve and allocate information) between team
members depend on knowledge type and whether the agent is present in a workflow
phase. Utilizing the results of these three studies, I proposed and Marc Ramsey developed
an extension of the Virtual Design Team (VDT) computational organization simulation
framework (Jin and Levitt 1996) to simulate the impacts of knowledge networks on
organizational performance in an enterprise with discontinuous membership (detailed in
Chapter 4). The cross-disciplinary mixed-method case-study research concludes that
discontinuity in organizations is a factor in the K-loss phenomenon in facility
development, and knowledge type and non-hierarchical information-processing of the
enterprise affects its organizational performance.
The rigor of attempting to answer a cross-disciplinary research question allows
me to cross two knowledge domains, i.e., dynamic knowledge flows and organization
theory; which will benefit the facility development domain. This dissertation will become
a stepping stone to many diverse areas of study, specifically in the theoretical and
application aspects of knowledge-based design of organizations operating in dynamic
environments. It is my hope that the construction industry can benefit when facility
8
owners have higher success rate of completing their facility projects from inception to
completion. Most importantly, facility owners will benefit from long-term continuity of
knowledge within their enterprises.
1.3 MOTIVATING BACKGROUND PROBLEM
My main research question motivated me to review literature from the knowledge flows
dynamics and organization theories. Much is presented in the following Chapters 2, 3, 4
and 5. In this introductory chapter, I would like to present the motivating background
problem that initiated my research interest, which is the inability to standardize the
facility development process in order to mitigate K-loss. Why? The reader can refer to
Chapter 2 for detailed literature review on the real estate and facility development in
general. I found real estate scholars—e.g., Fulton (1999), Peiser and Schwanke (1992),
Bookout (1990), Kone (1994), Schmitz (2000)—would make recommendations that their
facility development guidelines are general, but insist that each facility developer needs
to know the locality well before proceeding any further. In affordable housing
development projects, finance regulations added complexity to the already ambiguous
life-cycle process. Hence, no one has been able to standardize the facility development
process (Carrillo et al. 2004).
In an earlier study (Ibrahim 2001), I found that the facility developers view the
facility process evolving around three major phases: pre-development, development, and
property management. Pre-development is literally a feasibility phase where the major
components are selecting a site, analyzing the site, analyzing the title, determining the
governmental requirements, creating a conceptual design, doing market analysis,
9
determining product pricing and costing, doing financial analysis, and deciding whether
or not to proceed with the project. The development phase consists of three sub-phases,
i.e., entitlements, building permit, and construction. The entitlements phase is defined as
a period during which the developer applies for the official permission to develop and
construct the facility on a property within the governing jurisdiction. The building permit
phase is defined as a period during which architects and engineers collaborate on
completing the proposed facility’s design development and construction documents to
obtain building permits from the governing authorities. The construction phase represents
the building period from when the builder starts constructing physical work on site until
the project receives its certification of fitness for occupancy.
According to the city’s point of view (Landis 2001), cities look at the
development approval process as a means to gain revenue. Landis (2001) divides the
development process into three stages in relation to how cities or counties charge the
fees. They are planning fees, building permit and plan checking fees, and capital facilities
fees. Cities charge planning fees when facility developers apply for land-use approvals.
Building permit and plan checking fees are in effect when facility developers apply for
various site preparation and architectural approvals to build one or more structures on a
site. Capital facilities fees are charged when facility developers apply for connection of
infrastructure systems and public services to the facility. Planning and building permit
fees mostly cover on-site services and documents, while capital fees generally cover off-
site improvement and services. Many planning fees are set by local planning departments,
subject to planning commission and city council (or board of supervisors) reviews.
10
Planning and processing fees are typically due at application filing and are not refundable
if or when planning approvals are not forthcoming.
On the other hand, the American Institute of Architects (AIA 1997) describes the
facility development phases according to the architect’s scope of basic services. AIA
divides the development process into schematic design, design development, contract
documents, bidding or negotiation, and construction phases. Schematic design ends at
entitlements application, typically at the application for planning approval when design
development starts. Design development ends when architects complete a preliminary
integration of structural, mechanical, and electrical requirements into their schematic
designs. Contract documents include further refinement of all aspects of the building
design that can facilitate a successful bidding and eventual construction of the design by
contractors.
I compared the three viewpoints—namely the city’s, the facility developer’s, and
the architect’s—to see whether the three parties match in their definitions of a facility
development process. Table 1-1 presents the different phases of a residential development
process. In general, the city is not concerned with whether facility developers profit from
their development projects. Facility developers are concerned about the financial
sustainability of their facilities, while architects assist the facility developers to develop
their projects. From a facility developer’s point of view, how the design proposals
advance from entitlements’ schematic drawings to construction documents is not its
concern as long as it is aware that the architect is coordinating the design and
construction documents. This is because, during the entitlements approval to construction
period, facility developers are busy lining up their permanent financing in order to close
11
their construction loans (Ibrahim 2001). This scenario hints a dual side of the
development process prior to the construction phase: the well known AEC design-
construction process versus the developer’s public and financing process. The only
period when both processes require one another’s continuous interaction is during the
entitlements phase. During the entitlements process, architects are the ones preparing or
coordinating the bulk of planning and architectural documents for entitlements approval.
The documents provide the means for developers to cost and plan their development
schedule. They use these documents to obtain construction and permanent financing. I
posit that this is the period during which most of the K-loss occurrences starts
manifesting in the real estate development process.
Table 1-1. Comparison of a Typical Residential Development Process between Landis (2001), Ibrahim (2001), and the American Institute of Architects (1997)
Landis (2001) Early Development Process
Ibrahim (2001) Development Life Cycle
AIA (1997) Basic Services
N/A Feasibility Phase Schematic Design Phase Gaining land-use approvals Entitlements Phase
Getting various site preparation and architectural approvals through building permit and plan check.
Building Permit Phase
Design Development, Construction Documents, and Bidding/Negotiation Phases
Connecting the structure to infrastructure and public services
Construction Phase Construction Phase
N/A Post-Construction Phase N/A
Burton and Obel’s (2003) contingency theory framework would categorize the
facility development life cycle operating environment as having environmental
characteristics of high complexity, high uncertainty, and high equivocality. It has high
12
complexity because, despite having a functional organizational configuration, the facility
development organization also reflects a strong matrix configuration. There are also
many interdependencies between workflow processes in a development project. A facility
development project has high uncertainty because, despite having a general sequential
development activity schedule, each project is unique. Project Managers cannot
accurately predetermine which workflow path they need to concentrate on at any given
time. The operating environment has high equivocality, because there exist multiple and
conflicting interpretations, confusion, and lack of understanding among the stakeholders.
These are apparent especially when dealing with regulatory agencies, city officials, and
the public. Chapter 2 provides detailed examples that lead to these findings. The intricate
environmental characteristics point to the need for research on how we can ensure the
transfer of decisions and information (i.e., knowledge that enables the enterprise to act)
from one team to another efficiently while the process progresses, and while facility
developers maintain their feasibility.
1.4 POINT OF DEPARTURE
The identification of a possible period—the entitlements and building permit phases
(Ibrahim 2001) where I believed K-loss starts manifesting during a facility development
life cycle, versus the design development phase (AIA 1997)— helped this dissertation to
focus on the pre-construction phase. Chapter 2 explains in detail how I came to this
conclusion. Based on this finding, I concentrated on knowledge flows dynamics theory to
set my dissertation in a broader context. This section summarizes the literature gap and
identifies how my dissertation contributes towards reducing this knowledge flow gap.
13
Figure 1-1 positions this research in its broader context. I adapted Nissen’s (2002)
Notional Knowledge-Flow three-dimensional vector representational model that enables
researchers to visualize an enterprise’s knowledge flows. The global goal of knowledge
flow is expediting individual tacit knowledge to the inter-organizational explicit
knowledge. The knowledge flow path is not a straight line, as was established by Nonaka
(1994) in his spiral model, and by Nissen’s (2002) knowledge flow trajectory. For
simplicity, the two non-linear trajectories are omitted from the figure below.
Figure 1-1. Research contribution to dynamics knowledge-flow theory by amplifying tacit knowledge from individual to inter-organization (Adapted from Nissen 2002).
Referring to Figure 1-1, the first dimension is the epistemological dimension
defining the explicitness of knowledge. Polanyi (1967) was first to look at the
epistemological definitions of knowledge and divided it into tacit and explicit. The y-axis
14
of the vector model represents the scale of explicitness of knowledge. Nonaka (1994)
developed the ontological reach when he proposed the spiral SECI trajectory
(socialization, externalization, combination, internalization) as tacit and explicit
knowledge amplified from individual to group to organization to inter-organization. The
x-axis represents the ontological reach and provides a two-dimensional view of
knowledge flow. The third dimension of knowledge flow is its life cycle.
Nissen (2002) extends Nonaka’s dynamics of knowledge flow theory by
integrating the life-cycle process of knowledge flow through the enterprise when he
amalgamated from other scholars’ knowledge life cycle steps (Davenport and Prusak
1998; Depress and Chauvel 1999). He developed a six-step life cycle process for
knowledge: 1) creation, 2) organization, 3) formalization, 4) distribution, 5) application,
and 6) evolution. In this vector model, Nissen’s fourth dimension is how the flow time
occurs for the knowledge movement. This flow time can represent the ‘stickiness’ of the
knowledge flow within the enterprise. Von Hippel (1994) coined the term ‘stickiness’ on
how a needed information can ‘stick’ with the problem-solving capabilities in a different
location. Szulanski (2000) further developed the ‘stickiness’ measure of knowledge during
its transfer process within an organization.
I adapted Nissen’s knowledge life cycle stages, but generalized the stages to
creation, sharing, and application only. The reason is that this study is an early attempt in
modeling knowledge flows using computational organization theory (COT) modeling
techniques, and I would like to concentrate on the knowledge sharing aspect of the life
cycle stages. Much knowledge transfer literature is currently in the upper end of the
15
epistemological scale—i.e., explicit—where knowledge can be stored, retrieved, and used.
It is in the lower end of the epistemological scale that my dissertation will contribute.
PART 2:
MIXED-METHOD CASE-STUDY RESEARCH METHODOLOGY
1.5 METHOD
The need to answer a cross-disciplinary research question provides an opportunity to
develop a mixed-method case-study research methodology. With a case-study approach
(Yin 2003), I can explain the contextual causes why knowledge flows impact the
organizational performance of an enterprise with discontinuous membership. My research
design combines research methodologies from the field of anthropology (archival
ethnography), sociology (knowledge network analysis), and computer science and
engineering (computational organizational theory—COT). The logic of linking data to
the propositions will be explained in Section 1.9 of this chapter and the criteria for
interpreting and validating the findings will be explained in Section 1.10. Part 2
concludes with the validation methodology and limitations of each research step.
1.6 RESEARCH QUESTIONS
I developed my main research question to bridge the theoretical gap and my preliminary
observations in order to create an understanding on the knowledge loss phenomenon that
continues recurring in the facility development domain. My main research question is:
16
RQ: How do knowledge flows impact organizational performance in enterprises
with discontinuous membership? (See Chapter 4 for the answer).
I developed two sub-research questions (sub-RQs) that guided me in developing answers
to parts of the main research question. The first two sub-research questions are:
Sub-RQ1: How does knowledge flow across the life cycle of a complex process?
(See Chapter 3 for the answer).
Sub-RQ2: What are the operating environmental constructs that are
representative of how knowledge flows in a complex process with
discontinuous membership? (See Chapters 2 and 5 for the answer).
Table 1-2 illustrates the objectives of these sub-research questions based on the
choice of research methodology I chose to seek the answers. The complex process I
studied comes from the affordable housing domain. At the enterprise level, the
dissertation first seeks to understand the facility development life cycle process from the
facility developer’s perspective. The ethnographic study concentrates on formalizing the
facility development workflow process and the organization responsible for it by
identifying the environmental operating characteristics surrounding it.
17
Table 1-2. Mixed-method Case-study Steps for Multi-disciplinary Research
STEP RESEARCHQUESTION
RESEARCH METHOD
(Tool/ Source)
UNIT OF ANALYSIS
OBJECTIVES LIMITATIONS ON CHOICE OF METHOD
1 How does K flow across the life cycle phases of a complex process?
Ethnography (archival)
Enterprise -Understand life cycle process from owner’s view. -Identify potential K-loss reasons.
Does not measure K flows.
2 What are the operating environmental constructs that are representative of how K flows in a complex process with discontinuous membership?
COT (SimVision®)
Enterprise -Develop integrated K flows and organizational environmental constructs. -Cross-validate life cycle environment in COT tool.
Only at theoretical level because data for measurements are not possible yet.
3 How different are the K flows within a life cycle phase of a complex process?
Knowledge Network Analysis (KAME)
Individual -Determine whether there are different K flow behaviors among team members.
Does not link to workflow process.
4 How do K-flows impact the organizational performance in enterprises with discontinuous membership?
COT (VDT-KN)
Project Answers research question! Only proof-of-concept. Need further studies to determine exact behavior parameters.
18
The second sub-research question utilizes a computational organizational theory
(COT) to refine the environmental operating characteristics by integrating ideas from the
knowledge flow dynamics and organization theories to explain the complex process. I
used SimVision®, an agent-based COT tool developed by Jin and Levitt (1996), for this
step. The ethnographic study provides me with qualitative results; future studies can
gather more empirical data to calibrate the model quantitatively and validate it. I chose to
replicate the operating environmental characteristics on a COT tool because it is too high
a risk to ask a facility developer to participate in this research. Moreover, the successful
completion rate of real estate projects is so low that early research work can be wasted
when a selected facility project does not proceed to construction for any reason.
The ethnographic study and initial COT modeling can only visualize a single
direction of tasks, such as in a precedence diagram. Therefore, the need to measure the
number of knowledge flows introduces the use of a traditional social network analysis
tool, which was adapted for study of communication networks, to help measure the
number of communications taking place between team members. I developed a third sub-
research question for this step:
Sub-RQ3: How different are the knowledge flows within a life cycle phase of a
complex process? (See Chapter 3 for the answer).
A Knowledge Areas Mapping Exercise (KAME) (Contractor et al. in review;
Yuan et al. in review) was utilized to determine whether there are different knowledge
flow behaviors among team members due to the dominant knowledge type (i.e.,
19
explicitness of knowledge involved in tasks) within one life cycle phase. The limitation
of this method is that its results do not give me any indication of whether the process that
the members are working on influences the organizational performance.
The results from each research step gradually built up the required parts to answer
my main research question at the end. A knowledge network behavior extension was
added to the virtual design team (VDT) computational framework (Jin and Levitt 1996)
to test a proof-of-concept model that knowledge flows within the knowledge networks of
an enterprise’s organization affect the performance of the enterprise. The extension is
called Virtual Design Tool-Knowledge Network (VDT-KN).
1.7 PROPOSITIONS
The nature of ethnographic study requires no proposition to test (Creswell 2003). Any
propositions or hypotheses developed in this dissertation came after the ethnographic
study as reflected in the development of the sub-research questions above. They are as
follows:
Knowledge Network Analysis (see details in Chapter 3)
Hypothesis 1: In discontinuous membership organizations, less expert group
members tend to retrieve information from perceived experts in
their group.
Hypothesis 2: In discontinuous membership organizations, less expert group
members will allocate information to perceived experts in their
group.
20
Hypothesis 3: In discontinuous membership organizations, members will turn to
continuous members to augment their meta-knowledge of “who
knows what.”
COT modeling with Virtual Design Team - Knowledge Network (VDT-KN) (see
details in Chapter 4)
Hypothesis 1: A continuous member in a dynamic organization, who possesses
an accurate cognition of his other team members’ knowledge
skills, will be able to utilize his knowledge network to maintain the
organization’s performance when he turns to other team members
to complement his incomplete cognitive skills.
Hypothesis 2: A discontinuous member, who does not have an accurate cognition
of his other team members’ knowledge skills, can place the
organization at risk when he turns to other team members to
complement his incomplete cognitive skills.
1.8 UNIT OF ANALYSIS
The unit of analysis of my case-study depends on the research methodology step I used in
the cross-disciplinary mixed-method case-study. I used the enterprise, a non-profit facility
developer, as my unit of analysis for the ethnographic study and the COT modeling with
SimVision® (see Chapters 2, 4, and 5). I was a participant-researcher during a summer
internship in 2002, and had continued access to the organization for the next two years.
The Chief Operating Officer (COO) was my case-study gatekeeper who authorized my
21
access to everyone and all information in the office. I obtained my data from archival
records, 1993-2002 Development Reports in the enterprise’s Board of Directors Papers,
semi-structured interviews with executives and staff, and participation in selected
meetings.
In the knowledge network analysis (details in Chapter 3), my unit of analysis is at
the individual level. Information about the team members and the knowledge areas within
a phase came from the ethnographic study. The behavior parameters for the knowledge
network came from findings of the knowledge network analysis (see Chapter 3). In the
final step using the VDT-KN extension, I used the project as my unit of analysis and
based the model used in my experiments on the findings of my ethnographic research
described in Chapter 2.
1.9 LOGIC LINKING DATA TO PROPOSITION
In this cross-disciplinary mixed-method case-study research methodology, the
development of the sub-research questions facilitates how my results from each
preceding research methodology link to inform and guide the development of the
subsequent research methodology. My data are qualitative in relation to the succeeding
research methodology because the development of specific empirical constructs were not
yet possible at the time of this dissertation. For example, the ethnographic study (in
Chapter 2) provides qualitative results about the facility development operating
environment, while the COT with SimVision® (in Chapters 4 and 5) affirms the
discontinuous membership affect qualitatively, but without any claim of quantifying the
knowledge flows, since these extensions have not yet been calibrated and validated. In
22
the knowledge network analysis (in Chapter 3), the empirical results could not provide
specific numerical measurements of organizational performance.
Although the COT modeling with VDT-KN shows the effects of knowledge flows
in an enterprise with discontinuous membership, it is premature to quantify the amount of
knowledge flow because it involves knowledge type, reach, life cycle, and flow time
(Nissen in review). I rely mainly on the functional risk index of VDT-KN’s performance
quality measures and use two other quantitative measures in an ordinal way to explain the
qualitative performance of two test cases. The Baseline model uses Galbraith’s (1974 and
1977) hierarchical information-processing only, while the K-Baseline model integrated
Wegner’s (1987) non-hierarchical transactive memory information-processing algorithm.
In the final conclusion, this dissertation illustrates a proof-of-concept that knowledge
flows do affect the organizational performance in enterprises with discontinuous
membership.
1.10 CRITERIA FOR INTERPRETING FINDINGS
I used the internal and external validation method for interpreting the findings at each
research methodology step in my mixed-method case-study. The internal validity for the
ethnographic study uses pattern matching of the development life cycle phases, the
activities, and team structure of several projects’ development process (see Chapter 2).
The results provide sources for the succeeding research methodology steps. For external
validity, I compare the results of my qualitative and quantitative findings against
established organization and knowledge flow dynamics theories. The dominant theories
are contingency theory (Burton and Obel 2003) and transactive memory theory (Wegner
23
1987). This dissertation assumes that the facility development environment should be
stable when I use the COT modeling tools, and also while the transactive memory
hypotheses are being tested. The ultimate criterion in interpreting my dissertation is the
affirmation that there are qualitative differences in organizational performance metrics
when a non-knowledge-networked Baseline model is compared to a knowledge-
networked Baseline model. Details are presented in Chapter 4.
1.11 VALIDATION
In addition, I used Yin’s (2003) four-element validation process to validate my findings
as a whole, with each research methodology step having its independent four-step
validation mechanism. The four elements of validation are construct validity, internal
validity, external validity, and reliability. See Table 1-3 for detailed sequencing of the
mixed-method’s cross-validation process. The ethnographic study (Chapter 2) feeds into
the COT model that refined the discontinuous organization’s environmental constructs.
24
Table 1-3. Validation Tests for Mixed-method Case-study for Multi-disciplinary Research
TYPES
METHODS
ETHNOGRAPHY
COT (SimVision®)
K-NETWORK ANALYSIS
COT (VDT-KN)
CONSTRUCT VALIDITY
Data collection Composition
• 13 cases • Establish chain of
evidence from archives • Key informants review
drafts
• Interview project managers of worst and best cases.
• Best case • Establish K areas from
ethnography study • Determine K mapping
participants
• Conceptual case • Use K areas and K
perceptions of project members from knowledge network analysis.
INTERNAL VALIDITY
Data Analysis • Pattern matching • Explanation building
• Worst case representation • Matrix vs. non-matrix cases
• Social network analysis
• Multiple linear regression
• Individual level
• Pattern matching with expected results.
• Establish chain of evidence
EXTERNAL VALIDITY
Research Design • Compare with Environmental Factor in Contingency theory (Burton & Obel 2003)
• Compare with California’s planning fees structure (Landis 2001)
• Integrate dynamic K flows theories in model building.
• Compare with Environmental Factor in Contingency theory (Burton & Obel 2003)
• Compare with Transactive Memory Theory (Wegner 1987)
• Compare with Collective Action Theories (Marwell & Oliver 1993; Monge & Contractor 2003)
• Integrate transactive memory theory (Wegner 1987) in VDT model.
• Compare results with expected outcomes from transactive memory theory (Wegner 1987)
RELIABILITY
Data Collection • Able to use key events and documents to gather data on remaining 60 cases
• Replicate worst and best cases while maintaining environmental factor affects
• Thomsen et al. (1999) retrospective model.
• Duplicate KAME on other facility development projects.
• Thomsen et al. (1999) intellective experiments
25
Detailed analysis of the knowledge type within two sequential phases confirms
that the knowledge flow behaviors depend on the knowledge type and nature of a
member’s discontinuous factor (see Chapter 3). The non-hierarchical information-
processing from the knowledge network analysis (in Chapter 3) was then utilized to
develop the Virtual Design Team–Knowledge Network extension (VDT-KN). When
compared to a stable organization—albeit intellectively—a discontinuous organization is
at risk of having functional failure even though the overall organizational performance
shows almost no difference. This supports facility developers’ curiosity about why their
projects keep experiencing the K-loss phenomenon when their project managers seem to
be in control of the life cycle process.
1.12 LIMITATIONS
My dissertation is limited to demonstrating a proof-of-concept model that discontinuous
membership is a cause for K-loss phenomenon in the facility development. The metrics
for knowledge flow is not yet developed, hence I use knowledge network as representing
the flow of knowledge during communications to retrieve and allocate information
among the team members. The metric results provide qualitative reflections of the micro-
behaviors of the organization and the knowledge flow within. Neither does it involve
organizational learning over time.
26
PART 3:
CONTRIBUTIONS, IMPLICATIONS, AND FUTURE RESEARCH
1.13 CLAIMED CONTRIBUTIONS
Contribution 1. Establishment of discontinuous membership in enterprises as a
contributing factor to knowledge loss in complex product development process.
The ethnography results provide rich insights into the culture and operating environment
of a facility development organization. The four operational characteristics—multiple
concurrent and sequential phases, discontinuous organizational memberships, task
interdependencies, and knowledge form—can explain the K-loss phenomenon. I posit that
the task interdependencies between multiple concurrent phases make the facility
development process seem unstable due to changing critical paths when in fact, the work
process is stable within each phase. I also posit that discontinuous membership across
project phases promotes inefficient knowledge flow even when the organization is
working on a single facility project. Knowledge loss will tend to happen, and has the
potential to worsen when knowledge flow requires members to become aware of others
in concurrent organizations within the enterprise. The study identifies discontinuous
membership as the critical factor for knowledge loss when the enterprise is working on a
complex process in an uncertain and equivocal environment.
27
Contribution 2- Merging dynamic knowledge flow theory with organization theory
in establishing knowledge as another contingency factor, discontinuity as another
structural configuration measure, and reach as another design parameter property
measure in the design of organizational fit for discontinuous organizations.
My dissertation proposes adding knowledge as the seventh contingency factor for the
design of organizational structure, where tacit and explicit are its measures. It also
proposes discontinuity as another structural configuration measure, and reach as another
design parameter property measure. Three implications of knowledge flow for
organization design demonstrate the need to include knowledge flow in the design of
future organizations. The organization’s configuration misfit—having a formal procedure
versus handling uncertainty and equivocality (Burton and Obel 2004)—suggests the need
for finding ways to design a more dynamic organization that can handle complex
information formally. In addition, a discontinuous membership configuration does not
have a specific place in the organization’s structural configuration. It cannot be structured
in a simple matrix configuration because its configuration changes at different life-cycle
phases, and it is not a wholesome ad hoc configuration either due to the need for
hierarchy. Finally, the facility development organization exhibits a combination of non-
hierarchical and hierarchical information-processing systems in handling exceptions,
while Galbraith’s information processing view of organizations suggests a single- or
multi-dimensional (e.g., functional and project) hierarchical exception handling system.
The three implications guide us to conclude that knowledge flow must be seriously
considered in organization design, hence the recommendation to further study the
inclusion of Nissen’s knowledge flow dimensions as another contingency factor.
28
Contribution 3. Application of a well-established social network analysis
methodology to study knowledge flows in an engineering problem.
I used a well-established social network analysis approach (Wasserman and Faust 1997)
to conduct a knowledge network analysis. This part of my dissertation enabled me to
work with Professors Michelle Shumate of North Dakota State University, and Noshir
Contractor of the University of Illinois Urbana-Champaign. We utilized a Knowledge
Areas Mapping Exercise (KAME), an online survey instrument designed to collect data
about knowledge flow among team participants (Contractor et al. in review; Yuan et al. in
review). I acknowledge the contribution of Chunke Su, a doctoral student of Professor
Contractor, in setting up and testing the KAME protocol into the online format, and
Professor Shumate in conducting the empirical social network analysis.
Contribution 4. Extension of Galbraith’s (1974 and 1977) information-processing
theory in the Virtual Design Team (VDT) computational organizational theory
framework to include Wegner’s (1987) transactive memory theory.
My dissertation integrated Wegner’s (1987) horizontal transactive memory information-
processing method with Galbraith’s (1974 and 1977) hierarchical information-processing
for handling exceptions in two test case models. Chapter 5 describes the development of
this extension in detail and the results. I acknowledge the contribution of Marc Ramsey
who wrote and tested the programming code in this part of my dissertation. I found that
the inaccuracy of a new member’s cognitive knowledge skill about other members—i.e.,
especially when a new person joins a team—could increase the functional risk index
metric of the organization’s performance. It supports my ethnographic observations and
29
knowledge network analysis that both hierarchical and non-hierarchical information-
processing processes are taking place in an organization.
Contribution 5. Developing a cross-disciplinary research methodology using well-
established tools in other domains in solving an engineering problem.
I drew the findings from an ethnographic study and the knowledge network analysis from
the anthropology and sociology fields respectively, using data from the affordable
housing domain, to develop an extension of a computational organization theory (COT)
model. These research methodologies are acceptable within a case-study research
methodology (Yin 2003) and are discussed in detailed in Chapters 2 to 5.
1.14 IMPLICATIONS
At the conclusion of my dissertation, several implications stem from better understanding
of knowledge flow issues in respect of current organization design. The implications are
discussed in detail at the end of Chapters 3, 4, and 5.
First, I affirmed that the facility development enterprise, which has a combination
of complex, uncertain, and equivocal environmental characteristics, will always incur
knowledge loss because of its environmental characteristics. Unless researchers
understand the impacts of discontinuous membership on the enterprise, the K-loss
phenomenon will stay with the construction industry forever.
Second, discontinuous organization is not explicitly considered in the
organizational configurations and variables laid out in Burton and Obel’s (1998)
contingency theory. Organizations with discontinuous membership fall between ad hoc
30
and matrix configurations, and without any design recommendations. Based on my
findings, I would like to propose a seventh contingency factor—knowledge—to be
included in the diagnosis and design of organizational fit (see Chapter 5). Additional
measures are reach and discontinuous for design parameter property and structural
configuration respectively.
Third, current organization design is strongly based on Galbraith’s (1974 and
1977) information-processing theory. Following the validation trajectory proposed by
Thomsen et al. (1999), I proved “intellectively”—i.e., with idealized models of work
processes and organizations—that the facility development enterprise also has a non-
hierarchical information-processing system that can affect organizational performance
(detailed in Chapter 4). Moreover, both hierarchical and non-hierarchical knowledge
flows are very much governed by the dominant knowledge type—i.e., tacit vs. explicit—
in a given project phase and whether or not the team member is “continuous”—i.e.,
present in successive workflow phases—or discontinuous (see details in Chapter 3).
1.15 PRACTICAL IMPLICATIONS
I speculate that my finding about discontinuous membership in a complex process is the
primary cause for the knowledge-loss (K-loss) phenomenon will impact many diverse
areas of study, specifically in the theoretical and application aspects of a knowledge-
based organization operating in a dynamic environment.
1.15.1 Knowledge Management and Knowledge Flows
The implications of knowledge flows on the design of organizations and enhanced
understanding of knowledge flows could lead to the development of the next contingency
31
factor for organizational structure. I foresee that further understanding of knowledge flow
can assist researchers in mitigating K-loss in better designed knowledge management
systems for dynamic organizations. This has a special relevance for organizations
operating in uncertain or equivocal environments. My findings on the discontinuous
membership factor in a complex process as the key source of K-loss can help enterprises
to identify areas of potential K-loss during their processes. I have shown that in the
facility development life cycle, the critical period for potential K-loss is during the
building permit phase when the AEC design team is pursuing formal documentation
completion, while the facility owners are still negotiating with the regulatory authority to
obtain the development approval. I posit that the entitlements phase is the period when
facility development project managers should be especially diligent in managing
knowledge flow.
1.15.2 Organization Theory
The integration of Wegner’s (1987) non-hierarchical transactive memory information-
processing approach with Galbraith’s (1974 and 1977) hierarchical information-
processing in the VDT-KN COT modeling framework illustrates that knowledge
networks and flows of knowledge must be taken into consideration during organization
design. This dissertation proposes additional contingency fit parameters—reach,
discontinuity, and knowledge—for organizational design fit. Much more studies must be
done for empirical validation of this proposal. Repeating the above, further development
in this area would one day see the establishment of the next contingency factor—
knowledge—for consideration during the design of organizational fit.
32
Two branches of organization theory that could be impacted are organizational
learning and organizational evolution. VDT-KN can be further extended to simulate
organization learning. The characteristics of discontinuous membership could be further
defined for longitudinal study on organizational evolution. The use of COT can inform
organizational evolution researchers who are dedicated to following the evolution of an
enterprise over many years. I speculate that the findings from my dissertation can help to
further understand learning and evolution for dynamic organizations operating under
uncertain and equivocal environments.
1.15.3 Construction Industry
It is my intent that the construction industry, especially the affordable housing domain,
can benefit when facility owners have higher success rates in completing their projects
from conception to construction within time and schedule. On a longer term, facility
owners can benefit from long-term sustainability of the accumulated knowledge within
their organizations. I suggest a change in the curriculum of the construction industry’s
professionals to increase awareness of cross-disciplinary functions in the industry. I also
foresee a review of the basic services contract by professionals in the industry to cater to
better understanding of the operating environment of facility owners. Any revisions to the
basic services should integrate the facility owners’ need for flexibility during pre-
construction phases and for documenting commitments made during the entitlements
phase in ways that will be communicated to all involved participants in downstream
phases.
33
1.15.4 Cross-disciplinary Research
The cross-disciplinary mixed-method case-study research methodology is an example of
how a multi-disciplinary research team can each utilize the ‘tools of their trade’ for
collaborative research in engineering. I suggest more use of mixed-method research
methodologies in order for multi-disciplinary researchers to develop new theoretical
bases for emerging problems that require a cross-disciplinary answer. My case-study
research method produces four propositions (see Table 1-4 for details).
Proposition 1. Discontinuous membership in an enterprise promotes K-loss when
a work process has multiple, interdependent, and multiple concurrent and
sequential processes that are handling varying K-types in each phase (see
Chapter 2).
Proposition 2. Knowledge is another contingency factor for organizational design
fit (see Chapter 5).
Proposition 3. Transactive Memory (TM) in project organizations is supported by
communication to allocate information, when agents who have
information from the environment on topics outside of their own areas of
expertise, determine which other agents in the network could benefit from
this information, and pass the information on. TM is also supported in
communication to retrieve information, when agents who know they could
benefit from additional experts’ information coordinate the retrieval of
information from those they perceive as being appropriately skilled
experts (see Chapter 3).
34
Table 1-4. Summary of Results from Mixed-method Case-study for Multi-disciplinary Research Research Method
Results New Propositions for Future Research
ETHNOGRAPHIC STUDY
Identification of operating environmental constructs for facility development:
1) Multiple concurrent and sequential life cycle phases 2) Discontinuous membership 3) Task interdependency 4) Knowledge form
Discontinuous membership in an enterprise promotes K-loss when a work process has multiple, interdependent, concurrent and sequential processes that are handling varying K-types in each phase (see Chapter 2).
COT (SimVision®)
Additional organizational design fit parameters: 1) Reach for contingency design parameter fit for properties
configuration. 2) Discontinuous for contingency design parameter fit for
structural configuration. 3) Knowledge for contingency factor for situation fit.
Knowledge is another contingency factor for organizational design fit (see Chapter 5).
KNOWLEDGE NETWORK ANALYSIS
Characteristics of K flows in sequential phases of a facility development life cycle:
1) In K areas where tacit K dominates, continuous experts tend to retrieve and contribute their information to and from others.
2) In K areas where tacit K dominates, discontinuous experts tend to retrieve and contribute their information to and from other team members whom they perceive to be more continuous.
3) In K areas where explicit K dominates, functional experts tend to retrieve and contribute information to and from other experts.
Transactive Memory for the facility development domain: Communication to Allocate Information: The process by which agents who have information from the environment, on topics outside their own areas of expertise, determine which other agents in the network could benefit from this information, and pass the information on (see Chapter 3). Communication to Retrieve Information: The process by which agents who know they could benefit from other experts’ information coordinate the retrieval of information from those they perceive as being the most appropriately skilled experts (see Chapter 3).
COT (VDT-KN)
Impacts of discontinuous membership on organizational performance: development life cycle:
1) Macro level: Improvements on simulated duration and cost, but ‘negligible’ change to total work volume.
2) Micro level: Increase of waiting period, but reduction in coordination period.
Knowledge networks among team members can promote non-hierarchical knowledge flows that improve the organizational performance of an enterprise with discontinuous membership. But the success of such knowledge flows depends on the accuracy of each team member’s meta-knowledge about the skills of other team members, and discontinuous membership reduces the accuracy of this meta-knowledge (see Chapter 4).
35
Proposition 4. Knowledge networks among team members can promote non-
hierarchical knowledge flows that improve the organizational performance
of an enterprise with discontinuous membership. But the success of such
knowledge flows depends on the accuracy of each team member’s meta-
knowledge about the skills of other team members, and discontinuous
membership reduces the accuracy of this meta-knowledge (see Chapter 4).
1.16 SUGGESTED FUTURE RESEARCH
I am humbled that in the course of answering my main research question, I ended up with
more questions and more possibilities for future research in different domains. The power
of generalization is shown with the potential impacts of my dissertation across several
knowledge communities. In summarizing the key research areas that could facilitate
efficient knowledge flows, I list one recommendation for each domain that I have
discussed in the previous section. I believe that these recommendations can become
catalysts for many practitioners and researchers, but I leave the creativity on how to
pursue them to the reader.
In the knowledge management and knowledge flows domain, I am suggesting a
long-term goal to develop a flexible knowledge management system that could cater to
the need of an enterprise operating in a dynamic and uncertain environment with a
discontinuous membership nature of the networked communities (See Chapters 2, 3, 4,
and 5). In the organization domain, I suggest further research on establishing knowledge
as another contingency factor that is an enabler for contingency fit in organizational
design for discontinuous organization (See Chapters 4 and 5). In the construction
36
industry, I strongly recommend the American Institute of Architects review the scope of
its basic services so it play a leading role among the AEC team in mitigating knowledge
loss (See Chapter 2).
Recommendations for Practitioners
While scholars are figuring out how to mitigate K-loss during the facility development
life cycle, the best recommendation I can provide for facility developers is to provide
project assistants to the overloaded project managers so they can concentrate on
maintaining knowledge flows through the interdependent tasks in the complex process.
At this moment, many project managers are deemed “not working” because they are
always seen to be talking too much on the phone, having expensive lunches with
important people in the city, always giving instructions to peers to expedite their tasks,
etc. The ethnographic study (in Chapter 2) explains that they are actually working, but in
the tacit domain where socialization and internalization with peers are keys to knowledge
creation and transfer. With an assistant to take care of mundane office routines, such as
filing documents and photocopying, facility developers will allow project managers to
perform their best as the facilitators of knowledge flows in discontinuous organizations.
PART 4:
READER’S GUIDE
This dissertation consists of an introduction chapter, which provides an overview of my
dissertation and contributions. There is no concluding chapter. The four succeeding
37
chapters (Chapters 2 to 5) focus on the cross-disciplinary mixed-method case-study
research methodology and the limitations of each step. Each of the four subsequent
chapters will be submitted for publication as autonomous journal articles. Below are
descriptions for each chapter in my dissertation and the contributions of various co-
authors for that paper:
Chapter 2 describes the ethnographic study to understand the facility development
life cycle from the facility owner’s point of view. This paper has been submitted to the
Project Management Journal, and is currently under review. Ibrahim is responsible for
the overall content and results of the ethnographic study. Paulson assisted in reviewing
and editing the paper.
Chapter 3 describes the knowledge network analysis using Knowledge Areas
Mapping Exercise (KAME) from the field of sociology to determine whether there are
differences in the knowledge flow behaviors among team members in a sequential facility
development phase. This paper is targeted for submission to the Management Science
journal. Ibrahim is responsible for the overall paper, utilization of results from the
ethnographic study for the Knowledge Asset Mapping Exercise (KAME) exercise, and
the discussion on knowledge flow behaviors in an organization with discontinuous
membership. The KAME survey instrument was developed by Shumate with the
assistance of Chunke Su, a research assistant of Contractor. Shumate also performed the
statistical analysis on the data. Levitt assisted in reviewing and editing the paper.
Contractor provided access and resource of the KAME website at the University of
Illinois, Urbana-Champaign. Both Shumate and Contractor helped to determine the
individual level of analysis (betweeness, indegree, and outdegree of the social network’s
38
centrality measure) in studying knowledge flows in sequential phases. Development of
the KAME protocol started in February 2004 by Ibrahim and Shumate. Chunke Su
implemented the survey protocol into an online format in March 2004. Ibrahim
conducted the survey from April to June 2004. Shumate and Ibrahim analyzed the data in
July and August 2004.
Chapter 4 describes the extension of the Virtual Design Team (VDT) tool to
include Wegner’s (1987) transactive memory non-hierarchical information-processing
capability. This paper is targeted for submission to the Computational and Mathematical
Organization Theory journal. Ibrahim is responsible for the overall content of the paper.
The VDT extension code was written and tested by Ramsey. Levitt coordinated the
integration of the transactive memory behavior parameters into the VDT extension,
besides reviewing and editing the paper.
Chapter 5 describes the integration of dynamic knowledge flows and organization
theories in a COT tool for studying the organizational performance of a discontinuous
membership organization. Ibrahim is responsible for its overall content. Nissen assisted
in reviewing and editing the paper. This paper is targeted for submission to the
Knowledge Management Research and Practice journal. This is an extended version of
an earlier paper published in the Proceedings of the Thirty-Eighth Annual Hawaii
International Conference on System Sciences in Hawaii, January 3-6, 2005.
1.18 REFERENCES
Alavi, M., and D. E. Leidner. 2001. Review: Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly 25 (1): 107-136.
39
American Institute of Architects. 1997. Abbreviated standard form of agreement between owner and architect. AIA document B151-1997. New York: The American Institute of Architects.
Bookout, L. W. Jr. 1990. Residential development handbook. Washington, D.C.: ULI-
Urban Land Institutes. Burton, Richard M., and B. Obel. 2003. Strategic organizational diagnosis and design:
Developing theory for application. Boston: Kluwer Academic Publishers. Carillo, P., H. Robinson, A. Al-Ghassani, and A. Chimay. 2004. Knowledge management
in UK: Construction, strategies, resources and barriers. Project Management Journal 35 (1): 46-56.
Contractor, N., D. Brandon, M. Huang, E. T. Palazzolo, D. Steinley, C. Su, V. R. Suri, S.
A. Swarbrick, J. Templin, and S. Wasserman (in review). Multi-theoretical multilevel models of information retrieval in knowledge networks.
Cresswell, J. W. 2003. Research design: Qualitative, quantitative, and mixed methods
approach. Thousand Oaks: Sage Publications, Inc. Davenport, T. H., and L. Prusak. 1998. Successful knowledge management projects. Sloan
Management Review 37 (4): 53-65. Depress C., and D. Chauvel. 1999. Mastering information management: Part six-
knowledge management. Financial Times (8 March): 4-6. Fulton, W. 1999. Guide to California planning. Point Arena, CA: Solano Press Books. Galbraith, J. R. 1974. Organization design: An information processing view. Interfaces 4
(3): 28-36. Galbraith, J. R. 1977. Organization design. Reading, Massachusetts: Addison-Wesley. Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management
Journal 17 (Special Issue: Knowledge and the Firm): 109-122. Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for
affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Jin, Y., and R. E. Levitt. 1996. The virtual design team: A computational model of
project organizations. Computational and Mathematical Organization Theory 2 (3): 171-196.
40
Kone, L. D. 1994. Land development. Washington, D.C.: National Association of Home Builders.
Landis, J. 2001. Pay to play: Residential development fees in California cities and
counties, 1999. Department of Housing and Community Development, Division of Housing Policy Development, California.
Lawrence, P. R., and J. W. Lorsch 1967. Organization and environment: Managing
differentiation and integration. Boston: Graduate School of Business Administration, Harvard University.
Marwell, G., and P. Oliver. 1993. The critical mass in collective action: A micro-social
theory. Cambridge, UK: Cambridge University Press. Monge, P. E., and N. S. Contractor. 2003. Theories of communication networks. Oxford,
UK: Oxford University Press. Nissen, M. E. 2002. An extended model of knowledge-flow dynamics. Communications
of the Association for Information Systems 8: 251-266. Nissen, M. E., and R. E. Levitt. 2004. Agent-based modeling of knowledge dynamics.
Knowledge Management Research and Practice 2 (3). Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization
Science 5 (1): 14-37. Paulson, B. C., Jr. 1976. Designing to reduce construction cost. Journal of the
Construction Division 102 (CO4), Proc. Paper 12600, December: 587-592. Peiser, R. B., and D. Schwanke. 1992. Professional real estate development- The ULI
guide to the business. Washington, D.C.: Dearborn Financial Publishing, Inc. and ULI-The Urban Land Institute.
Polanyi, M. 1967. The tacit dimension. London: Routledge and Keoan Paul. Schmitz, A. 2000. Multifamily housing development handbook. Washington, D.C.: ULI-
the Urban Land Institute. Szulanski, G. 2000. The process of knowledge transfer: A diachronic analysis of
stickiness. Organizational Behavior and Human Decision Processes 82 (1): 9-27. Thomsen, J., R. E. Levitt, J. C. Kunz, C. I. Nass, and D. B. Fridsma. 1999. A trajectory
for validating computational emulation models of organizations. Journal of Computational & Mathematical Organization Theory 5 (4) Dec: 385-401.
41
Von Hippel, E. 1994. "Sticky information" and the locus of problem solving: Implications for innovation. Management Science 40 (4): 429-439.
Wasserman, S., and K. Faust. 1999. Social network analysis: Methods and applications.
Cambridge: Cambridge University Press. Wegner, D. M. 1987. Transactive memory: A contemporary analysis of the group mind.
Edited by B. Mullen, and G. R. Goethals. Theories of Group Behavior. New York: Springer-Verlag, 185-208.
Yin, R. K. 2003. Case-study research: Design and methods. Thousand Oaks: Sage
Publications, Inc. Yuan, Y., J. Fulk, M. Shumate, P. Monge, J. A. Bryant, and M. Matsaganis (in review).
Individual participation in organizational information commons: the impact of team level social influence and technology-specific competence.
42
CHAPTER 2
DISCONTINUITY IN ORGANIZATIONS:
HOW ENVIRONMENTAL CHARACTERISTICS CONTRIBUTE TO
THE PROJECT’S KNOWLEDGE LOSS PHENOMENON
Rahinah Ibrahim1 and Boyd C. Paulson, Jr.2
2.1 ABSTRACT
We conducted an ethnographic study of a large affordable housing organization to
understand why knowledge loss (K-loss) continues recurring during facility development.
We found four operational characteristics—multiple concurrent and sequential phases,
discontinuous organizational memberships, task interdependencies, and knowledge
form—which explain the environmental characteristics that contribute to the K-loss
phenomenon. The finding suggests a study of organizational performance, which must
include emerging dynamic knowledge flow theories with well established organization
theories. Further application of this study will lead to the development of a knowledge
management system that caters to dynamic or temporal teams within larger enterprises.
Key Words: Organization discontinuity, knowledge flows, environmental contingencies,
process design, organization design.
1 Stanford University. 2 Ibid.
43
2.2 INTRODUCTION
The construction industry understands very well that incomplete knowledge transfer can
cause unnecessary rework and delay (Paulson 1976; Jin and Levitt 1996). For instance, a
facility developer agreed to preserve an oak grove at one corner of a property as one of the
development approval terms with a city council. Several months down the facility
development process, his building permit was rejected (hence, a six-month delay) because
his mechanical engineer submitted a building plan that routed the water supply piping
system through this oak grove. The mechanical engineer, who was not aware of the
preservation commitment, located the piping route in that corner because it was the
location for all major water intake points to the site. In another instance, facility
developers lost valuable operating revenues for ‘forgetting’ to deliver an agreed item. In
this case, a funding program required a play structure in an affordable family housing
project. As the design progressed, the play structure became a flat playground area. A few
years after the project completion, the funding agency fined the developer for not
providing the play structure. It also requested the property developer to build a new play
structure or return the fund to the agency. Why do these problems occur, even when the
facility development organization has explicit information or maintains one project
manager throughout the facility development life-cycle process? It negates Cohen and
Levinthal’s (1990) absorptive capacity theory that an organization’s knowledge is built
upon its prior knowledge because somehow that knowledge is missing from the
organization.
Knowledge loss (K-loss) matters when it impacts a project’s schedule and cost.
We define knowledge as a set of commitments and beliefs of its holder that enable the
44
holder to undertake certain action (Nonaka 1994). The criterion for using the term
‘knowledge’ in this paper is its enabling action entity that allows the beholder of a
knowledge entity to undertake certain action. Explicit knowledge is the selected and
applicable group of facts that is transmittable in a formal systematic language that
enables its beholder to take some action to complete a task; and tacit knowledge is the
entity of “knowing how” that an individual or an enterprise possesses in selecting and
applying a group of facts that enables action to complete a task (Polanyi 1967; Nonaka
1994). On the other hand, information is the selective collection of facts that an agent can
use to perform a task, while data are facts that an individual or enterprise can use to
analyze or make a decision.
A grave consequence of incomplete knowledge transfer is the abandonment of a
product development project when emerging information can turn it infeasible. Emerging
information can complicate the process, extend the delivery schedule, or increase the cost
so much that a product development project is no longer worth pursuing. Our research is
motivated by the need to facilitate the transfer of tacit knowledge among members of a
facility development team with discontinuous memberships. Discontinuous membership
is an organizational operational situation where team members can join and leave the
organization to perform their specific roles while the workflow process continues. To
guide the reader, we use the term organization to represent an entity that comprises of
several team members to work on a process. We use the term enterprise to represent an
entity that comprises several organizations to complete a process. A process can be
sequential or concurrent.
45
Our main research goal is to develop a flexible knowledge management system
that captures both tacit and explicit knowledge during a facility development life cycle
process. The challenge is how to develop such a user-friendly knowledge management
system that captures the inherent tacit knowledge of individuals or the enterprise. Since
many facility developers complain about recurring K-loss despite their diligent measures,
our study turned to ethnographic method to understand the operating environment culture
that surrounds the formal textbook process. We were encouraged to pursue in this
direction when scholars such as Schreiber and Carley (2003) acknowledged that among
barriers in knowledge transfer in an organization are: not knowing which members have
the desired knowledge, not knowing whether they exist, and not knowing what
knowledge they hold. Their study identified two data types—task and referential—and
determined how they are different. They described task data as a purely technical process
whereby a member queries the database and obtains the results. On the other hand,
referential data is a social process facilitated by technology. A fundamental study by
Nonaka (1994) posits that many employees tend to seek knowledge from individual
experts on a personal basis (i.e., socialization to transform tacit knowledge to explicit
knowledge among individuals), but the organizations in Schreiber and Carley’s (2003)
study use information technology to facilitate knowledge transfer. Neither study
integrates the transfer of individual’s nor repository’s knowledge based on the work
process where the employees are involved. However, Schreiber and Carley (2003)
highlight the need to understand task complexities that an organization faces.
In order to provide a platform to capture tacit knowledge in a facility development
process, understanding the environmental culture of the facility developer’s organization
46
is critical. We conducted an ethnographic study (Spradley 1980) within a case study
research (Yin 2003) at an affordable housing development organization to find out
whether or not there was something amiss from facility planning textbooks that could
explain this recurring K-loss phenomenon. This study presents the results of the
ethnographic study. We present the basic premises of our ethnographic study, how we
determined the facility development and financing milestones, and how we developed the
life cycle phases from the developer’s perspective. Then, we present the analysis on how
the four environmental themes—i.e., concurrent and sequential phases, task
interdependencies, discontinuous membership, and knowledge forms—contribute
towards the K-loss phenomenon, and its validation approach. We conclude the paper with
a discussion on how the findings will impact organization and knowledge flow theories,
and recommendations for future studies.
2.3 POINT OF DEPARTURE
Unlike a case-study (Yin 2003) and grounded theory (Glaser and Strauss 1999), an
ethnographic study does not require a theoretical point of departure (Spradley 1980).
Instead, we refer to real estate development scholars to provide background information
for our observations. Scholars such as Fulton (1999), Peiser and Schwanke (1992),
Bookout (1990), Kone (1994), Schmitz (2000), etc. provide practical guides to land
development processes and procedures for real estate development in the United States.
However, Carrillo et al. (2004) found in a survey across major UK construction
organizations that the lack of standard work processes is the main barrier to
implementing a good knowledge management strategy in the construction industry. The
47
finding reflects an unsuccessful knowledge management strategy to fulfill the need to
share tacit knowledge of key employees in these construction organizations. Their
conclusion correlates with the real estate development scholars’ recommendations that
their guidelines are general and each facility development project needs to know the
locality well before proceeding any further. In affordable housing development, finance
adds complexity to the already ambiguous life cycle process. Other non-profit groups,
such as Neighborhood Reinvestment Corporation (NRC 1994) and California
Redevelopment Association (CRA 1998), and governmental agencies (such as the US
Housing and Urban Development Department (HUD)) assist potential non-profit
developers by providing guidelines on available financing programs.
At the American Society of Civil Engineers’ 4th International Joint Symposium on
Information Technology (IT) in Civil Engineering, participants discussed the need to
understand the economic impacts of IT usage in civil engineering. Participants asked for
IT tools that create better, more task-oriented views of complex project information that
involves different interfaces for viewing data. The organizer (Garrett et al. 2004)
concluded that there is still much challenging research to be done to bring emerging,
cost-effective information and communication technology to civil engineering practice.
Our main research contributes in fulfilling this need.
48
2.4 ETHNOGRAPHY RESEARCH METHOD
2.4.1 Overview
The essential core of ethnography is the concern for the meaning of actions and events to
the people we seek to understand (Spradley 1980). We wanted to study the project
managers involved in the facility development for they would make constant use of these
complex meaning systems to organize their behavior, to understand themselves and
others, and to make sense out of the world in which they work. The purpose of our
ethnographic study is to know whether there is something unique going on in the facility
development life cycle process that makes exception handling due to missing explicit
information a common phenomenon. We focused on the workflow processes and the
people responsible for these processes to provide us insights into the cultural and
operating environment of a facility development enterprise. Our ethnographic research
question is, what is the operating environment of a facility development enterprise?
2.4.2 Data collection
Our unit of analysis is one of the major affordable housing developers in the San
Francisco Bay Area. The affordable housing developer has completed 73 projects in the
Bay Area, and was managing about 5,000 units of affordable housing in year 2002. When
the first author became a summer intern at its central office in 2002, there were seven
project managers handling fourteen projects at various stages of the facility development
life cycle. She was a participant-observer during the initial three months data collection
period, and had continued access to the organization for the next two years as an
observer. She reported to the chief operating officer, the gatekeeper, who gave her access
to documents and human resources in the office. The major sources of data are archival
49
documents of 73 projects, and interviews with selected executives and staff. All
interviews were manually recorded and transcribed before the end of the day. The first
author met weekly with the chief operating officer to present her analysis based on the
interviews and document search of the previous week. It was during these meetings that
the first author decided whether she needed to adjust or redirect the research schedule for
the following weeks in view of emerging discoveries. The chief operating officer would
accommodate the changes accordingly.
2.4.3 Data Analysis
The study selected thirteen cases in the facility developer’s archive, which have almost
complete information (see Table 2-1). There are eight family housing projects, two single
residency occupation (SRO) projects, two senior housing projects, and one special needs
project. The SRO projects provide small efficiency studio units primarily designed to
accommodate the needs of one person. A special needs project is typically intended to
provide apartments for developmentally disabled adults who can be self-sufficient with a
little help from social service providers. The size of the affordable housing projects
ranges from 28 to 148 units. The developer completed them from 1992 to 2003. We
charted the major milestone dates for these cases and identified the most constant events
in terms of their sequencing and occurrences. A common iteration is charting the events,
then reconfirming them with the responsible project manager or staff member who had
participated in the project, and finally revising the chart. We compared the major event
milestones among the cases. Upon affirmation of the major development milestones by
the staff, we selected two cases—worst case versus best case—to study the organizational
changes that occurred during the facility development life cycle until the first year of the
50
facility’s operation. We used SimVision®, an agent-based computational organization
theory (COT) modeling tool (Jin and Levitt 1996), to build a high-level facility
development life cycle process which includes an organization responsible over each life
cycle phase to emulate the operating environment, hence validating the ethnographic
findings. Details of how we built and tested these models are published separately
(Ibrahim and Nissen 2004 and 2005).
Table 2-1. List of Affordable Housing Development Cases
Case No.
Year Completed
Number of Units
Housing Type
1 2001 43 Family 2 2001 148 Family 3 2001 74 Seniors 4 2000 70 Family 5 2001 80 Special Needs 6 2001 28 Family 7 2003 71 Family 8 2003 148 SRO 9 1994 121 SRO 10 1995 98 Family 11 1992 107 Seniors 12 1993 64 Family 13 1995 76 Family
2.4.4 Limitations of study
This study concentrates on the pre-operation phases of a facility development. We chose
affordable housing facilities because the process is richer in context, and more complex
due to its financing requirements. We limited the scope of this study to (1) determining
the milestones for a facility development life cycle, (2) understanding how the
organization evolved during its life cycle, and (3) identifying the operational environment
characteristics caused by the life cycle process.
51
2.4.5 Validation
We validated the ethnographic findings using Yin’s (2003) case study validation
approach. They are 1) establishing construct and internal validity, 2) establishing
external validation, and 3) using a computational organization theory (COT) model. In
establishing construct and internal validity (Yin 2003), we obtained affirmation of results
by key informants such as the project managers and the chief operating officer when they
reviewed the draft of the findings. We compared the general findings with existing
organization theories for external validation, specifically Burton and Obel’s (2003)
contingency theory. The contingency theory describes the operating environment as high
in uncertainty, high in complexity, and high in equivocality. The results are validated if
they correspond positively.
2.5 DETERMINING AFFORDABLE HOUSING DEVELOPMENT
MILESTONES
A facility development project starts when a developer becomes seriously interested in a
property and ends when the facility starts its operation. The first step in our ethnographic
study was determining the architectural-engineering-construction milestones during the
facility development period. We explain why we chose these time and cost milestones
below.
Architect’s first sketch submission or architect’s first fee proposal. These two
items are the first formal indications that the developer began to seriously consider
developing a property. The architect usually would sketch out the site plan to find out
how many units were feasible so the project managers can conduct an initial feasibility
52
analysis. Another alternative event is when the developer responded to a city’s request for
proposal (RFP). The developer usually pays the architects to prepare conceptual designs
to complement its submission. Similarly acceptable was the architect’s initial fee
proposal that covered a conceptual design not meant for submission purposes. Purchase
of property was not a good starting point because the closing month varied from project
to project due to terms and conditions stipulated in the purchaser agreements. The
developer’s normal practice is to request the architects to submit the formal professional
contract after it is comfortable with the property’s feasibility analysis. The initial
architect’s professional fee contract proposal may or may not include consulting
engineers’ scope of work. In most cases, contracts for the architects came many months
after they made their first sketches.
Application for development permit. This event marks a significant commitment
by the developer to obtain a formal permission—an entitlement or a development
permit—from the governing authority to develop an affordable housing project. We note
that the development application date is a little peculiar because authority submissions
varied depending on their requirements. It depended on whether or not the developer had
to submit for planned development approval. A developer submits for a planned
development approval when it intends to construct a facility that is beyond the standard
zoning regulation for that property. For example, a developer intends to build a five-unit
housing block instead of the allowable four units. Until today, most cities and counties
do not have standard ‘entitlements’ processes, a common alternative name which real
estate professionals use for this formal development approval process.
53
Receipt of the development permit. This date marks the developer’s success in the
entitlements process. It is the most consistent achievement of the entitlements process.
Upon receipt of the development permit, major activities take place since the affordable
housing projects had overcome any public opposition and could then move forward with
financing arrangements and finalizing the closure of the property’s purchase.
Receipt of the building permit or signing of building contract. At this point, all
construction documents (i.e., plans and specifications) were completed and the general
contractors were ready to start construction. The study did not choose the date for
building permit submission because this date varies among the housing projects. Some
housing projects received their building permits within a month of getting their
development approval while some took more than twelve months. However, within a
month of receiving the building permit, construction work would commence on site. We
also noted that the site handing-over in the contracts normally occurred several weeks
after signing the building contracts. When the building contracts are missing from the
files, there are occasional notices to the respective builders to commence work at site.
54
Receipt of certificate of fitness for occupancy (CFO). Prior to tenants move-in, all
facilities on the properties must have their certificates of fitness for occupancy (CFO)
from the governing authorities. All permanent funding programs require proof of
construction completion before disbursing their permanent loans. With a CFO, a project
was deemed complete from the perspective of design-construction staff. However, to
project managers, the development processes completed only when they received the
Internal Revenue Service’s (IRS) Form 8609 for the partnership company. The
application for this form required a copy of the CFO, but due to its late delivery—usually
within six to twelve months of tenants’ move-in dates—this date was not a good choice.
For 1992-1996 housing projects where no CFOs were available, the study found the CFO
dates on the properties’ Form 8609 certificates. The Form 8609 is the acknowledgement
from the State of California that the project will receive tax-credit exemptions for being
an affordable housing property.
55
Figure 2-1. Financing and architectural-engineering-construction interdependencies
during a facility development life cycle.
56
Figure 2-1 illustrates the sequence of major milestones during a facility
development life cycle. The sequential events are: identifying a potential site, applying
for a development permit, obtaining a development permit, starting construction, and
completing construction. Figure 2-1 also charts when the project managers assume the
completion of the standard American Institute of Architects (AIA) scope of architect’s
basic services (American Institute of Architects 1997) with our ethnographic
observations. There is no discrepancy in the schematic design and construction phases
among the Architecture-Engineering-Construction (AEC) and the Owner teams.
However, there are overlapping ambiguities between the times the Owner submitted the
development permit application until the start of a project’s construction. The project
managers viewed the design development at about sixty percent completion, and at the
most at seventy percent completion, when the projects received their development
permits. In addition, they only accepted the completion of the construction documents at
the time of the building permit application after they had an independent construction
estimator evaluated the project construction cost. The developer was not willing to risk
applying for insufficient permanent financing (details in the following section) when
bidding or negotiation may take too long for its financing application purposes. The
study notes these vague design development and construction documents phases for
possible explanation of the K-loss phenomenon later.
2.6 DETERMINING AFFORDABLE HOUSING FINANCING MILESTONES
Financing requirements make an affordable housing facility development more complex
than a similar for-profit’s. Affordable housing finance requires a combination of equity
57
partnership programs with multiple permanent soft and hard loans (see Table 2-2), most
of which require time-consuming applications and approval processes by government
agencies. The study found some cost estimates that reflected approximately the time
frame of architectural-engineering-construction events from documents that project
managers produced for financing applications. The developer submitted these documents
to obtain funding for the housing projects. Among the documents were:
Affordable Housing Program (AHP), Rental Housing Construction Program
(RHCP), or Redevelopment Agency (RDA) Applications. Developers can submit AHP and
RDA applications any time during the development process, provided they receive some
confirmation from the governing city or county. From 1990 until 1995, RHCP was
dominant. Cities or counties allocated and administered these funds. The developer
would submit its applications at about the same time that it submitted the projects’
entitlements applications. By this time, architects had already completed the schematic
designs. The developer usually asked for increments whenever its development cost
exceeded its earlier budget. Sometimes, it applied for additional funding more than once.
Most cities and counties were flexible on this because they did not want the affordable
housing projects to fail. Hence, there was no risk for the developer to apply before it
received the development approvals. AHP, RHCP, and RDA funds generally subsidized
the developer’s early development costs such as land leases, consultant fees, legal fees,
off-site infrastructure costs, etc.
58
Table 2-2. Examples of Equity Investment Programs, Permanent Soft and Hard Loans for Affordable Housing Development (CRA 1998)
ACRONYM
FUNDING PROGRAM
PURPOSE
Equity Investment Program LIHTC Low Income Housing Tax
Credit Allows investors in qualified low-income rental housing developments to receive a “tax credit” against their federal income tax liability for a period of 10 years. Equity investors usually become majority limited partners.
Permanent Soft Loans RDA Redevelopment Agency Funds Permanent soft loan provider. A public body
created to designate redevelopment project areas, supervise and coordinate planning of a project area and implement the development program. In all but 14 communities in California, the agency is composed of the governing body of the community (city council or board of supervisors).
AHP The Affordable Housing Program
Federal grant which provides subsidies to assist financial institutions in supporting the creation and preservation of housing for lower income families and individuals of its members affiliates. Subsidies are awarded to qualified projects submitted by members and selected through funding competitions held by each financial institution.
CDBG Community Development Block Grant
Under Title I of the Housing and Community Development Act of 1974, communities of over 50,000 people are entitled to receive direct federal funding to encourage more broadly conceived community development projects and expand housing opportunities for low- and moderate-income persons.
Permanent Hard Loans CHFA California Housing Finance
Agency A California state agency which provides below market interest rate financing for the development of affordable single-family (owner-occupied) and multifamily housing.
59
California Low-Income Housing Tax-credit (LIHTC) Application. Unlike the
AHP and RHCP applications, the LIHTC applications were more stringent. The points
systems applied to rank competing projects depended on the readiness of developers to
start construction after they received funding allocation. The LIHTC funding awards
were one-time only awards. Therefore, the developer must obtain its best development
estimate to finalize the tax-credit amount it was applying for. Hence, the developer
tended to wait until it received the building approvals and complete negotiation of the
building contracts based on final construction documents. However, if the project did not
get the funds, developers had to wait for the next application round and delay any
construction on the site. LIHTC applications are bi-annual: March and July.
Final LIHTC Cost Report. At the end of the project, an appointed company
auditor audited the project’s total development costs and issued final cost certificates to
the affordable housing developer. The final amounts recorded in these certificates were
the basis to close tax-credit partnerships for the projects. Closing of these partnerships
meant that full payment of the permanent funding for projects was available to pay off
the respective projects’ construction loans. In cases where the final cost certificates are
not available, the study refers to the final cost reports prepared by project managers to
LIHTC or AHP upon the completion of the projects.
2.7 FACILITY DEVELOPMENT PHASES
The ethnographic architectural-engineering-construction event milestones allowed the
study to track development schedule changes through at least five sequential
development phases and at least two concurrent phases (see Figure 2-2). The sequential
60
phases are feasibility, entitlements, building permit, construction, and property
management, while the concurrent phases are development financing and asset
management.
2.7.1 Sequential facility development phases
The facility development milestones allow us to divide the sequential life cycle
process into four sequential phases. The milestones denote either the start or the end of a
phase.
Feasibility Phase. This phase is defined as the period during which the developer
ascertains whether the housing project is profitable enough to justify the risks inherent in
a facility development process (Ibrahim 2001). Among the major components the
developer considers in its due-diligence feasibility analysis are site selection, site
analysis, title analysis, governmental requirements study, product design, market
analysis, product cost estimation, and financial feasibility analysis. These lead to a
decision on whether or not the developer wants to proceed. This phase starts when the
developer becomes interested in the property, and ends when the developer formally
submits an application for the housing project’s development permit.
61
Figure 2-2. Multiple concurrent and sequential phases in a typical facility development life cycle
with a different organization in each phase.
62
The study observed the project managers and executives utilizing their well-
established contacts in seeking potential sites and financing sources, and negotiating the
entitlements process. These sources would inform the project managers of the availability
of land or finance, and advise the developer about the best way to obtain the development
permit. By the time the proposed housing development enters the planning application
stage, most of the financing sources have given positive indications that the respective
proposal has a good chance for implementation. Otherwise, the developer will abandon
the proposal or wait until a better opportunity emerges.
Entitlements Phase. This phase is defined as a period during which the developer
applies for the official permission to develop and construct the facility on a property
within the governing jurisdiction. It starts when the developer formally submitted the
development permit application and ends when it obtained the development permit.
Obtaining the development permit allows the developer to build the project on that site
according to the concept it was approved for, and it is no longer at risk of being opposed
by parties beyond the control of the developer. Financing institutions will not consider
any formal applications for a permanent or construction loan without this development
permit. During the entitlements phase, the three major components are taking control of
the land or property, obtaining the entitlements from the government authority, and
securing financial commitments (Ibrahim 2001). A financial commitment is a conditional
approval from a permanent financing institution, which is subject to receipt of all
required authority approvals prior to the finalization of any loan release.
The entitlements process varies from city to city. For the City of Palo Alto, among
the major elements in obtaining its development permit are obtaining the Planning
63
Commission approval, the Architectural Review Committee approval, and the City
Council approval. More time will be required if a housing project needs to undergo a
Planned Urban Development (PUD) or Planned Development (PD) approval process
when the developer intends to build outside the city’s standard zoning regulations.
Although this period seems more technical, the most challenging part is the exposure of
the proposed housing project to the scrutiny of its neighbors, i.e., the public. The
developer uses its best negotiation skills—maneuvering politically and publicly for social
acceptance—in order to obtain the development permit. Opponents to the proposed
housing project will use all the avenues available during the process to stop the project,
while the developer arranges for public forums and bus tours to successful affordable
housing projects to allay the public’s fears of having such a project in their backyards.
Building Permit Phase. A developer needs a building permit before it can
construct a facility on a property. The building permit phase is defined as a period during
which architects and engineers collaborate on completing the proposed facility’s design
development and construction documents to obtain building permits from the governing
authorities. This phase starts when the developer receives a project’s development
permits and continues until it receives the building permits from various building
departments allowing the developer to build the facility. The major components during
this phase are acquiring the site, obtaining the building permit, obtaining the construction
financing, selecting the builder, preparing the construction documents, and product
marketing (Ibrahim 2001). The study notes that while the developer would bid and
negotiate with builders, it concurrently worked hard to apply and negotiate for permanent
financing to support the construction loan. During this period, the developer would allow
64
the design team to collaborate in completing the design development documents and,
eventually, the construction documents. The developer will appoint new consultants as
recommended by the existing design team should the team needs their expertise.
Construction Phase. This phase represents the building period from when builders
start constructing physical work on site until the project receives its certification—
Certificate of Fitness for Occupancy (CFO)—from the authority that the facility is safe
for occupation. From the developer’s viewpoint, this period marks the culmination of all
the hard work to put together the facility design with the complex financial schemes. This
is the least eventful period to the developer’s organization if the project manager did a
good job in preparing the design and finance for the housing project. However, any
exception occurring during this period is significant as it impacts schedule and cost the
worst. In the developer’s case, the construction phase for most of the housing project
cases ranged from twelve to 24 months.
Property Management Phase. This phase begins when the housing development
goes into operation. It starts when the developer receives the CFO from the governing
authority that tenants can occupy the facility. The critical operating period starts after
construction completion until the developer has complied with all the terms and
conditions of the permanent soft and hard loans. The terms of these hard and soft loans
varies from at least fifteen years to at most about sixty years. After this initial period, the
developer has no more obligations to its creditors. In the case of for-profit developers,
they can consider obtaining maximum revenue from their facility operations. However, in
the case of affordable housing developers, many eventually initiate a refinancing scheme
65
in order to refurbish the facility while ensuring itself of operating the facility for another
long-term period.
The study notes that there is no clear demarcation when the entitlements phase
actually started among the housing case studies, but there was a definitive ending with
the issuance of development approval or permit. We note this ambiguity in Figure 2-1
when there are occasions when the developer becomes very sure of getting a proposed
housing project development approval prior to any planning application submission. An
instance is when the developer applies for a RDA fund to purchase the site. The
developer obtains this fund from the city council, i.e., the final authority giving the
development permit in the entitlements process. The same city council would not object
to the development permit when the same developer for the proposed project would later
apply for the planning approval several months into the process. We propose
standardizing this dynamic period by combining both phases into the feasibility-
entitlements phase. The chart in Figure 2-1 illustrates how we combined the two phases
into one.
2.7.2 Concurrent facility development phases
Although the ethnographic study can distinguish among the sequential facility
development life cycle phases, there are ambiguities in the developer’s organizational
structure which suggest different processes going on concurrently during the sequential
facility development life cycle phases. They are the development finance and asset
management processes. For overall congruency, the study names both processes as
phases—development finance and asset management phases. In both phases, the
66
organization maintains similar members throughout several sequential facility
development phases (details in Section 2.8).
Development Finance Phase. This phase starts as soon as a project manager
ascertains that the housing project has potential to proceed further upon the conclusion of
the due-diligence analysis. The project manager obtains the approval from the board of
directors before committing further to pay the deposit to obtain site control and to hire the
architect to prepare a conceptual proposal for the site. In the course of compiling
available finance programs to support the proposed housing project, the project manager
will come across favorable financing schemes. The project manager then designs a total
finance package in lieu of the most favorable finance program. In our study, the
developer has a well-established reputation with the Low Income Housing Tax Credit
(LIHTC) program (refer to Table 2-2). Upon receipt of the development permit, the
project manager will submit multiple applications for the various finance programs. The
project manager will obtain all the permanent soft and hard loan commitments prior to
closing the construction loan, which, in addition, requires a building permit and
ownership of the site. The finance team is responsible for approving payments to the
consultants and builder during the feasibility-entitlements phase until the end when the
housing project has closed all of its permanent soft and hard loans. Then, the finance
team will transfer the finance requirements to the asset management team for long-term
operational sustainability.
Asset Management Phase. In a larger development organization, such as the
developer in our study, an asset management team is created by staff from the property
management team when the number of units in the housing portfolio grows, and many
67
earlier projects have reached maturity with their long-term finance programs. Unlike
property management, which oversees the operation of the portfolio, the asset
management team acts as the owner representative throughout the development life cycle.
Closing the land acquisition is its first task in a new housing development, while
purchasing the limited partner equity share is first in older facility portfolios. A
shareholder buy-out provides an opportunity to the developer either to run the property as
is, to rehabilitate the property, or to sell the property to a third party. The developer
mainly decides to rehabilitate these properties by upgrading the facility to meet current
tenants’ needs. In doing so, the asset management team more often works closely with
the design consultants and builder, and not the development project managers, to
refinance the property from similar finance programs available. Sometimes, the role of
the asset management staff is similar to the role of the project managers. They are
initially involved in older properties that require large capital improvement programs
before handing the projects to the development project managers. However, this
rehabilitation fund is usually a small fraction compared to new development funding
requirements. In addition, due to its role in overseeing capital improvement projects, the
asset management staff does in some way oversee all major capital improvement projects
by the property management’s maintenance staff.
The study notes that the development finance and asset management tasks fall
among optional additional services architects could provide to the owners for additional
fees. The development organization we observed had its own team of permanent staff,
besides it is very much involved in the operational aspects of most of its rental properties.
68
2.8 THE EVOLVING ORGANIZATION
Once we understood the sequential and concurrent phases of the facility
development life cycle, we determined the organizational structure responsible for each
phase. We obtained the project manager’s input in charting the estimated full-time-
equivalent (FTE) of each team member on a selected project. In the best case example,
there are a total of six organizations over the facility life cycle. Referring to Table 2-3,
the feasibility-entitlements phase has twelve members, the building permit phase has ten
members, the construction phase has twelve members, the property management phase
has nine members, and the development finance has seven members, while the asset
management has four members. The study finds that the developer’s organization
depends on external consultants assisting its staff during the sequential feasibility-
entitlements, building permit, and construction phases, while having full-time staff of its
own during the property management phase. On the other hand, the concurrent phases—
i.e., development finance and asset management—has about the same number of internal
and external members. Among the reasons behind these organizational changes are that
the developer’s staff is “not expert in technical matters,” the developer “…. cannot afford
to be liable for technical matters,” it “…hire(s) as and when we need the expertise like
preparing EIA report,” and it “…cannot afford to pay the consultants on a retainer basis.”
In Table 2-3, a full-time-equivalent (FTE) value corresponds to an eight-hour day in a 5-
day work week. The project manager for the best case study estimated the FTE
contribution by each member of his development team based on his knowledge of how
many projects he assumed each member would handle during each phase. For short-term
consultants, such as the title company or the environmental engineer, their FTEs are 1.0
69
because the project manager assumed that their staff will concentrate on one project
within a short period of time to complete a task. The project manager, for example, was
handling three to four housing projects at various development phases throughout the best
case’s development life cycle.
Table 2-3. Staff’s Position and Contributing Fulltime Equivalent (FTE) Allocations for Different Facility Development Life Cycle Phase in an Affordable Housing Case-Study
Agent’s Position Phase
FE BP CO PM DPF AM
OWNER Executive Director 0.40 0.20 0.20 0.20 0.20 0.20 Project Manager 0.50 0.20 0.20 0.20 0.25 0.15 Services Director 0.10 0.10 Accounting Department 0.50 Chief Operating Officer 0.30 Public Relations Executive 1.00 Regional Manager 0.30 Compliance Specialist 1.00 Property Manager 0.30 Site Manager 1.00 A-E CONSULTANTS & BUILDER Title Company 1.00 Environmental Engineer 1.00 Surveyor 1.00 1.00 Architect 1.00 4.00 0.50 Civil Engineer 0.50 1.00 0.10 Landscape Architect 0.50 1.00 0.10 Geotech Engineer 1.00 Financial Consultant 1.00 1.00 1.00 General Contractor 0.10 1.00 2.00 Value Engineer 1.00 1.00 Wood Structural Engineer 0.25 0.10 Concrete Structural Engineer 0.25 0.10 MEP Engineer 0.50 0.10 3rd Party Inspector 0.10 Geotech Inspector 0.10 Legal Advisor 0.50 0.15 Auditor 1.00 Note: FE = Feasibility-Entitlements; BP = Building Permit; CO = Construction; PM = Property Management; DPF = Development Project Finance; AM = Asset Management; 1FTE = 8-hour per day.
70
2.9 OPERATIONAL CONSTRUCTS
Based on the ethnographic study results, we developed four operating environment
constructs for the facility development life cycle that could explain how knowledge flow
impacts organizational performance. They are multiple concurrent and sequential phases,
discontinuous membership, task interdependencies, and knowledge form. We describe
the four constructs below.
Multiple concurrent and sequential phases. The facility development process is
complex in general, but the affordable housing development process is more complex due
to the financial and regulatory constraints that state and federal programs impose on their
developments and operations (Ibrahim 2001). Our ethnographic study finds the
feasibility, entitlements, building permit, construction, and property management phases
to occur in a sequential order, while it finds development finance and asset management
to be concurrent with the sequential phases. The succeeding phase continues upon the
completion of the milestones of the preceding phase, such as the feasibility phase ended
with the submission of the planning application, while the entitlements phase commences
with the planning application submission. On the other hand, the study finds that the
project managers have to commence development finance and asset management tasks
separately upon the completion of a task in the sequential phases. For example, the
developer needs to finalize the land acquisition, i.e., prepare legal and financial
documents, while the architectural-engineering-construction team members prepare the
schematic design documents.
Discontinuous membership. The second major construct represents a unique
organizational character—a dynamic organizational structure that varies across different
71
facility development life cycle phases. The study attributes the dynamics of the evolving
organization to the requirement for different skill sets which the development team needs
in order to complete the tasks in a particular phase. It finds some team members
contributing to multiple facility development life cycle phases, but the frequency and
intensity of their participation vary across such phases (e.g., the architect is involved in
three sequential phases that require design and construction tasks). Other team members
served in only a single phase (e.g., the environmental engineer in the feasibility-
entitlements phase). For instance, the feasibility-entitlements phase has a total of twelve
members, but the building permit phase has a total of ten members. The organization
changes when five original members leave, but it obtains three different new members.
On the other hand, the development finance phase has a constant seven team members
throughout. Please refer to Table 2-3 for discontinuous membership details for each
member of the selected project.
Task interdependencies. The third major construct acknowledges that each phase
has tasks that are interdependent with those in concurrent phases. Referring to Figure 2-1,
the event milestones are the major convergent points for concurrent workflows during the
facility development life cycle process. For instance, facility developers require building
permits before starting construction, but they need to finalize the construction loan before
handing the site to the general contractor to start construction. The tasks to obtain a
building permit and handing over the site are sequential in the building permit phase
(refer Figure 2-1). However, the task to finalize the construction loan is in another
concurrent phase—i.e., the development finance.
72
Despite a risky outcome that a facility project may not see its implementation, the
study found that the project managers were not so concerned about the uncertainties and
complexity of the facility development life cycle process. Responses such as “….we will
find other sources of finance,” or “….we’ll have the designers work on the construction
documents, while we wait for the next (LIHTC) application round” are common. It
reflects a readjustment of the critical path in the overall facility development process.
Among the common causes for change in the critical path are: delay in getting the
relevant approvals, failure to obtain the applied financing, rework due to additional
requirements, etc. Despite the fact that all the project managers mentioned that there
were no two similar projects they had handled, the ethnography study allows us to link
the common workflow sequences on their selected projects. We chart a general summary
of the task interdependencies for the selected best project in Figure 2-1.
Knowledge form. The final major construct identifies two forms of knowledge
dominating during different facility development life cycle phases. Specifically, tacit
knowledge dominates during the early feasibility and entitlements phase, while explicit
knowledge is dominant during the later building permit, construction, and property
management phases. The study found that the developer project managers obtained tacit
knowledge by socializing and internalizing the actions and sayings of the local elected
officials and the public that supports them, while they ensured transfer of explicit
knowledge among the team members during the design and financing application
processes. It found that these experienced project managers are very comfortable in their
social and political operating environment which enables them to maneuver socially,
politically, and financially during the complex process to ‘smooth’ the sequence of the
73
architectural-engineering-construction process. The study observed a number of remarks
such as, “…I’ll call so-and-so at the city hall to find out what’s going on,” or “….please
arrange a lunch meeting with so-and-so so (that) I can clarify the details….” Explicit
knowledge flows are represented by the sharing of documents each team member passes
on to others to complete their tasks.
2.10 VALIDATION
We validated the ethnographic findings by using Yin’s (2003) case study approach. First,
we established construct and internal validity. Then, we established eternal validity
followed by using a computational organization theory (COT) modeling for proof of
concept (Thomsen et al. 1999).
Constructs and Internal Validation. In establishing construct and internal validity
(Yin 2003), key informants such as the project managers and the chief operating officer
reviewed the draft of the findings. They affirmed the findings verbally. The chief
operating officer found the weekly briefings enlightening as she gradually understood the
functions and impacts of each life cycle phase on an affordable housing project. The
central office also arranged for a presentation to its central office staff during a monthly
staff meeting in late summer. About fifty executives and staff attended the presentation.
At the end of the presentation, many in the audience appreciated the interdependency
finding that explains why some of the project managers seem “pushy” on many
occasions. Despite the fact that most of the executives and staff the first author
interviewed were very dedicated to their work and the developer’s organization, they did
not quite understand why they need to worry about other tasks outside of their own
74
departments. The study observed the project managers taking the role of the main
interdepartmental coordinators throughout the facility development process, such as
calling for group meetings, carrying documents to and from the respective individuals,
and initiating one-on-one meetings when necessary. The study concludes that the main
reason is because other than the project managers, employees were unable to see the
repercussions of their tasks on another task in a different workflow process that was
handled by another department’s staff.
Another construct validation came from the executive director who exclaimed,
“That is exactly how my brain works to coordinate so many tasks!” where all the project
managers agreed amidst the laughter of others. On that day, the project managers gained
a new level of respect from their co-workers for their ability to coordinate multiple tasks.
During several follow up visits after the presentation, the first author felt the central
office’s working environment was becoming more pleasant due to better understanding
of the team members’ roles and responsibilities to one another. The developer provides a
profound validation by accepting the interdependency finding and initiated their
affordable housing development manual. The proposed manual takes into account the
interdependency requirements and long-term impacts from different department staff
members, especially with regards to those critical decisions project managers will make
during the feasibility-entitlements phase.
External Validation. We compare the general findings with existing organization
and knowledge flow theories for external validation. Burton and Obel’s (2003)
contingency theory describes the operating environment as high in complexity, high in
uncertainty, and high in equivocality. It has high complexity because, despite having a
75
functional organizational configuration, the facility development organization also
reflects a strong matrix configuration. For example, it is common for a single project
manager to handle several development projects concurrently. There are also many
interdependencies between workflow processes in a development project. For instance,
the facility development team needs to work with its finance and property management
teams internally, while working with external design consultants and regulatory agencies
to complete the development project. A facility development project has high uncertainty
because, despite having a general sequential development activity schedule, each one is
unique. Project managers cannot accurately predetermine which workflow path they need
to concentrate on at any given time. For example, facility developers cannot be sure
which program will fund a particular facility development project, and each funding
program has different requirements and application procedures. The operating
environment has high equivocality, because there exist multiple and conflicting
interpretations, confusion, and lack of understanding among the stakeholders. These are
apparent especially when dealing with regulatory agencies, city officials, and the public.
Nonaka’s (1994) SECI model—socialization, externalization, combination, and
internalization—describes the spiral process of knowledge life among a group's
interactions. The study finds different forms of knowledge dominating during different
facility development life cycle phases. Specifically, tacit knowledge dominates during the
early feasibility and entitlements phases, while explicit knowledge is dominant during the
later building permit, construction, and property management phases. Tacit knowledge
(Nonaka 1994) is rooted deeply in action, commitment, and involvement in a specific
context. As such, it can be very difficult to articulate and share, as this study found that
76
the project managers to be confident of “putting things” together. Explicit knowledge is
transmittable in formal, systematic language. As such, it can be articulated and shared via
plans, drawings, documents and databases. The study found that the developer’s project
managers obtained tacit knowledge by socializing and internalizing the actions and
sayings of the local elected officials and the public that supports them, while they ensured
transfer of explicit knowledge among the team members during the design and financing
application processes.
Computational Organization Theory (COT) Model. We cross-validated our
ethnographic findings by developing a simple model of the worst and best cases of our
ethnographic study using an agent-based tool—SimVision®—to simulate the operating
environment of a facility development life cycle. The models consist of high-level
workflow processes for the different phases and illustrate the team members responsible
for the tasks. We established the interdependency links between the various phases. The
simulation results affirm three of the four operating environment constructs: multiple
concurrent and sequential phases, discontinuous participation, and task
interdependencies. The knowledge form was not measurable using this COT tool. Please
refer to Ibrahim and Nissen (2004 and 2005), which describes how we developed the
computational models.
With the constructs internally and externally validated using Yin’s (2003) case
study method, we can now further study how knowledge flows impact organizational
performance during the facility development life cycle.
77
2.11 CONCLUSIONS AND FUTURE STUDIES
We conducted an ethnographic study of an affordable housing organization to understand
why knowledge loss (K-loss) continues to recur during the facility development life
cycle. The study found that project managers handled at least two concurrent phases (i.e.
development finance and asset management) throughout the sequential facility
development life cycle that were known to the architectural-engineering-construction
team. The study renamed the sequential phases in the facility development life cycle to
feasibility, entitlements, building permit, construction, and property management based
on the goals of the respective phases, which describe the overall workflow tasks
completed by the team responsible for each phase.
When compared to the architect’s standard basic services, the study found that the
most likely potential period for K-loss to occur was during the period when the developer
submitted the development application until the builder started construction on the
property. During the design development phase, the architect may recommend other
professionals as necessary to complete the facility design, while the developer has to
comply with various external requirements, namely public opinion, if it wants to see the
housing project obtain the development permit. The new team members may not be
aware of previously agreed decisions made by senior team members. This is especially
so if a senior team member is no longer with the team. The situation is aggravated when
goal of the design team differs from that of the developer.
Due to the professional fee constraints, most design team members aim to
complete the design development and construction documentations as soon as possible.
Unfortunately, the developer’s need for design flexibility during the entitlements phase
78
does not augur well for the functional approach of the technical team members.
Therefore, there is a high probability of inefficient knowledge movement when not every
member of the development team is aware of changes that are happening in another
concurrent phase. The phenomenon explains why the mechanical engineer missed the
preservation of the oak grove when he designed the water supply routing. He came on
board the development team during the building permit phase as recommended by the
architect during the design development phase. We recommend further studies on this
dichotomous situation where the flexible need of the developer clashes with the
functional professional process. Moreover, we recommend further studies on how to
converge the different goals in different facility development phases without
compromising the quality and increasing the cost of the overall process. We also would
like to suggest to the American Institute of Architects to evaluate whether there is need to
review the contractual obligations to be more flexible while having the ability to
financially sustain their design organizations.
The ethnographic study found a unique and dynamic organizational structure
which changes throughout the sequential major phases, while maintaining the
organizational structure in the concurrent phases throughout the facility development life
cycle process. The need for different team memberships depends on the need of different
skill sets to complete the tasks for the respective phases. There are many studies on
organization as an entity that evolves (Bacharach et al. 1996; Allmendinger and Hackman
1996; Dyck and Starke 1999), but there is no specific study of organization with
discontinuous membership during an on-going work process. Using the discontinuous
membership paradigm, we proved how the mechanical engineer made an error in routing
79
the water pipes through the oak grove. Using the multiple concurrent phases' paradigm,
our study proved how the play structure ends up as a flat playground in another K-loss
incident. It was simply that the design team was trying to minimize the construction cost
by cutting down ‘non-essential’ items from the program requirements. Since the finance
team thought that the play structure requirement had been codified (i.e., in the finance
program agreement), the developer was confident that it would be included. However, it
is common practice not to issue a copy of the developer’s financing agreement to non-
finance team members, who in this case happened to be the architectural-engineering-
construction team members.
These findings led us to suggest that new members have difficulties ‘knowing’
when and how to retrieve available information regardless of whether the existing
information is possessed within the developer’s database system. The ‘knowing’ of when
and how to use certain information is tacit knowledge to individuals, despite being
explicit knowledge in the repository. It also involves ‘who’ should actually become
aware of this information. This defeats the purpose of having a usable knowledge
management system because the knowledge management systems will act as a repository
entity only. We recommend further study into the impact of discontinuous memberships
on knowledge flows. In addition, we recommend further study on how the disruptive
knowledge flows would impact organizational performance in order for us to design a
better organization for a facility development project. Any knowledge flows study should
integrate how different knowledge forms would impact knowledge flows, especially tacit
knowledge, and eventually how they would impact organizational performance.
80
The final major finding from the ethnographic study is the existence of task
interdependencies between the work processes in different sequential and concurrent
phases. Although the COT model could not measure knowledge flows, SimVision®
allows us to view the dynamic change on the overall critical paths of the facility
development life cycle spanning over multiple concurrent and sequential phases. The
critical path changes correspond to various exceptions that the project managers
highlighted. Among the exceptions are delays in getting an approval, failure to obtain
financing, ‘missed’ information causing rework, etc. The interdependencies cause the
critical path to shift in multiple phases, but the overall work process does not change.
Only members who are aware of the overall tasks, such as the project managers, can take
mitigative actions to reduce the impacts in the relevant phase. Hence, the ethnographic
study provides an answer as to why the project managers and many scholars (such as
Carillo et al. 2004) comment that “…no (facility) project is the same.” We are
recommending further research on the critical path change due to task interdependencies
to facilitate knowledge transfer in a complex environment.
In conclusion, the ethnography results provide rich insights into the culture and
operating environment of a facility development organization. The study found that the
facility development life cycle has several consistent milestones—having a potential site,
applying for development permit, obtaining development permit, starting construction,
and completing construction. The four operational characteristics—multiple concurrent
and sequential phases, discontinuous organizational memberships, task
interdependencies, and knowledge form—can explain the K-loss phenomenon. We posit
that the task interdependencies between multiple concurrent phases make the facility
81
development process seems unstable due to changing critical paths when, in fact, the
work process is stable within its own phasing. We claim that discontinuous memberships
promote inefficient knowledge transfer even when the organization is working on a single
facility project. K-loss will tend to happen, and has the potential to worsen when
knowledge transfer requires members to be aware of others in the organization. Our
study suggests future study of organizational performance, which must include emerging
dynamic knowledge flow theories with well established organization theories.
Specifically, the study recommends further investigation into dynamic knowledge flow
theories to understand how an organization can benefit from efficient knowledge
transfers to improve organizational performance. Further application of this study will
lead to the development of a knowledge management system that caters to dynamic or
temporal teams within larger enterprises.
2.12 ACKNOWLEDGEMENTS
This paper is part of the first author’s doctoral research at Stanford University, which is
sponsored by the Ministry of Science, Technology, and Innovation of Malaysia in
affiliation with Universiti Putra Malaysia. Additional support was provided by the UPS
Foundation endowment at Stanford University. We acknowledge the contributions of
Professors Ray Levitt of Stanford University, Palo Alto; and Mark Nissen of Naval
Postgraduate School, Monterey, California. We extend a special appreciation to Fran
Wagstaff, Anna Kramer, Kevin Brown, and Susan Russell for their assistance throughout
this study.
82
2.13 REFERENCES Allmendinger, J. and J. R. Hackman. 1996. Organizations in changing environments: The
case of East German symphony orchestras. Administrative Science Quarterly 41 (3): 337-369.
American Institute of Architects. 1997. Abbreviated standard form of agreement between owner and architect. AIA document B151-1997. New York: The American Institute of Architects.
Bacharach, S. B., P. Bamberger, et al. 1996. The organizational transformation process: The micropolitics of dissonance reduction and the alignment of logics of action. Administrative Science Quarterly 41 (3): 477-506.
Bookout, L. W. Jr. 1990. Residential development handbook. Washington, D.C.: ULI-Urban Land Institutes.
Burton, Richard M., and B. Obel. 2003. Strategic organizational diagnosis and design: Developing theory for application. Boston: Kluwer Academic Publishers.
California Redevelopment Association. 1998. California affordable housing handbook:
Strategies for planning and development. Sacramento, CA: California Redevelopment Association.
Carillo, P., H. Robinson, et al. 2004. Knowledge management in UK: Construction, strategies, resources and barriers. Project Management Journal 35 (1): 46-56.
Cohen, W. M., and D. A. Levinthal. 1990. Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly 35 (1, Special Issue: Technology, Organizations, and Innovation): 128-152.
Dyck, B., and F. A. Starke. 1999. The formation of breakaway organizations:
Observations and a process model. Administrative Science Quarterly 44 (4): 792-822.
Fulton, W. 1999. Guide to California planning. Point Arena, CA: Solano Press Books.
Garrett, James H., I. J. Flood, et al. 2004. Information technology in civil engineering-
Future trends. Journal of Computing in Civil Engineering 18 (3): 185-186.
Glasser, B. G., and A. L. Strauss. 1999. The discovery of grounded theory: Strategies for qualitative research. New York: Aldine De Gruyter.
83
Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Ibrahim, R., and M. E. Nissen. 2004. “Simulating environmental contingencies using SimVision®,” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pensylvannia, June 27-29 by the North American Association for Computational Social and Organization Science.
Ibrahim, R., and M. E. Nissen. 2005. “Developing a knowledge-based organizational performance model for discontinuous participatory enterprises,” Proceedings of the Thirty-Eighth Annual Hawaii International Conference on System Sciences in Hawaii, January 3-6 by College of Business Administration, University of Hawaii at Manoa.
Jin, Y., and R. E. Levitt. 1996. The virtual design team: A computational model of project organizations. Computational and Mathematical Organization Theory 2 (3): 171-196.
Kone, L. D. 1994. Land development. Washington, D.C.: National Association of Home Builders.
Neighborhood Reinvestment Corporation. 1994. How to be your own developer: Making the development decision. Washington, D.C.: Neighborhood Reinvestment Corporation.
Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization Science 5 (1): 14-37.
Paulson, B. C., Jr. 1976. Designing to reduce construction cost. Journal of the Construction Division 102 (CO4), Proc. Paper 12600, December: 587-592.
Peiser, R. B., and D. Schwanke. 1992. Professional real estate development- The ULI guide to the business. Washington, D.C.: Dearborn Financial Publishing, Inc. and ULI-The Urban Land Institute.
Polanyi, M. 1967. The tacit dimension. London: Routledge and Keoan Paul.
Schindler, M., and M. J. Eppler. 2003. Harvesting project knowledge: a review of project learning methods and success factors. International Journal of Project Management 21 (3): 219-228.
Schmitz, A. 2000. Multifamily housing development handbook. Washington, D.C.: ULI-the Urban Land Institute.
84
Schreiber, C., and K. Carley. 2003. “The impact of databases on knowledge transfer: Simulation providing theory,” Proceedings of the NAACSOS Conference 2003 in Pittsburgh, Pensylvannia, June 26-28 by the North American Association for Computational Social and Organization Science.
Spradley, J. P. 1980. Participant observation. Fort Worth: Harcourt College Publishers.
Thomsen, J., R. E. Levitt, J. C. Kunz, C. I. Nass, and D. B. Fridsma. 1999. A trajectory for validating computational emulation models of organizations. Journal of Computational & Mathematical Organiztion Theory 5 (4) Dec: 385-401.
U.S. Department of Housing and Urban Development. Available at http://www.hud.gov/.
Yin, R. K. 2003. Case study research: Design and methods. Thousand Oaks: Sage Publications, Inc.
85
CHAPTER 3
DISCONTINUITY IN ORGANIZATIONS:
KNOWLEDGE FLOW BEHAVIORS IN SEQUENTIAL
WORKFLOW PROCESSES
Rahinah Ibrahim3, Michelle Shumate4, Raymond Levitt5, and Noshir Contractor6
3.1 ABSTRACT
Maintaining product feasibility and managing knowledge flows are difficult if an
organization has to operate in an equivocal environment. We build upon an ethnographic
study and computational organization theory (COT) modeling to study how knowledge
flow—defined as allocation and retrieval of information—occur in two sequential
workflow processes in a discontinuous membership organization. Using a knowledge
mapping instrument, we collected information about knowledge retrieval, knowledge
allocation and perceived expertise during various project phases from 19 participants of
an affordable housing project. We conclude that experts who are continuous members of
a discontinuous organization facilitate the flow of knowledge in that organization. We
also find that functional experts are self-sufficient in explicit-dominant knowledge areas.
However, in tacit-dominant knowledge areas, the presence of the experts is critical for
3 Stanford University 4 North Dakota State University 5 Stanford University 6 University of Illinois Urbana-Champaign
86
knowledge flow. We propose new propositions for communication to retrieve and
allocate information for facility development organizations. The study highlights the
need to consider knowledge flows to be based on its knowledge type when designing
organizations operating in a dynamic and complex environment.
Key Words: Discontinuous membership, knowledge flows, knowledge network analysis,
functional information-processing, transactive memory, and organization design.
3.2 INTRODUCTION
After nine months of negotiations, a facility project obtained its development permit that
enabled its owner to apply for permanent financing. While the owner worked with its
finance team on the financing aspects, the architect requested that the owner expand the
design team to include new team members to help prepare the building documents for
obtaining the project’s building permit. Among the new members were a mechanical
engineer, an electrical engineer, and a concrete structural specialist. The new design team
successfully submitted the project’s building permit application after three months of
working together.
After two months of waiting, the building department informed the architect that
the building permit application was rejected because the project had ignored the oak
grove conservation requirement at one corner of the property. Apparently, the mechanical
engineer had routed a water pipe through the oak grove because it was the shortest route
from the major water main to the project. The city authority had earlier stipulated the
conservation as one of the development permit’s conditions, but that knowledge seemed
87
missing within the design team. There was no other alternative but to amend the building
documents, but among the effects of this missed implementation included the owner
postponing its financing application, delaying its bidding process, incurring additional
financing charges, and most importantly delaying its return on investment. This case
demonstrates that a facility developer may have difficulty maintaining a project’s
feasibility while it maintains knowledge flows of information during multiple processes.
Previous theory and research have addressed information retrieval and allocation
patterns among employees. Wegner’s (1987 and 1995) transactive memory theory poses
that individuals will retrieve information from and allocate information to those who they
perceive to be experts in that information area. However, previous research on
transactive memory and knowledge management fail to address the problem of
discontinuous membership illustrated above. Discontinuous membership in an
organization occurs where a position in an organizational structure is added or deleted
while the process is on-going. It differs from turnover, which occurs where the
incumbent of a position in an organizational structure is replaced with another incumbent
to fulfill the same position’s role while the process is on-going. More specifically in this
case, the team members and leaders vary during the phases of the facility development
process. Discontinuous organization refers to a group of workers who experience
structural organizational changes due to discontinuous membership of one or more of its
members while the process progresses. Organizational managers often utilize
discontinuous members because different skill sets are needed in order to complete the
tasks in different phases of a single workflow process.
88
We define knowledge as a set of commitments and beliefs of its holder that
enables the holder to undertake certain action (Nonaka 1994). The criterion for using the
term ‘knowledge’ in this paper is its enabling action property that allows the holder of a
knowledge entity to undertake certain actions. Explicit knowledge is the selected and
applicable group of facts that is transmittable in a formal systematic language that
enables its holder to take some action to complete a task; and tacit knowledge is the entity
of “knowing how” that an individual or an enterprise possesses in selecting and applying
a group of facts that enables action to complete a task (Polanyi 1967; Nonaka 1994). On
the other hand, information is the selective collection or group of facts that an agent can
use to perform a task, while data are facts that an individual or enterprise can use to
compose the information used to analyze or make a decision.
An earlier ethnographic study by Ibrahim and Paulson (2005), results described in
Section 3.3.4, found that a facility development process has multiple sequential and
concurrent phases. Each phase requires a different skill set because it has a different goal
(ibid.). The purpose of this article is to determine whether or not there are any differences
in the knowledge flow behaviors—i.e., communication to retrieve information (CRI) and
communication to allocate information (CAI)—within the facility development life cycle
due to organizational discontinuous membership. We first introduce the literature on
knowledge flows and transactive memory theory (Wegner 1987 and 1995) in
organization theory. Then we introduce the discontinuous membership patterns on the
facility management project we examined. In the remainder of this article, we report the
method, results and implications of this study.
89
3.3 LITERATURE REVIEW
3.3.1 Organization Theory
We reviewed literature on organization, but found much of it concentrates on
organization formation and behavior (March and Simon 1958; Cyert and March 1963;
Galbraith 1974; Mintzberg 1992; Scott 1998; Burton and Obel 2003). Most researchers
focus upon hierarchy as the basic structure for organizing complex social activity where
cooperation among members is achieved through vertically imposed bureaucratic
processes (Grant 1996). March and Simon (1958) identify the use of rules or program to
coordinate behavior between interdependent subtasks. Galbraith’s information-
processing model is an extension of Lawrence’s and Lorsch’s (1967 in Scott 2003)
contingency theory where the efficiency of an organization depends upon it adapting to
its environmental context. The information-processing model of organization (Galbraith
1974) proposes that decision-makers need to process information well during exception-
handling for the organization to perform well. Galbraith argues that the greater the task
uncertainty, the greater the amount of information that participants in an organization
must process. As an organization faces greater uncertainty, its members face situations
for which they have no rules. At this point, the hierarchy is employed on an exception
basis (ibid., p. 86) where lower-ranked staff would seek guidance or information from
their supervisors. Built upon Galbraith’s information-processing model of organization,
Burton and Obel (2003) extended the contingency theory and developed six contingency
factors for contingency fit during the design of organizations. They are management style,
climate, size/ownership, environment, technology, and strategy.
90
3.3.2 Knowledge Management and Dynamic Knowledge Flows Theories
Knowledge management is an increasing concern in the design of organizations. In
knowledge management literature, Alavi and Leidner (2001) note the abundance of
literature on knowledge creation, knowledge storage, and knowledge retrieval. However,
the literature concerning knowledge transfer is scarce (ibid.) and the study of knowledge
flow dynamics is recent (Nonaka 1994). Similarly, Carlile and Rebentisch (2003)
highlight that most approaches to knowledge management in organizations emphasize
storage and retrieval processes. Based on Kogut and Zander’s (1992) knowledge of the
firm theory, Nonaka (1994) supports that knowledge resides within individuals.
Therefore, he argues that organizational membership plays a critical role in articulating
and amplifying that knowledge. Kogut and Zander (1992) posit that firms are more
successful in transferring knowledge within organizations than between organizations.
Central to their argument is that although knowledge is held by individuals, it is also
expressed in regularities by which members cooperate in a social community. Nonaka
(1994) proposes four modes of knowledge transfer mechanism—socialization,
externalization, combination, and internalization (SECI)—in a dynamic spiral
epistemological relationship between tacit and explicit knowledge as it extends its
ontological reach from individual to inter-organizational.
Nissen (2002) extends Nonaka’s dynamics of knowledge flow theory by
integrating the life cycle process of knowledge flow through the enterprise: 1) creation, 2)
organization, 3) formalization, 4) distribution, 5) application, and 6) evolution. This six-
step knowledge life cycle was an amalgamation of earlier views of knowledge life cycle,
which were proposed by Davenport and Prusak (1998), and Depress and Chauvel (1999).
91
Von Hippel (1994) coined the term ‘stickiness’ on how needed info can ‘stick’ with the
problem-solving capabilities in a different location. Stickiness connotes the difficulty
experienced in the process in which an organization recreates and maintain a complex,
causally ambiguous set of routines in a new setting (Szulanski 2000). In Nissen’s later
work (under review), he states that new organizational forms may obtain and even
dominate through a focus on dynamic knowledge flows. Nissen’s work provides discrete
qualitative categories for potential operationalization of knowledge flow in an enterprise.
His four knowledge flow dimensions are type of knowledge (tacit versus explicit), level
of socialization associated with the knowledge (individual, group, organization, and inter-
organization), activities of knowledge work (create, share, apply, etc.), and flow time.
In order to understand knowledge creation by individuals, Grant (1996)
conceptualizes that the firm is an institution for integrating knowledge at the next
organization level. Grant attempts to devise mechanisms for integrating individuals’
specialized knowledge. He proposes four mechanisms to coordinate the integration of
knowledge within an enterprise: (a) having rules and directives to enable the conversion
of tacit knowledge to explicit knowledge; (b) sequencing of the workflow process that
minimizes communication, but ensures the input of an expert in a different time slot; (c)
creating routines to support complex patterns of interactions between individuals in the
absence of rules, directives, or even significant verbal communication; and (d)
establishing group problem solving and decision making. The resulting knowledge-based
firm theory has implications for the basis of organizational capability, the principles of
organization design (in particular, the analysis of hierarchy and the distribution of
decision-making authority), and the determinants of the horizontal and vertical
92
boundaries of the firm. Grant’s knowledge-based view of the firm encourages us to
perceive interdependence as an element of organizational design and the subject of
managerial choice rather than exogenously driven by the prevailing production
technology. Grant emphasizes knowledge application and the role of the individual as
the primary actor in knowledge creation and the principal repository of knowledge.
However, he points to further research need on knowledge-based theory of the firm that
will embrace knowledge creation and application.
An emerging trend among knowledge flows research is the utilization of
computational models to simulate knowledge flows (Levitt and Nissen 2002). Prior
Computational Organizational Theory (COT) research has examined the work processes
and information flows associated with project- or task-based organizations (Carley and
Prietula 1994; Levitt et al. 1994). The advantage of utilizing such computational
simulation assistance is that it can provide hypothetical data to analyze a unique
operating environment that is hard to test in real life. A study by Schreiber and Carley
(2003) encourages this research when they acknowledge that among barriers to
knowledge transfer in an organization are: 1) not knowing which members have the
desired knowledge, 2) not knowing whether they exist, and 3) not knowing what
knowledge they hold. They found that 53 percent of received answers by information
seekers contained referral information using information technology. Their study
identified two data types—task and referential—and determine how they are different.
They described task data as a purely technical process wherein a member queries the
database and obtains the results. On the other hand, referential data is obtained through a
social process facilitated by technology. While Nonaka (1994) argues that many
93
employees tend to seek knowledge from individual experts on a personal basis (i.e.,
socialization to transform tacit knowledge to explicit knowledge among individuals), the
organizations in Schreiber and Carley’s (2003) study use information technology to
facilitate knowledge transfer. However, neither study integrates the transfer of
individual’s or repository’s knowledge based on the work process in which the
employees are involved, but Schreiber and Carley (2003) do highlight the need to
understand task complexities that an organization faces. It is in this light that we seek to
utilize a knowledge network analysis tool—common in the sociological communication
field—to examine the differences, if any, between knowledge flow behaviors in different
facility development life cycle phases.
Since organization theory has emphasized vertical information-processing
structure in organization design, horizontal (or non-hierarchical) information-processing
structure for decision-making is of particular interest to us. Recent scholars such as
Lambert and Shaw (2003) have turned to Wegner’s (1987 and 1995) transactive memory
theory to explain the existence of this non-hierarchical information-processing structure
in organizations. We define knowledge flow as communication to retrieve or allocate
information between individuals in an organization. We define the communication to
retrieve and allocate information as knowledge flow because the flow of selected
information would ‘enable’ action by either the source or recipient. We would like to
explore whether transactive memory theory plays a role in knowledge flows in a
discontinuous organization.
94
3.3.3 Transactive Memory Theory
Wegner (1987 and 1995) describes transactive memory as a shared system for encoding,
storing, and retrieving information. The three key processes of a transactive memory
system are (a) directory updating, where people learn what others are likely to know; (b)
information allocation, where new information is communicated to the person whose
expertise will facilitate its storage; and (c) retrieval coordination, which is a plan for
retrieving needed information on any topic based on knowledge of the relative expertise
of the individuals in the memory system.
Wegner (1987 and 1995) and later Moreland (1999) posed that organizational
teams may act like a large brain, where individuals store information to be combined with
others’ information. When individuals in that team need information they go to the
“expert node” or expert member. When individuals gain new information related to a
particular expert area, they allocate it to an “expert node”. As individuals gain new
information through experiences about the expertise of the others in their “brain”, they
update their individual directory of where information should be and is stored. Although
the purpose of this project is to examine the knowledge flow processes in discontinuous
membership teams, we expect to uncover similar patterns posed by transactive memory
theory.
(H1): In discontinuous membership organizations, less expert group members
tend to retrieve information from perceived experts in their group.
(H2): In discontinuous membership organizations, less expert group members will
allocate information to perceived experts in their group.
95
A number of communication scholars, such as Monge and Contractor (2004);
Contractor et al. (in review); Palazzolo et al. (in review), and Yuan et al. (2005) examine
the social influence on development of knowledge networks among individuals. Other
contemporary scholars (such as Lambert and Shaw 2002; Hollingshead 1998b) have
examined whether or not a hierarchical decision-making process is applicable in today’s
fast-track project delivery environment. Lambert and Shaw (2002) merge transactive
memory theory (Wegner 1987 and 1995) with information-processing views of
organizational work processes (Galbraith 1974). Lambert and Shaw argue that in current
high-technology environments where participants have equal access to information,
organizations can minimize the need to perform hierarchical decision-making processes if
participants know from whom to seek information. While Lambert and Shaw focus on
“who knows what”, Hollingshead’s (1998b) study focuses on the function of
communication behaviors. Lambert and Shaw, and Hollingshead do not consider the
aggregate level of knowledge the enterprise possesses. In addition, Ibrahim and Nissen
(2005) point to the need to include non-hierarchical information-processing in dynamic
knowledge flows within an organization if scholars want to study performance in
contemporary organizations.
Transactive memory systems may operate more effectively in teams where
individuals have higher interdependence and have developed a more convergent
cognitive map of task-person-expertise relationships (Hollingshead 1998a). The
knowledge of “who knows what” enables peers to consult independent team members in
order to complete their own tasks, especially when their supervisor lacks the level of
96
expert knowledge needed. In a temporal organization, in particular, the supervisor
primarily acts as a bridge spanner or chief coordinator. Thus, other experts in the team
may have more specialized expertise than the supervisor. A product development team,
specifically the facility development organization, reflects a norm of discontinuous
membership while the team works on the workflow process. In this instance, the
knowledge flows may be governed by two additional mechanisms. The first mechanism
is seeking information from a mediator when one’s directory of expertise is incomplete.
Wegner (1987 and 1995) describes this process as seeking information from “who knows
who knows what.” For example, the remaining team members would know “who knows
what” and can provide the resource to newly joined team members when they seek
information. A second mechanism particular to discontinuous membership teams may be
to seek information from those who were in the previous phase of the project and to
allocate information to those who will continue in the next phase of the project. In this
mechanism, it is not proper embedding of knowledge in experts, but continuing the life of
the knowledge that is the key factor governing knowledge flow patterns.
An effective transactive memory system has several advantages for a group
process. Foremost is the expansion of an individual’s expertise when the individual gains
access to others’ domains of expertise. Another is that an individual also gains access to
new knowledge that is created through integrations occurring within the transactive
process. Integration affirms the need to have a group in the first place, showing all
members the utility of coming together to remember because the group exerts a strong
directive pressure on what is to be encoded, stored, and retrieved and places a special
premium on integrative transactions (Wegner, p. 197). The third advantage is the
97
possibility of others processing the knowledge and making decisions even when the
individual is not available. Finally, Mooreland (1999) finds that groups that develop
effective transactive memory systems can complete tasks more efficiently. It is unknown
if these advantages of effective transactive memory systems in continuous membership
organizations will apply equally to discontinuous membership organization.
3.3.4 Results from Ethnographic Study
A previous ethnographic study of an affordable housing development project was
conducted to examine knowledge flows (Ibrahim and Paulson 2005). The purpose of this
earlier study was to understand why knowledge flow problems occur despite the
developers having invested in information technology tools for knowledge management.
They found several unique environmental characteristics regarding the facility
development life cycle that may cause interruptions in knowledge flow.
Ibrahim and Paulson (2005) listed the major characteristics as 1) having multiple
concurrent and sequential workflows, 2) having discontinuous membership, 3) having
multiple task interdependencies, and 4) displaying different knowledge form (i.e., tacit- or
explicit-dominant knowledge areas). These characteristics were further refined using a
computational organization theory tool, i.e., SimVision®, to model three constructs
(Ibrahim and Nissen 2005). They are work complexity, team knowledge transaction, and
knowledge flow. The refined constructs explained the complexity of the facility
development process in general, and revealed that the affordable housing development
process is even more complex due to the financial and regulatory constraints imposed by
state and federal programs on their development and operations (Ibrahim and Paulson
2005).
98
Earlier work by Ibrahim (2001) illustrates the division of the sequential phases of
the facility development process into five phases: 1) feasibility, 2) entitlements, 3)
building permit, 4) construction, and 5) property management phases. Ibrahim and
Paulson (2005) later combined the two early phases into feasibility-entitlements phase for
better clarity. Given the discontinuous nature of this type of organization, it is unknown
what additional mechanisms will govern knowledge flow. However, as discussed
previously, it is possible that both “who knows who know what” and concern for the
remembrance of knowledge may impact discontinuous membership organizations
differently than continuous membership organizations. Therefore, we pose:
(H3): In discontinuous membership organizations, members will turn to
continuous members to augment their knowledge of “who knows what.”
The facility development life cycle displays multiple task interdependencies
between the concurrent phases. The final major unique characteristics found different
forms of knowledge dominate during different facility development life cycle phases.
Specifically, tacit knowledge dominates during the early feasibility-entitlements phase,
while explicit knowledge is dominant during the later building permit, construction, and
property management phases. Tacit knowledge is rooted deeply in action, commitment,
and involvement in a specific context (Polanyi 1967). As such it can be very difficult to
articulate and share. Explicit knowledge is transmittable in formal, systematic language.
As such it can be articulated and shared via plans, drawings, documents and databases.
Facility developers obtain tacit knowledge by socializing and internalizing the actions
99
and sayings of the local elected officials and the public that supports them. The nature of
knowledge that dominates each phase may impact knowledge flows as well. Therefore,
we ask:
(RQ1) How does the tacitness versus explicitness of knowledge affect knowledge
flows within a workflow process?
3.4 RESEARCH METHOD
The Discontinuous Organization
The discontinuous organization is an affordable housing project team, and the complex
process we selected is the project’s pre-construction process. We divided the process into
three phases (Ibrahim and Paulson 2005): feasibility-entitlements, building permit, and
project financing. The feasibility-entitlements and building permit phases form a
sequential process, while project financing is a concurrent phase to that sequential
process. We assumed the cumulative data from all three phases as belonging to a
discontinuous organization. However, for the purpose of our study, we limit our study to
data from two sequential phases, which are feasibility-entitlements and building permit,
to study knowledge flow behaviors due to tacit- and explicit-dominant knowledge type in
each phase. Figure 3-1 illustrates how Ibrahim and Paulson (2005) divide the facility life
cycle into its concurrent and sequential phases. Please refer to Appendix 1-2 for a larger
representation for a typical affordable housing project.
100
Figure 3-1. Multiple concurrent and sequential phases in a typical facility development life cycle with a different organization in each phase (Adapted from Ibrahim and Paulson (2005), Fig. 2-2).
101
Participants
The participants in the current research were all part of an affordable housing project
team in the San Francisco Bay Area. The project is a housing project with 148 single
resident occupancy units belonging to a non-profit housing developer, who is one of the
three largest affordable housing developers in that region. There were 19 members of this
project that worked in three of the five life cycle phases. Four of the members were staff
of the non-profit housing developer, while the others were appointed external consultants
in various capacities.
KAME Procedure
Data were collected using an online survey and Knowledge Asset Mapping Exercise or
KAME (Contractor, et al., in review; Yuan, et al., in review). The KAME instrument was
customized for each team based on a detailed protocol completed by the team leader, i.e.,
the project manager. (The pre-KAME interview protocol is appended in Appendix 3-1).
Responses to the protocol provided information regarding the key knowledge areas, key
tasks, and the assignment of various team members during the phases of the project. The
major knowledge areas for the feasibility-entitlements phase are architectural-
engineering-construction (AEC), development project finance (DPF), and regulatory and
authority requirements (RAR).
In the building permit phase, the major knowledge areas are AEC documentation
process and bidding process. AEC issues and regulatory and authority requirements are
major knowledge areas for the project financing phase. The KAME included questions
regarding perceived expertise, information allocation, information retrieval, and task
assignments. Network data in the KAME were collected using interactive Java applets
102
that allowed participants to indicate their network ties by clicking on images. Eighteen of
the nineteen members completed the knowledge mapping exercise (95 % response rate).
Measures
Perceived Expertise. Participants were asked to rate the amount of expertise each team
member possessed in the specified knowledge areas for the team. The response set for
this question included “expert,” “intermediate,” “beginner,” and “none.” Expert was
defined as “one of the team’s most knowledgeable people on the topic.” Intermediate was
defined as “a clear understanding of the topic.” Beginner was defined as “a basic
understanding of the topic” and none was defined as “not familiar with this topic.”
Perceived expertise was the project manager’s ranking of each of the members of the
team. This measure was utilized in order to compare results with data on members’
knowledge levels that had been collected for an earlier computational model of the same
team processes (Ibrahim and Nissen 2003 and 2004).
Continuous versus discontinuous membership. Continuous membership was
defined as being present in the previous and successive phases of the sequential process.
For instance, if a member is in the initial feasibility-entitlements phase and will be
present in the succeeding building permit phase, she or he is rated as continuous.
Otherwise, she or he is rated discontinuous. Similarly if a member is present in the later
phase and was present in the previous phase, she or he has a continuous attribute. If the
member was not present in the previous phase, then she or he has a discontinuous
attribute. Individuals’ assignment to project phases was indicated by the project
manager. Continuous membership was coded as a binary variable where one equaled
continuous membership and zero equaled discontinuous membership.
103
Information Retrieval and Information Allocation. Information Retrieval (IR) was
defined as a matrix of information retrieved for each team for each knowledge area.
Information retrieval was also measured via an online applet for each knowledge area by
which individuals were asked, “Please indicate how frequently you have retrieved
information about <Knowledge Area> from your colleagues during this project.” The
response set was a five-point Likert scale where one through five represented “none,”
“seldom,” “sometimes,” “often,” and “very often,” respectively.
Information Allocation (IA) was defined as a matrix of information allocation for
each team for each knowledge area. In these items individuals were asked, “In your
work, you may receive or create information about <Knowledge Area>. Using the
adjacent screen, please indicate how often you have provided unsolicited information
(i.e., information that you distributed that was not requested from others) about
<Knowledge Area> to your colleagues during this project.” The same response sets were
used, substituting allocation for retrieval in the response.
Analysis
In order to test the hypotheses, betweeness, indegree and outdegree centrality measures
were computed for each information retrieval and information allocation network.
Betweeness centrality is defined as “the probability that a distinct actor, i, is “involved”
(i.e., on the chosen geodesic path) in the communication between two actors”
(Wasserman and Faust 1994, p. 190). Indegree centrality is defined as the number of
communication links directed toward a single node. Outdegree centrality is defined as the
number of communication links directed away from a particular node. After each node’s
centrality score was calculated, a multiple regression analysis was conducted where
104
continuous membership and perceived level of expertise were utilized to predict the
betweeness, indegree and outdegree centrality respectively. These results were utilized to
test the hypotheses H1 to H3.
Additionally, a multiple regression quadratic assignment procedure (MRQAP), a
network analytic technique, was used to examine the research question in this project
(Krackhardt 1988). MRQAP is similar to multiple regression analysis in traditional
multivariate statistics. Like multiple regression, multiple independent variables
simultaneously predict a single dependent variable. However, in MRQAP, both the
independent variables and dependent variables are networks of relations. Additionally,
the significance of beta weights are calculated differently in MRQAP (for more on the
differences, please see Krackhardt 1988). MRQAPs were run for allocation and retrieval
networks in each knowledge area in both the feasibility-entitlements and the building
permit phases. In these MRQAPs, presence in a previous phase for building permit and
continuous presence in the feasibility-entitlements phase was entered as an independent
variable. Additionally, the agent’s expertise as rated by the team manager was entered as
an independent variable. Finally, the individual’s presence in the examined phase was
entered as a control variable.
3.5 RESULTS
Knowledge flows in discontinuous membership organizations seem to be impacted not
only by the distribution of expertise but also by members’ continuous vs. discontinuous
participation. Hypothesis 1 and 2 posed that discontinuous membership organizations
would seek information from and allocate information to those who were experts.
105
Hypothesis 3 posed that members would turn to continuous members to augment their
knowledge of “who knows what”. Table 3-1 shows the comparison of standard
coefficients of independent variables being continuous and perceived expertise against
the dependent variable for information retrieval and information allocation using the
betweeness, indegree, and outdegree measures for all three phases in the discontinuous
organization.
Table 3-1. Comparison of Standard Coefficients of Being Continuous and Having Perceived Expertise on Betweeness, Indegree, and Outdegree Centrality
for Feasibility-Entitlements and Building Permit Phases Retrieval (K-Inflow)
Allocation (K-Outflow) Variables
Betweeness In- degree
Out-degree
Betweeness In-degree Out-degree
Constant 5.152 16.431 16.097 4.971 15.428 16.736 Being Continuous
.373 ** .527 *** .464*** .386 *** .522*** .449 ***
Perceived Expertise
.123 .320*** .446*** .186 .331*** .413***
R-Squared .196 .536 .605 .250 .541 .544 F-score 10.992*** 51.922*** 69.013*** 14.98*** 53.14*** 53.69***
NOTE: N = 95, df = 2 * p < 0.05 (two-tailed) ** p < 0.01 (two-tailed) *** p < 0.001 (two-tailed)
Hypothesis 1 posed that individuals would retrieve information from more expert
members of the group. Indeed, individuals who were perceived as having higher
expertise did have higher indegree centrality (β = 0.320, p < 0.001). Therefore, H1 is
supported. However, a unique phenomenon was also discovered in these measures.
106
Expert members also tended to retrieve information from other members at a greater level
than non-expert members (β = 0.446, p < 0.001). The knowledge retrieval results
illustrate that in a discontinuous organization, while expert members would tend to wait
for lesser expert members to retrieve knowledge from them, they would also tend to seek
information from other members.
Hypothesis 2 posed that individuals would allocate information to more expert
members of the group. Again, consistent with transactive memory theory, individuals did
tend to allocate information to those with higher expertise (β = 0.331, p < 0.001).
Therefore, hypothesis 2 was supported. However, again we found an interesting
additional knowledge allocation pattern. Experts in this discontinuous organization also
tended to allocate information to others more than their less expert counterparts (β =
0.413, p < 0.001).
Hypothesis 3 posed that in a discontinuous membership organization, members
will turn to continuous members to augment their knowledge of “who knows what.”
Individuals who were continuous members did have higher betweeness centrality in both
the knowledge retrieval (β = 0.373, p < 0.01) and knowledge allocation (β = 0.386 p <
0.001) networks. Both the knowledge retrieval and allocation behaviors show that both
continuous and expert members do turn to other members in their network to augment
their knowledge of “who knows what” when their cognitive knowledge networks are
incomplete. Therefore hypothesis three was strongly supported.
The research sought to discover whether knowledge type (i.e., tacit or explicit)
could provide some additional insight into the knowledge flow patterns in discontinuous
organizations. Particularly, we were interested in the above finding that experts tended to
107
allocate and retrieve information more often than did their non-expert counterparts. We
conducted several MRQAPs for each identified knowledge areas within two independent
sequential phases. During the first feasibility-entitlements phase (see Table 3-2), the
knowledge areas are architectural-engineering-construction (AEC), development project
financing, and regulatory and authority requirements. This phase displays tacit-dominant
knowledge areas.
Table 3-2. Comparison of MRQAP Coefficients Predicting Knowledge Networks in the Feasibility-Entitlements Phase (Phase 1)
Retrieval (K-Inflow)
Allocation (K-Outflow) Variables
AEC
Dvlp. Project Finance
Reg./Auth. Requirements
AEC
Dvlp. Project Finance
Reg./Auth. Requirements
Constant .716 .553 .891 .631 .653 .176
Agent will be present in Phase 2
-0.042 .237* .322 .214 .170 -0.082
Agent is in current phase
.114 .212* .271** .153 .237* .240*
Perceived Expertise .186 .111 -0.074 .214* .091 .407
R-Squared .048 .151 .153 .127 .125 .180
NOTE: N = 342, df = 3 * p < 0.05 (two-tailed) ** p < 0.01 (two-tailed) *** p < 0.001 (two-tailed)
Ibrahim and Paulson (2005) found that the developer project managers obtained
tacit knowledge by socializing and internalizing the actions and sayings of the local
elected officials and the public that supports them, while they ensured transfer of explicit
knowledge among the team members during the design and financing application
108
processes. Their study found that these experienced project managers are very
comfortable in their social and political operating environment. This enables them to
maneuver socially, politically, and financially during the complex process to ‘smoothen’
the sequence of the architectural-engineering-construction process. The study observed a
number of remarks such as, “…I’ll call so-and-so at the city hall to find out what’s going
on,” or “….Please arrange a lunch meeting with so-and-so so (that) I can clarify the
details….”
In the feasibility-entitlement phase, team members did not tend to allocate
information to or retrieve information from those who have expertise. The only
exception to this is the AEC knowledge area, the only explicit-dominant knowledge area
in this phase. Explicit knowledge flows are represented by the sharing of documents each
team member passes on to others to complete their tasks. For instance, the structural
engineer would wait for the architect’s documents before designing the structural system
for the housing project. In that knowledge area, expert members tend to contribute
significantly in knowledge allocation, but not in its knowledge retrieval. In the building
permit phase (see Table 3-3), which Ibrahim and Paulson (2005) identified as mainly
consisting of explicit-dominant knowledge areas, the agents’ expertise was consistently
predictive of information allocation and retrieval patterns. Results show that having
perceived expertise facilitates knowledge retrieval and allocation.
Addressing RQ1, knowledge flow is influenced by the discontinuous nature of the
organization. In tacit-dominant knowledge areas such as the development project
financing and regulatory and authority requirements, neither expertise nor being
continuous predicted the information retrieval and allocation patterns. In explicit-
109
dominant knowledge areas, such as AEC documentation process and bidding process,
knowledge flow behaviors are supported by expert members.
Table 3-3. Comparison of MRQAP Coefficients Predicting Knowledge Networks in the Building Permit Phase (Phase 2)
Retrieval (K-Inflow)
Allocation (K-Outflow) Variables
AEC Documentation Process
Bidding Process
AEC Documentation Process
Bidding Process
Constant .332 .795 .502 .685
Agent is present in Phase 2
.057 -0.069 -0.008 -0.122
Agent was present in Phase 1
.025 .010 -0.003 -0.036
Perceived Expertise
.366** .249* .358*** .363**
R-Squared .130 .066 .129 .132
NOTE: N = 342, df = 3 * p < 0.05 (two-tailed) ** p < 0.01 (two-tailed) *** p < 0.001 (two-tailed)
3.6 DISCUSSION
Knowledge flow behaviors could impact the efficiency of knowledge transfer in a
complex process. There are two major findings from our study. First, is that knowledge
flow behaviors in a discontinuous organization are qualitatively different from those of a
stable organization depending on the “expertise” and “continuous” nature of the
members. Second, knowledge flow behaviors depend on the knowledge type—tacit- or
explicit-dominant—of the knowledge areas in a workflow process. In the first finding,
110
expert members would tend to wait for lesser expert members to retrieve knowledge from
them, while they would also tend to seek knowledge actively from other members during
knowledge retrieval. Likewise, the experts would also tend to wait for lesser expert
members to allocate knowledge to them, while they would also tend to provide
knowledge to other members during knowledge allocation. These knowledge flow
behaviors support Wegner’s (1987 and 1995) transactive memory theory; and they extend
Wegner’s theory for knowledge retrieval and knowledge allocation by identifying these
unique behaviors for discontinuous organization.
We revisited Ibrahim and Paulson’s (2005) ethnographic study to seek
explanation for this new observation during knowledge retrieval. The construction
industry is known to be a high risk industry where the success rate for development
projects to be completed through construction can be as low as three percent and as high
as fourteen percent. The need to support every member of the group by each independent
member is thus strong for the sake of the whole group. The external consultants’
incentive is always towards earning bigger consultancy fees should the project progress
towards construction. By the time construction starts, the consultants are entitled to
seventy percent of their professional fees; whereas they would only obtain about fifteen
percent of their fees during the feasibility phase. Given this incentive, we are not
surprised by this outcome.
Transactive memory explains that there are two sources of information people use
to decide who is to be the acknowledged location of a set of labeled knowledge in the
group (Wegner 1987 and 1995). First this is based on the individual’s personal expertise.
Second, it is determined through the circumstantial knowledge responsibility that accrues
111
as a result of how the knowledge has been encountered by the group. We find both to be
true in the facility development organization. For example, the architectural information
is encoded under the architecture label among the building submission documents.
External encoding, however, requires that the location be encoded internally with the
label, but the item itself need not even be known (ibid.). For instance, the mechanical
engineer in the design team would only need to label the location of the architectural
design information under the architecture label, but he does not need to know what the
details are about. When the mechanical engineer requires the building plan, she or he
would easily retrieve this knowledge using the ‘architecture’ label. This knowledge
retrieving behavior by the experts is encouraged by the fact that all members are at least
trained experts in their fields, so they know “who knows what” for information they
require to complete their task.
Using the same arguments, we could explain how experts would tend to allocate
knowledge more to continuous members in the discontinuous organization. Supporting
the organization’s survival for the long term benefit of the individual (Ibrahim and
Paulson 2005), expert members willingly allocate any knowledge they feel could benefit
other members in the team. Using transactive memory’s encoding process, we pose that
expert members would decide that certain knowledge is better stored by other experts,
who may retrieve and use that knowledge to benefit their professional tasks. The
incentive to do so is driven by the long term benefit to the individual experts if the group
survives. Therefore, the knowledge flow behaviors in discontinuous organization support
and extend Wegner’s (1987 and 1995) transactive memory theory.
112
The study finds continuous members do play a role in knowledge retrieval and
knowledge allocation in a discontinuous organization. Results show that members tend to
turn to both expert and continuous members in their network to augment their knowledge
of “who knows what” when their cognitive knowledge networks are incomplete. Based
on the ethnographic study by Ibrahim and Paulson (2005), we similarly argue that the
early members have big incentives to help the development project progress towards
construction. Hence, they are willingly available for knowledge retrieval, and in addition,
willingly share knowledge with new members of the organization since they want to see
that the project survives.
It is also possible that continuity elevates the status of a member in a
discontinuous organization (Thomas-Hunt et al. 2003) through the value they add in
carrying the organization’s knowledge (Cohen and Levinthal 1990) and meta-knowledge
(of who knows what) into future life cycle phases. Thomas-Hunt et al.’s (2003) study
suggests that social status can promote the differential emphasis of shared, own, and
other unique knowledge, but at the same time also the biased evaluations of member’s
knowledge and contribution. Given that discontinuity could spur knowledge transfer
through interruptions (Zellmer-Bruhn 2003) caused by organizational changes, we posit
that the absorptive capacity of continuous members actually enables the firm to ‘carry’
earlier knowledge where this added value of the continuous member helps promote the
prominence of continuous members as enablers of knowledge flows in discontinuous
organization. Further study is recommended to measure the relationship of continuity to
the prominence of a discontinuous member.
113
The second major finding is that knowledge flow behaviors depend on the
knowledge type—i.e., tacit- or explicit-dominant—of that the knowledge areas in a
workflow process. The feasibility-entitlements phase is tacit-dominant while the building
permit phase is explicit-dominant. We show that team members in a tacit-dominant
knowledge area did not tend to allocate information to or retrieve information from those
who have expertise (refer Table 3-2). The only exception to this is the AEC knowledge
area, the one explicit-dominant knowledge area in the feasibility-entitlements phase. The
building permit phase, which has mainly explicit-dominant knowledge areas, reflects
similar functional expertise knowledge flow behaviors. Here, expertise consistently
predicts information allocation and retrieval patterns. Our results show that transactive
memory does not predict knowledge flows in tacit-dominant knowledge areas.
We revert to Nonaka’s (1994) dynamic knowledge flows theory, which explains
that the transfer of tacit knowledge occurs mainly through socialization and
internalization of the “know how” by members of the organization. For knowledge to
flow well, team members have to be within the same physical location or within the
communication infrastructure of that organization (Palazollo et al. in review). This is
empirically supported by the significant results that both knowledge retrieval and
knowledge allocation for development project finance, and regulatory and authority
requirements occurred when “agent is in the current phase” during feasibility-
entitlements phase. Hence, we claim that knowledge type does influence knowledge flow
behaviors in stable organization, and transactive memory cannot support knowledge flow
in discontinuous organization in this regard. If this effect is obvious in a stable tacit-
dominant organization, the need to consider knowledge type in any knowledge
114
management design and system is even more pertinent for discontinuous organizations
with mixed knowledge types. We recommend knowledge type inclusion in future studies
on organization design and knowledge management.
There are several implications for research on dynamic knowledge flow behaviors
in organizations based on results of this study. Firstly, a complex product development
process—e.g., housing development—requires additional considerations to improve its
current knowledge management system. Zellmer-Bruhn (2003), who studied how
interruptive events affect team knowledge acquisition, provided several suggestions. We
equate “interruptive events” to an equivocal and uncertain environment from Burton and
Obel’s (2003) contingency theory. She identified that some types of interruptions may
influence knowledge acquisition effort where teams experiencing redirection and high
performance reported higher knowledge transfer effort. Redirection involves formal
planning and incorporating new members, and as such may focus teams on their
practices—to consider whether they are in need to change (ibid.). High performance may
invoke the opportunity for slack search, or it may be that high-performing teams are
compelled to actively examine their routines so they can identify what is working to
codify it for potential transfer to teams with performance problems (ibid.). In other
words, structural changes positively influenced acquisition. Based on this reason,
Zellmur-Bruhn supports the usefulness of distinguishing distinct knowledge processes
that would enhance a team’s performance if the organization changed. Hence, we also
recommend identifying the dominant knowledge type for all knowledge areas within a
complex life cycle process.
115
This study illustrates that knowledge type (Nissen in review; Ibrahim and Nissen
2005) influences team members information allocation and retrieval behaviors. We found
that in functional knowledge areas (such as architectural-engineering-construction) where
explicit knowledge dominates; knowledge flows are facilitated by experts within the
team, and having continuous membership does not influence information allocation and
retrieval behaviors. On the other hand, we found that generally in tacit-dominant
knowledge areas (such as development project financing, and regulatory and authority
requirements) there are additional factors to consider. Being a continuous member and
being present in that particular phase do seem to play some roles in information
allocation or retrieval, in addition to perceived expertise. Nissen (in review) suggests the
possibility that an organization could be designed based on its knowledge flows. Our
finding that transactive memory does not promote knowledge flow in tacit-dominant
organization supports this future area of research.
We also recommend that future research consider the product development
process together with the changing organization for each life cycle phase. Zellmer-Bruhn
(2003) indirectly points to this need when she recommends future research to examine
timing during the process to improve knowledge acquisition. We further recommend that
researchers take the opportunity during this process examination to further identify
potential knowledge loss locations during the process and/or the team member
responsible for critical tasks in these locations.
Secondly, future knowledge management systems and incentive programs for
organizational members must consider the knowledge flow behaviors caused by the
combination of different knowledge areas’ dominant knowledge type in a single life cycle
116
process. The combination of tacit- or explicit-knowledge type depends on which
knowledge area is required to perform the task. We have shown that different dominant
knowledge types have different knowledge flow characteristics. Our study does not
include organizational learning (Levinthal 1991; Levitt and March 1988). Further study is
recommended to determine the impact of knowledge type on organizational learning
processes and outcomes.
The third implication is that information-processing within an organization must
include non-hierarchical information-processing to accurately represent knowledge flows.
This finding points to the need to revisit Galbraith’s (1974) information-processing
theory of the organization, in which the vertical structural hierarchy still dominates in the
exception handling and decision-making process. Our study found non-hierarchical
communications made by agents present within the same workflow process. We believe
this non-hierarchical knowledge flow does affect organizational performance as proven
by different knowledge flow behaviors in two independent phases. Hence, we propose the
development of new hypotheses and propositions that integrate the non-hierarchical
knowledge flow with Galbraith’s (1974) information-processing theory.
This study establishes how environmental characteristics of a workflow process
influence team members’ knowledge flow behaviors via communicating to retrieve or
allocate information within an organization with discontinuous membership. We will
seek to obtain further validation of our findings by the development of a computational
organizational modeling tool to model and measure this new-found discontinuous
membership factor and how it would impact organizational performance. We also
117
recommend further studies to develop organization theories to cater to the design of
dynamic organizations that include knowledge flows and discontinuous membership.
3.7 CONCLUSIONS
This study is the first to utilize a social network analysis tool (i.e., KAME) to
study knowledge flow behaviors as a means to understand the knowledge loss
phenomenon in the construction industry. We sought to find if there were differences in
knowledge flow behaviors of a stable versus discontinuous organization. We found that
knowledge flow behaviors in a discontinuous organization are different from those of a
stable organization depending on the “expertise” and “continuous” nature of the
organization’s members. Moreover, the knowledge flow behaviors depend on the
knowledge type—i.e., tacit- or explicit-dominant—of that knowledge area in a workflow
process. The study found that the prominence of a team member depends on the
continuity attribute of that person. Our study establishes how environmental
characteristics of knowledge flow behaviors can impact how agents facilitate knowledge
retrieval and knowledge allocation within the organization. The different methodologies
of knowledge transfer due to different knowledge types affect the efficiency of
knowledge transfer in complex product development process. These findings lend further
support to the hypothesis that discontinuity in organizations is a factor contributing to the
knowledge loss phenomenon inherent in complex processes, and highlight the need to
consider knowledge flow in organization design.
118
3.8 ACKNOWLEDGEMENTS
This paper is part of the first author’s doctoral research at Stanford University,
which is sponsored by the Ministry of Science, Technology, and Innovation of Malaysia
in affiliation with Universiti Putra Malaysia. Additional support was provided by the UPS
Foundation Endowment at Stanford University. We acknowledge the contributions of
Professors Boyd C. Paulson, Jr. of Stanford University, Palo Alto, California; and Mark
E. Nissen of Naval Postgraduate School, Monterey, California. We extend a special
appreciation to Fran Wagstaff, Anna Kramer, Kevin Brown, Mara Blitzer, and Chunke
Su for their assistance throughout this study.
3.9 REFERENCES
Alavi, M. and D. E. Leidner. 2001. Review: Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly 25 (1): 107-136.
Burton, R. M., and B. Obel. 2003. Strategic organizational diagnosis and design:
Developing theory for application. Boston: Kluwer Academic Publishers. Carley, K., and M. Prietula. 1994. Computational organization theory. Hillsdale, NJ:
Erlbaum. Carlile, P. R., and E. S. Rebentisch, E. S. 2003. Into the black box: The knowledge
transformation cycle. Management Science 49 (9): 1180-1195. Contractor, N., D. Brandon, M. Huang, E. T. Palazzolo, D. Steinley, C. Su, V. R. Suri, S.
A. Swarbrick, J. Templin, and S. Wasserman (in review). Multi-theoretical multilevel models of information retrieval in knowledge networks.
Cyert, R. M., and J. G. March. 1963. A behavioral theory of the firm. Englewood Cliffs:
Prentice-Hall. Davenport, T. H., and L. Prusak. 1998. Successful knowledge management projects. Sloan
Management Review 37 (4): 53-65.
119
Depress, C., and D. Chauvel. 1999. Mastering information management: Part six- knowledge management. Financial Times (8 March): 4-6.
Dyck, B., and F. A. Starke. 1999. The formation of breakaway organizations:
observations and a process model. Administrative Science Quarterly 44 (4): 792-822.
Galbraith, J. R. 1974. Organization design: An information processing view. Interfaces 4
(3): 28-36. Galbraith, J. R. 1977. Organization design. Reading, Massachusetts: Addison-Wesley. Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management
Journal 17 (Special Issue: Knowledge and the Firm): 109-122. Hollingshead, A. B. 1998a. Cognitive interdependence and convergent expectations.
Journal of Personality and Social Psychology 81 (6): 1080-1089. Hollingshead, A. B. 1998b. Retrieval processes in transactive memory systems. Journal
of Personality and Social Psychology 74 (3): 659-671. Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for
affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Ibrahim, R., and M. E. Nissen. 2003. “Emerging technology to model dynamic
knowledge creation and flow among construction industry stakeholders during the critical feasibility-entitlements phase,” Proceedings of The ASCE 4th Joint Symposium on IT in Civil Engineering in Nashville, Tennessee, November 15-16 by the American Society of Civil Engineers.
Ibrahim, R., and M. E. Nissen. 2004. “Simulating environmental contingencies using
SimVision®,” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pennsylvania, June 27-29 by the North American Association for Computational Social and Organization Science.
Ibrahim, R., and B. C. Paulson, Jr. 2005. Discontinuity in organizations: How
environmental characteristics contribute to the project's knowledge loss phenomenon. CRGP Working Paper No. 12, Stanford University, CA.
Kogut, B., and U. Zander. 1992. Knowledge of the firm, combinative capabilities, and the
replication of technology. Organization Science 3 (3, Focused Issue: Management of Technology): 383-397.
Krackhardt, D., and J. R. Hanson. 1993. Informal networks: The company behind the
chart. Harvard Business Review 71: 104.
120
Lambert, M. H., and B. Shaw. 2002. Transactive memory and exception handling in
high-performance project teams. CIFE Technical Report 137, Stanford University.
Lawrence, P. R., and J. W. Lorsch 1967. Organization and environment: Managing
differentiation and integration. Boston: Graduate School of Business Administration, Harvard University.
Levinthal, D. A. 1991. Organizational adaptation and environmental selection-
interrelated processes of change. Organization Science 2 (1, Special Issue: Organizational Learning: Papers in Honor of (and by) James G. March): 140-145.
Levitt, B., and J. G. March. 1988. Organizational learning. Annual Review of Sociology
14: 319-340. Levitt, R. E., and M. E. Nissen. 2002. Toward simulation models of knowledge-intensive
work processes. CIFE Working Paper No. 77, Stanford University. Levitt, R. E., G. P. Cohen, J. C. Kunz, C. I. Nass, T. R. Christiansen, and Y. Jin. 1994.
The Virtual Design Team: Simulating how organization structure and information processing tools affect team performance. Edited by Carley, K. M. and M. J. Prietula. Computational and Mathematical Organization Theory. Hillsdale, NJ: Lawrence Erlbaum.
Lewin, A. Y., and H. W. Volberda. 1999. Prolegomena on coevolution: A framework for
research on strategy and new organizational forms. Organization Science 10 (5, Focused Issue: Coevolution of Strategy and New Organizational Forms): 519-534.
March, J. G., and H. A. Simon. 1958. Organizations. New York: John Wiley & Sons,
Inc. Mintzberg, H., and F. Westley. 1992. Cycles of organizational change. Strategic
Management Journal 13 (Special Issue: Fundamental Themes in Strategy Process Research): 39-59.
Monge, P. E., and N. S. Contractor. 2003. Theories of communication networks. Oxford,
UK: Oxford University Press. Moreland, R. 1999. Transactive memory: Learning who knows what in work groups and
organizations. In Levine (Ed.) Shared cognition in organizations: The management of knowledge. Hillsdale, NJ: Lawrence Erlbaum Associates.
Nissen, M. E. (in review). Understanding knowledge-flow patterns for design.
121
Nissen, M. E., and R. E. Levitt. 2002. Dynamic models of knowledge-flow dynamics. CIFE Working Paper No. 76 (November), Stanford University.
Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization
Science 5 (1): 14-37. Palazzolo, E. T., D. Serb, Y. She, C. Su, and N. S. Contractor (in review). Co-Evolution
of communications and knowledge networks as transactive memory systems: Using computational models for theoretical integration and extensions.
Polanyi, M. 1967. The tacit dimension. London, Routledge and Keoan Paul. Ren, Y. 2004. “Why and when does transactive memory matter: Opening the black box”.
Proceedings of the NAACSOS Conference 2003 in Pittsburgh, Pennsylvania, June 24-26 by the North American Association for Computational Social and Organization Science.
Schreiber, C., and K. Carley. 2003. “The impact of databases on knowledge transfer:
Simulation providing theory.” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pennsylvania, June 26-28 by the North American Association for Computational Social and Organization Science.
Scott, J. 2000. Social network analysis. Sage Publications, London. Scott, W. R. 2003. Organizations: rational, natural, and open systems. New Jersey:
Prentice Hall. Szulanski, G. 2000. The process of knowledge transfer: A diachronic analysis of
stickiness. Organizational Behavior and Human Decision Processes 82 (1): 9-27. Thomas-Hunt, M. C., T. Y. Ogden, and M. A. Neale. 2003. Who's really sharing? Effects
of social and expert status on knowledge exchange within groups. Management Science 49 (4): 464-477.
Von Hippel, E. 1994. "Sticky information" and the locus of problem solving: Implications
for innovation. Management Science 40 (4): 429-439. Wasserman, S., and K. Faust. 1999. Social network analysis: Methods and applications.
Cambridge: Cambridge University Press. Wegner, D. M. 1987. Transactive memory: A contemporary analysis of the group mind.
Edited by B. Mullen, and G. R. Goethals. Theories of group behavior. New York: Springer-Verlag, 185-208.
Wegner, D. M. (1995). A computer network model of human transactive memory. Social
Cognition, 13(3): 319-339.
122
Yuan, Y., J. Fulk, M. Shumate, P. Monge, J. A. Bryant, and M. Matsaganis (in review).
Individual participation in organizational information commons: the impact of team level social influence and technology-specific competence.
Zellmer-Bruhn, M. E. 2003. Interruptive events and team knowledge acquisition.
Management Science 49 (4): 514-528.
123
CHAPTER 4
DISCONTINUITY IN ORGANIZATIONS:
IMPACTS OF KNOWLEDGE FLOWS ON
ORGANIZATIONAL PERFORMANCE
Rahinah Ibrahim7, Raymond Levitt8, and Marc Ramsey9
4.1 ABSTRACT
Maintaining product feasibility and managing knowledge flows are difficult if a complex
process is operating under an equivocal environment and by an organization with
discontinuous membership across project phases. In this paper, we built upon findings
from an ethnographic study and data from a knowledge network analysis of an affordable
housing development project to extend an agent-based computational organization theory
(COT) modeling tool. We integrated Wegner’s transactive memory theory—that a team
member will turn to his knowledge network to augment his incomplete cognitive skills—
with Galbraith’s information-processing view of organizations to test whether or not
discontinuous membership could impact an organization’s performance through
intellective computational experiments. We found that there was no significant difference
in the total work volume of organizations with a continuous versus discontinuous team
7 Stanford University 8 Ibid. 9 Ibid.
124
member; however, the inaccuracy about other team members’ cognitive skills by the new
member could cause a detrimental effect on the functional quality of the tasks for which
she or he is responsible. The detriment of the functional quality supports our earlier
ethnographic findings that new members are at risk of becoming sources of knowledge
loss when they are not aware of prior knowledge within the organization. The study
highlights the need to consider knowledge flows when designing organizations,
especially if they operate in a dynamic and complex environment.
Key Words: Discontinuous membership, knowledge flows, organization design,
functional information-processing, transactive memory.
4.2 INTRODUCTION
During a financing program negotiation to develop a low-income family housing project,
the project developer agreed to provide a children’s play structure for the housing
complex. Somehow, as the project progressed, the play structure eventually became an
open play area by the time the property started operating. When the program sponsor
found this out a year later, it fined the property owner and requested it to return the full
amount or build the play structure. The property owner paid the fine and built the play
structure, but at its own cost which had to be diverted from other future housing projects.
This case negates Cohen and Levinthal’s (1990) absorptive capacity theory because the
prior knowledge about agreeing to build a play structure for the housing project was
missing as the housing development team continued its development.
125
The purpose of our study is to determine how knowledge flows can impact the
organizational performance of a discontinuous organization. A discontinuous
organization is a group of positions operating in an environment where one or more team
positions would join or leave the group while the work process is still on-going, due to
the need for different skill requirements to complete different parts of the process
(Ibrahim and Paulson 2005). It differs from turnover, which is an operational situation
where the incumbent of a position in an organizational structure is replaced with another
agent to fulfill the same position’s role while the process is on-going (ibid.). In affordable
housing development, the process is complex with multiple sequential and concurrent life
cycle phases. It is unique because each has its own organization overseeing the process
within one phase. The ‘non-standardization’ of the facility development process was also
a reason for the inability of experts to design a knowledge management system for the
construction industry (Carillo et al. 2004). A study by Ibrahim et al. (2005) on knowledge
flow behaviors suggests that the non-hierarchical (or informal) knowledge networks
within the life cycle phase may influence the organizational performance of the project
team.
Most theory and practice of organization theory are based on Galbraith’s (1974
and 1977) hierarchical information-processing theory of the firm to design an
organization (Burton and Obel 2003; Jin and Levitt 1996). Earlier organization scholars
(such as Lawrence and Lorsch (1967), and Scott (2003)) have suggested the integration
of informal structure within the hierarchical decision-making structure of an organization
that would enhance an organization’s capability in responding to changing operating
environment. Moreover, Ibrahim et al.’s (2005) study highlights the need to consider
126
non-hierarchical information-processing process in the design of discontinuous
organization in addition to considering the knowledge type (tacitness versus explicitness)
dominance of the knowledge areas within a process phase.
Our long term objective is to extend the representation and reasoning of the extant
VDT in integrating Wegner’s (1987) transactive memory theory with Galbraith’s (1974
and 1977) hierarchical information-processing theory for the design of knowledge-
networked organizations. This study is our first attempt to prove that discontinuity in an
organization could affect its organizational performance by way of disrupting knowledge
flows among the team members. We extended an agent-based computational organization
tool called the Virtual Design Team (VDT) (Jin and Levitt 1996; Kunz et al. 1998) to test
this hypothesis. We first present the literature review on discontinuous organization, our
research method, how we extend the VDT tool, and the results we obtained. Then, we
conclude with discussions on the results and provide recommendations for further study.
To guide the reader, we use the term organization to represent an entity that
comprises several team members working on a process. We use the term enterprise to
represent an entity that comprises several organizations to complete a process. A process
can be sequential or concurrent. A sequential process may comprise of several different
organizations working on different parts of the process. In a concurrent process, a
different organization could work on different parts of the process independently. We
define knowledge as a set of commitments and beliefs of its holder that enables the holder
to undertake certain action (Nonaka 1994). The criterion for using the term ‘knowledge’
in this paper is its enabling action entity that allows the beholder of a knowledge entity to
undertake certain action. Explicit knowledge is the selected and applicable group of facts
127
that is transmittable in a formal systematic language that enables its beholder to take
some action to complete a task; and tacit knowledge is the entity of “knowing how” that
an individual or an enterprise possesses in selecting and applying a group of facts that
enables action to complete a task (Polanyi 1971; Nonaka 1994). On the other hand,
information is the selective collection of facts that an agent can use to perform a task,
while data are facts that an individual or enterprise can use to analyze or make a
decision.
4.3 LITERATURE REVIEW
4.3.1 Discontinuity in Organizations
Discontinuous membership is the operating situation where one or more team members
join or leave an organization while the work process is still on-going due to the need for
different skill requirements to complete successive parts of the process (Ibrahim and
Paulson 2005). A study by Ibrahim and Paulson finds several unique operating
environmental characteristics in the facility development organization that explain the
knowledge loss (K-loss) phenomenon inherent and on-going in complex processes despite
investment in latest information technology by the project owners. The facility
development process is complex in general, but the affordable housing development
process is more complex due to the financial and regulatory constraints that state and
federal programs impose on their developments and operations (Ibrahim 2001; NRC
1994). Ibrahim (2001) divides the sequential phases of the facility development life cycle
process into feasibility, entitlements, building permit, construction, and property
management phases. Ibrahim and Paulson (2005) found that among the unique
128
environmental characteristics are 1) the facility development life cycle process consists of
several sequential and concurrent phases; 2) each life cycle phase has different workflow
process that requires different skill sets for the team to complete the tasks; 3) tasks in a
workflow are interdependent with some tasks in another or several workflows of the life
cycle phases; and 4) individual phases are dominated by different knowledge types (i.e.
tacit versus explicit knowledge). Ibrahim and Nissen (2005) further refined the four
environmental characteristics into three constructs for the development of a knowledge-
based organizational performance model: knowledge flow, work complexity, and team
knowledge transaction. Knowledge flow (Nissen in review) includes explicitness, reach,
life cycle, and flow time of dominant knowledge in a phase. Work complexity considers
the ratio of tasks and workflow concurrency in the complex process. Team knowledge
transaction addresses the discontinuous nature of the team members in response to the
knowledge flow within the organization, and several studies have recommended further
research on the effects of knowledge flows on organizational performance due to this
discontinuous attribute (Ibrahim and Paulson 2005; Ibrahim et al. 2005).
4.3.2 Knowledge Flow Dynamics
Knowledge management is increasingly concerned in the design of organizations, the
effects of knowledge flow on organizational performance is emerging as the measuring
factor on how scholars reviewed knowledge flow in organization (Nissen and Levitt
2002). The development of literature concerning knowledge transfer is recent (Alavi and
Leidner 2001; Carlile and Rebentisch 2003) where scholars noted an abundance of
knowledge management literature in the areas of knowledge creation, knowledge storage,
and knowledge retrieval. The study of knowledge flow dynamics is recent (more so
129
since Nonaka 1994). Central to knowledge transfer scholars’ arguments (e.g., Nonaka
1994; Kogut and Zander 1992) is that knowledge is held by individuals, and regularities
in an organization depict the success of knowledge transfer in an organization. Nonaka
(1994) supports that knowledge resides within individuals and organizational
membership play a critical role in articulating and amplifying that knowledge. Kogut and
Zander (1992), however, pose that firms are more successful in transferring knowledge
within organizations than between organizations. Nonaka (1994) proposes four modes of
knowledge transfer mechanism—socialization, externalization, combination, and
internalization (SECI)—for internal and external knowledge transfer of an organization.
The SECI mechanism is a dynamic spiral epistemological relationship between tacit and
explicit knowledge as it extends its ontological reach from individual to inter-
organizational of that organization.
Nissen (2002) extends Nonaka’s dynamics of knowledge flow theory by
integrating the life cycle process of knowledge flow through the enterprise: 1) creation, 2)
organization, 3) formalization, 4) distribution, 5) application, and 6) evolution. This six-
step knowledge life cycle was an amalgamation of earlier views of knowledge life cycle,
which were proposed by Davenport and Prusak (1998), and Depress and Chauvel (1999).
Von Hippel (1994) coined the term ‘stickiness’ on how needed info can ‘stick’ with the
problem-solving capabilities in a different location. Stickiness connotes the difficulty
experienced in the process in which an organization recreates and maintain a complex,
causally ambiguous set of routines in a new setting (Szulanski 2000). In Nissen’s later
work (in review), he states that new organizational forms may obtain and even dominate
through a focus on dynamic knowledge flows. Nissen’s work provides discrete
130
qualitative categories for potential operationalization of knowledge flow in enterprise, but
these are yet to be measured empirically. His four knowledge flow dimensions are type of
knowledge (tacit versus explicit), level of socialization associated with the knowledge
(individual, group, organization, and inter-organization), activities of knowledge work
(create, share, apply, etc.), and flow time.
In order to understand knowledge creation by individuals, Grant (1996)
conceptualizes that the firm is an institution for integrating knowledge at the next
organization level. Grant attempts to devise mechanisms for integrating individuals’
specialized knowledge. He proposes four mechanisms to coordinate the integration of
knowledge within an enterprise: (a) having rules and directives to enable the conversion
of tacit knowledge to explicit knowledge; (b) sequencing of the workflow process that
minimizes communication, but ensures the input of an expert in a different time slot; (c)
creating routines to support complex patterns of interactions between individuals in the
absence of rules, directives, or even significant verbal communication; and (d)
establishing group problem solving and decision making. The resulting knowledge-based
firm theory has implications for the basis of organizational capability, the principles of
organization design (in particular, the analysis of hierarchy and the distribution of
decision-making authority), and the determinants of the horizontal and vertical
boundaries of the firm. Grant’s knowledge-based view of the firm encourages us to
perceive interdependence as an element of organizational design and the subject of
managerial choice rather than exogenously driven by the prevailing production
technology. Grant emphasizes knowledge application and the role of the individual as
the primary actor in knowledge creation and the principal repository of knowledge.
131
However, he points to further research need on knowledge-based theory of the firm that
will embrace knowledge creation and application.
4.3.3 Organization
We also reviewed literature on organizations to see whether it would enlighten the
knowledge-dominant difference within different life cycle phases. However, much
literature concentrates on the organization formation and behavior (March and Simon
1958; Cyert and March 1963; Galbraith 1974; Mintzberg and Westley 1992; Scott 2003;
Burton and Obel 2003). Most focus upon hierarchy as the basic structure for organizing
complex social activity where cooperation among members is achieved through vertically
imposed bureaucratic processes (Grant 1996). March and Simon (1958) identify the use
of rules or program to coordinate behavior between interdependent subtasks. Galbraith’s
information-processing model is an extension of Lawrence and Lorsch’s (1967)
contingency theory where the efficiency of an organization depends upon it adapting to
its environmental context. The information-processing model of organization (Galbraith
1974 and 1977) proposes that decision-makers need to process information well during
exception-handling if the organization wants to perform well. Galbraith argues that the
greater the task uncertainty, the greater the amount of information that participants in an
organization must process.
As an organization faces greater uncertainty, its members face situations for which
they have no rules. At this point, the hierarchy is employed on an exception basis (ibid., p.
86) where lower-ranked staff would seek guidance or information from their supervisors.
Built upon Galbraith’s information-processing model of organization, Burton and Obel
(2003) extended the contingency theory and developed six contingency factors for
132
contingency fit during the design of organizations. They are management style, climate,
size/ownership, environment, technology, and strategy. Since organization theory has
emphasized vertical information-processing structure in organization design, non-
hierarchical information-processing structure for decision-making is of particular interest
to us. Recent scholars such as Lambert and Shaw (2003) have turned to Wegner’s (1987)
transactive memory theory to explain the existence of this non-hierarchical information-
processing structure in organizations. We would like to explore the possibility of
integrating transactive memory theory with Galbraith’s (1974 and 1977) information-
processing theory in organization design.
4.3.4 Transactive Memory
A follow up study by Ibrahim et al. (2005) on Ibrahim and Paulson’s (2005) study also
recommends the integration of non-hierarchical information-processing within the
common hierarchical information-processing of an organization. An emerging theory,
which is establishing itself as the prime foundation for this purpose, is Wegner’s (1987)
transactive memory theory. Transactive memory theory involves understanding the
informal knowledge transfer—communication to retrieve or allocate information—
between individuals in an organization. Wegner (1987; in Hollingshead 1998b), describes
transactive memory as a shared system for encoding, storing, and retrieving information.
The three key processes of a transactive memory system are (a) directory updating, where
people learn what others are likely to know; (b) information allocation, where new
information is communicated to the person whose expertise will facilitate its storage; and
(c) retrieval coordination, which is a plan for retrieving needed information on any topic
based on knowledge of the relative expertise of the individuals in the memory system.
133
Wegner (1987) and later Moreland (1999) posed that organizational teams may
act like a large brain, where individuals store information to be combined with others.
When individuals in that team need information they go to the “expert node” or expert
member. When individuals gain new information related to a particular expert area, they
allocate it to an expert node. As individuals gain new information through experiences
about the expertise of the others in their “brain”, they update their individual directory of
where information should be and is stored. Ibrahim et al. (2005) hypothesized that the
behavior of members in transferring knowledge among them are similar between a
discontinuous and a stable organization. Although the purpose of this project is
determining the impact of knowledge flow processes in discontinuous membership
teams, we too expect similar patterns posed by transactive memory theory.
(H1) A continuous member in a dynamic organization, who possesses an accurate
cognition of his other team members’ knowledge skill, will be able to utilize his
knowledge network to maintain the organization’s performance when he turns to
other team members to complement his incomplete cognitive skills.
A number of communications scholars, (such as Monge and Contractor (2004);
Contractor et al. (in review); Palazzolo et al. (in review), and Yuan et al. (in review)) are
examining the social influence on development of knowledge networks among
individuals. Other contemporary scholars (such as Lambert and Shaw 2002; Hollingshead
1998b) are studying whether or not hierarchical decision-making process is applicable in
today’s fast-track project delivery. Lambert and Shaw (2002) merge transactive memory
134
theory (Wegner 1987) with the information-processing views of organizational work
processes. Lambert and Shaw argue that in current high-technology environment where
participants have equal access to information, organizations can minimize the need to
perform hierarchical decision-making process if participants know from whom to seek
information. While Lambert and Shaw focus on “who knows what,” Hollingshead’s
(1998b) study focuses on the function of communication behaviors. Lambert and Shaw,
and Hollingshead do not consider the expertise level of knowledge the enterprise
possesses. In addition, Ibrahim and Nissen (2005) suggested integration of non-
hierarchical information-processing for better representation of dynamic knowledge
flows within an organization.
Transactive memory is more important in temporal organizations where most
members have the skills and expertise to perform their tasks independently or as peers in
product development project teams (Ibrahim et al. 2005). The system may operate more
effectively in teams where individuals have higher interdependence and have developed a
more convergent cognitive map of tasks-person-expertise relationships (Hollingshead
1998a). Ibrahim et al. (2005) explains that the knowledge of “who knows what” enables
peers to consult independent team members in order to complete their own tasks,
especially when their supervisor lacks the cumulative level of expert knowledge needed
of all the team members combined. In a temporal organization, in particular, the
supervisor primarily acts as a bridge spanner or the chief coordinator. There is a thin line
of hierarchical authority in such situation. Thus, other experts in the team may have more
specialized expertise than the supervisor. A product development team, specifically the
facility development organization, reflects a norm of discontinuous membership while
135
the team works on the workflow process (Ibrahim and Paulson 2005). In this instance,
Ibrahim et al. (2005) suggest that the transactive memory in knowledge flows may be
governed by two additional mechanisms. The first mechanism is seeking information
from a mediator when one’s directory of expertise is incomplete. Wegner (1995)
describes this process of seeking information as taking on a higher dimension, when team
members need to know “who knows who, who knows what.” For example, the
remaining team members would know “who knows what” and can provide the resource
to newly joined team members when they seek information. The second mechanism
particular to discontinuous membership teams may be to seek information from those
who were in the previous phase of the project and to allocate information to those who
will continue in the next phase of the project (Ibrahim et al. 2005). In this mechanism, it
is not proper embedding of knowledge in experts, but continuing the life of the
knowledge that is the key factor governing knowledge flow patterns (ibid.).
Wegner (1987) acknowledges that an effective transactive memory system has
several advantages for a group process. Foremost is the expansion of an individual’s
expertise when the individual gains others’ domains of expertise. Another is that an
individual also gains access to new knowledge that is created through integrations
occurring within the transactive process. Integration affirms the need to have a group in
the first place, showing all members the utility of coming together to remember because
the group exerts a strong directive pressure on what is to be encoded, stored, and
retrieved and places a special premium on integrative transactions (ibid.). The third
advantage is the possibility of others processing the knowledge and making decisions
even when the individual is not available. Finally, Mooreland (1999 in Ibrahim et al.
136
2005) finds that groups that develop effective transactive memory systems can complete
tasks more efficiently. In our study, we would like to determine if the effective
transactive memory systems in stable membership organizations will apply equally to
discontinuous membership organization.
Despite the advantage of a group having a larger pool of knowledge expertise, the
transactive memory system is prone to errors made by individual members (Wegner
1987). Several sources for errors are 1) the label that is used at the beginning of the
transactive retrieval attempt can be translated into another label by another member, 2)
incomplete specification of paths of knowledge responsibility within the group may leave
certain individuals not knowing who is expert in important domains of knowledge, 3)
lack of information when other members were left out during the creative reproduction of
different retrieved knowledge, and 4) low quality information could cause a group to
make poor decisions. The biggest disadvantage is that, when a group dissolves, formerly
interdependent individuals are left with useless and potentially troublesome individual
remnants of what was once a transactive system. Similarly, Ibrahim et al. (2005) propose
that when a new member joins in, he would not know what others had done or their
knowledge expertise.
(H2) A discontinuous member, who does not have an accurate cognition of his
other team members’ knowledge skills, can put the organization at risk when he
turns to other team members to complement his incomplete cognitive skills.
137
4.3.5 Knowledge Flow and Computational Organization Theory
To study knowledge flow in organizations, three items are important: the process under
study, the organization responsible, and how they are connected. Both process and
organization must be linked because the flow of knowledge is assumed to occur with the
passage of time as the process progresses. We found the Virtual Design Team (VDT) tool
closest in fulfilling our needs. In fact, the first attempt to simulate knowledge flow in
organization design using a computational organization (COT) tool was made by Nissen
and Levitt (2002) using VDT. In their study, they draw upon current research advances in
COT to describe a dynamic representational environment used for formal organizational
modeling, and employed this environment to describe a computational model of an
enterprise’s knowledge-flow process. Their reason for utilizing the Virtual Design Team
(VDT) Research Program was due to its established reputation as a planned accumulation
of collaborative research over two decades to develop rich theory-based models of
organizational processes. VDT uses an agent-based representation (Cohen 1992; Kunz et
al. 1998) that conducted research on micro-level organizational behaviors, which were
formalized to reflect well-accepted organization theory (Levitt et al. 1999). Extensive
empirical validation projects (e.g., Christiansen 1993; Thomsen 1998) further established
its reliability. Nissen and Levitt (2002) also describe the limitation of VDT. Unlike
mathematical representations and analyzable micro-behaviors of physical systems, the
dynamics of organizations are influenced by a variety of social, technical and cultural
factors, and are difficult to verify experimentally (ibid.). They are yet amenable to
numerical representation, mathematical analysis or precise measurement (ibid.). Nissen
138
and Levitt expect ambiguity when people and social interactions—distinct from physical
systems—drive the behavior of organizations.
Nissen and Levitt (2002) point the advantage of the VDT tool to its well-
embedded time-tested collection of micro-theories that lend themselves to qualitative
representation and analysis. Among the collection are Galbraith’s (1974 and 1977)
information-processing abstraction, March and Simon’s (1958) bounded rationality
assumption, and Thompson’s (1967) task interdependence contingencies. Although the
representation is qualitative (e.g., lacking precision offered by numerical models), Nissen
and Levitt (2002) accept that it becomes formal (e.g., people viewing the model agree on
exactly what it describes), reliable (e.g., the same sets of organizational conditions and
environmental factors generate the same sets of behaviors) and explicit (e.g., much
ambiguity inherent in natural language is obviated) through computational modeling.
Once a model has been validated to accurately emulate the qualitative behaviors of the
field organization it represents, it can be used to examine a multitude of cases under
controlled conditions, hence offering great promise for theory development (ibid.). The
approach they took to modeling knowledge flow dynamics was due to the strong
integration of qualitative and quantitative models, enhanced by strong reliance upon well-
established theory and commitment to empirical validation, that their study differs from
extant knowledge-flow research to provide new insight into the dynamics of
organizational behavior. While we support their arguments in our attempt to extend
Wegner’s (1987) transactive memory in Galbraith’s (1974 and 1977) information-
processing concept, Ibrahim et al. (2005) provide an empirical support that affirmed the
existence of such non-hierarchical process. Ibrahim et al. found different knowledge flow
139
behaviors caused by the different explicitness level of the knowledge areas involved
where, while functional expertise is reliable to facilitate knowledge flow in explicit-
dominant process, tacit-dominant process relies on the presence of the organization’s
members during that process.
4.4 RESEARCH METHOD
Computational Organization Theory Tool (SimVision®)
We used SimVision®, educational version 3.11.1, which was developed by Vité
Corporation and is licensed from ePM, LLC, Austin Texas. Refer http://www.epm.cc/
for details.
A virtual concept design project
Using SimVision®, we created an idealized work process model consisting of ten
sequential and concurrent tasks conducted by seven team members to represent a
hypothetical conceptual design project (see Figure 4-1). In SimVision®, a project is
modeled as a set of graphical objects that represent the work process performed by
members of one or more organizations to achieve a key business milestone. Each project
is composed of tasks, milestones, positions, meetings, and the links among these
components. We analyze this virtual conceptual design project at the Project level.
140
Figure 4-1. The organization and workflow for the hypothetical concept design project.
Test Case 1: Baseline
We set the project parameters to MEDIUM for team experience, LOW for centralization,
MEDIUM for formalization, MEDIUM for matrix-strength, 0.70 rating for
communication-probability, 0.15 rating for noise-problem, and 0.15 rating for project-
exception-probability. Due to having a small number of task volumes in the test cases, we
set the behavior parameters for noise-problem, functional-exception-probabilities, and
project-exception-probabilities somewhat higher than the normal construction industry
practice in order to amplify the effects of knowledge flows on outcomes.
Team members’ parameters. Table 4-1 lists the organization’s micro-behaviors
and knowledge skills (K-skills). The micro-behaviors include fulltime-equivalent (FTE),
role, application experience, and salary; while the knowledge skills we represented are
141
architectural-engineering-construction (AEC), regulatory and authority requirement
(RAR), and development project finance (DPF). We selected the environmental
consultant (EC) to test the accuracy of an agent’s knowledge network (Knetwork) in this
organization. Table 4-2 lists the accurate and inaccurate cognitive perception by the EC
towards the other agents in the organization. Each cognitive perception is appointed
LOW, MEDIUM, or HIGH expertise. For the Baseline case, all six knowledge links in
the EC’s Knetwork have similar K-skill cognitions as the members’ K-skills in the
project’s position.
Table 4-1. Baseline Parameters for Discontinuous Membership Organization MICRO-BEHAVIORS KNOWLEDGE
SKILLS ID POSITION INI
FTE ROLE APP EXP
SALARY
AEC
RAR DPF
1 Project Manager PM 0.3 PM M 80 M M M 2 Architect AR 1.0 SL H 150 H M L 3 Civil Engineer CE 0.5 ST H 140 M M L 4 Landscape
Architect LA 0.5 ST M 100 M L -
5 Financial Consultant
FC 1.0 ST H 200 - H H
6 Environmental Consultant
EC 1.0 ST L 100 L L -
7 Geotech Consultant
GC 1.0 ST M 100 M L -
Table 4-2. Environmental Consultant’s Knowledge Cognition of Other Team Members in Baseline and X-Baseline Cases
EC’s Knowledge Cognition
PM
AR
CE
LA
FC
GC
Accurate (Baseline)
M M M L H L
Inaccurate (X-Baseline)
L L M M - M
142
Test Case 2: X-Baseline
The X-Baseline case is the inaccurate Knetwork model. It has similar project parameters
as the Baseline model. To represent the inaccurate knowledge cognition (see Table 4-2),
we made several changes to the team members’ parameters. Two members (e.g., PM and
AR) who have HIGH K-skills to solve the exception were changed to low, so the new
member will miss the opportunity to obtain correct feedback. The two LOW K-skilled
members (LA and GC) were assigned higher skill levels, so that the new member will
obtain incorrect feedback. Only one member out of the five in the new member’s
Knetwork has the accurate K-skill to solve the exception.
Analysis Method
We ran 50 trials for each simulation, and made ten simulation runs (N=500). We
compared the Baseline and X-Baseline cases on selected performance variables at
project, task, and position levels.
Limitation and validation
This computational simulation experiment was conducted and validated only for an
idealized case—i.e., “Intellective Validation” (Thomsen et al. 1999)—to emulate the
behaviors of an idealized organization with discontinuous membership. Although
idealized, the organization’s staffing is relatively closely based on an earlier Knetwork
study using knowledge network analysis (Ibrahim et al. 2005). We ran the same
simulations ten times to provide us a larger population, N = 500 trials, which we perform
a two-tail t-test at 95% for each parameter we compare.
143
4.5 FUNCTIONAL EXCEPTIONS PARAMETER EXTENSION FOR
KNOWLEDGE NETWORK
The hierarchical information-processing model used by VDT incorporated two types of
exceptions: project exceptions and functional exceptions. Project exceptions represent
problems that occur in the interfaces between teams working in parallel on related tasks
within the project. Functional exceptions represent unanticipated technical problems in
the implementation of individual tasks. In the original VDT model, the two types of
exceptions were generated and tracked separately, but the same exception handling
behavior was used in both cases. In particular, when an agent encountered either type of
exception, the decision would be made to pass the exception to someone higher in the
project hierarchy based solely on the agent’s own role and the centralization parameter
setting for the project organization. VDT-KN retains this behavior for project exceptions,
but adds a new peer-to-peer relationship—we called knowledge-network (Knetwork)
information-processing—for establishing functionally-based exception handling
pathways.
We describe below the manner how the functional exceptions will be handled via
the Knetwork information-processing methodology. Based on Hypothesis 1—when an
agent seeks other members in his Knetwork to complement his incomplete cognitive
skill—the VDT code will randomly decide who to pass the exception to basing on the
perceived K-skill level of the agents. The following steps explain the iteration and the
effects on the project, task, and position caused by the flow of exception-handling by
various agents in the organization.
144
Step 1: Generation of functional exceptions from task. A task is divided into sub-
tasks, which are either one work-day in length, or one-twentieth of the nominal duration
of the task, whichever is less. The completion of a subtask represents an abstract
decision point, where the agent effectively decides whether the subtask has been
completed successfully, or has failed, leading to initiation of a project or functional
exception. When a subtask is completed (i.e., when one work-day is completed for tasks
requiring more than 20 days to complete), a functional exception will be generated
according to the current task functional exception probability. The initial task functional
exception probability is a function of the project functional exception probability,
requirement complexity, and K-skill of the agent. The factor for the requirement
complexity is 1.5 if HIGH, 1.0 if MEDIUM, and 0.67 if LOW. The K-skill factor of the
agent corresponds to his application experience. When the assigned agent has the
required skill, his skill factor corresponds to his application experience. The higher the
application experience he has, the lower the number of exceptions will be generated. For
example, if the assigned agent has HIGH application experience, his K-skill factor is 0.5
compared to 0.7 and 0.9 for a MEDIUM or LOW application experience. If the assigned
agent’s K-skill does not match the task’s required K-skill, the application experience is
set at default at 2.0 for HIGH, 2.5 for MEDIUM, and 3.5 for LOW. It means that there
will be more exceptions being generated when the K-skill does not match the task
requirement.
Step 2: Who to pass the exception to? When a functional exception is generated,
the assigned agent will either resolve the exception himself, or send it to a member of his
Knetwork with the same or higher perceived rating of the K-skill required by the task.
145
This assignment differs from VDT where functional exceptions are passed up the
supervision hierarchy. If there is at least one Knetwork member who is believed to have
the necessary skill, the probability that it will be passed to the network is determined by
the assigned agent’s K-skill rating. The probability is 0.2 if K-skill rating is HIGH, 0.5 if
MEDIUM, and 0.8 if LOW or NONE. This means that higher K-skill agents tend to
resolve functional exceptions themselves, while lower K-skilled agents tend to pass these
exceptions to the Knetwork for solution. When a functional exception is passed to the
Knetwork, and there is more than one agent with the highest available skill, one of these
agents will be selected at random to provide the solution.
Step 3: Affects on coordination volume. A functional exception, whether resolved
by the originating agent or a member of that agent’s Knetwork, requires coordination
work to obtain a solution. This coordination volume is 0.15 times the work volume of the
originating subtask. This is changed from the original VDT model, where functional
exceptions had fixed coordination volumes of 30 FTE-minutes for a PM, and 60 FTE-
minutes for agents with other roles. This change was made to avoid disproportionate
increases in coordination volume resulting from exceptions generated by short duration
subtasks.
Step 4: How does the agent respond to an exception? An agent deciding the
outcome of a functional exception will make a decision based on its rating for the K-skill
required by the task that originated the exception. There are three possible outcomes:
rework, which requires rework of 100% of the originating subtask work volume, correct,
which requires reworking 50% of the subtask work volume, and ignore, which requires
no additional rework. When the agent’s K-skill level is higher, there is a greater
146
likelihood that the agent will require reworking some or all of the subtask volume.
Greater amounts of rework cause a decrease in the task functional exception probability,
which reduces the likelihood of future exceptions. For example, if the agent’s K-skill is
HIGH; the factors for rework, correction, and ignore are 0.65, 0.30, and 0.05
respectively. If the agent’s K-skill is MEDIUM, the factors for rework, correction, and
ignore are 0.40, 0.40, and 0.20 respectively. And if the agent’s K-skill is LOW, the
factors for rework, correction, and ignore are 0.05, 0.35, and 0.60 respectively.
The agent who is assigned to implement the decision will incur additional
coordination volume due to the additional workload of processing the decision. The
coordination volume increases by a factor of 0.10 times the subtask’s volume. In
addition, if the decision was to rework the subtask, the rework volume of the associated
task is increased by a factor of 1.0 times the subtask volume. If the decision was to
correct the subtask, the task rework volume is increased by 0.5 times the subtask volume.
Step 5: Affects on the functional error probability. The functional error
probability of the exception task will be adjusted according to the decisions made by the
agent who makes the decision on the exception. It is based on the decision-maker’s K-
skill and the type of decision he makes. The probability adjustment is:
1.0 + (decision-weight – 1.0) * volume-weight
where the volume-weight is the nominal subtask work volume of 8 FTE-hours. The
decision-weight depends on the agent’s K-skill. If the agent’s K-skill is HIGH, the
rework factor is 0.90, the corrected factor is 0.95, and ignored factor is 1.05. On the
other hand, if the agent’s K-skill is LOW, the rework factor is 0.95, the corrected factor is
1.10, and ignored factor is 1.20. The functional error probability reflects the need to do
147
rework on a task when exceptions are being ignored by less K-skilled agent. The higher
the functional error probability, the higher the functional risk index is.
If the originating agent has an incorrect perception of the skills of the members of
its Knetwork, there is increased likelihood that functional exceptions will be directed to
agents with inappropriate skill sets. This results in decreased likelihood of reworking
flawed subtasks, and increases the task functional exception probability.
4.6 RESULTS AND ANALYSIS
Our computational simulations illustrate that a new member in a discontinuous
membership organization who uses his inaccurate cognition of his other team members to
complement his incomplete cognitive skill, can expose his task to higher risk of failure,
which in turn can be detrimental to his overall team’s organizational performance in the
long run. The results support H1 and H2, and subsequently our main research question on
how knowledge flows can impact the organizational performance of a discontinuous
organization. At the project level (see Table 4-3), there is minimal change to the overall
simulated durations and total work volumes of both Baseline and X-Baseline cases.
Rework volume is reduced (by 10%) in the X-Baseline case because less skilled team
members are ignoring the exceptions that the EC had allocated to them for solutions. The
coordination volume shows a slight increase because the positive and negative
increments of multiple members roughly cancelled one another out. On the other hand,
the wait volume has a huge increase (57%) in the X-Baseline case because more
exceptions are occurring due to the increasing error rate caused by inaccurate Knetwork’s
skill matching. The project’s FRI increases from 0.346 to 0.535 respectively from the
148
Baseline to X-Baseline case. FRI is a measure of the likelihood that components
produced by a project have defects. In the X-Baseline case, the EC’s task to complete
environmental report has a higher likelihood of having errors.
Table 4-3. Comparison of Selected Organizational Performance Measures for Continuous Versus Discontinuous Membership
PERFORMANCE MEASURES
BASELINE (SD)
X-BASELINE (SD)
PERFORMANCE CHANGE
PROJECT Simulated Duration (days)
189.259 (7.059)* 189.048 (8.119)* 0%
Total Work Volume (days)
119.052 (3.673)* 119.709 (4.768)* +1%
Rework Volume (days)
13.161 (2.445)* 11.828 (2.619)* -10%
Coordination Volume (days)
11.853 (0.790)* 12.119 (1.081)* +2%
Wait Volume (days)
3.037 (1.844)* 4.762 (1.881)* +57%
FRI 0.346 (0.073)* 0.535 (0.067)* +0.189 TASK FRI 0.315 (0.101)* 0.667 (0.076)* +0.352 Functional Exceptions Probability
0.197 0.479 +0.282
Rework Volume (days)
7.084 5.713 -19%
Coordination Volume (days)
1.400 1.685 +20%
Wait Volume (days)
2.865 4.554 +59%
Note: * p < 0.05 (two-tail t-test); N=500
At the task level, the FRI in the X-Baseline case doubled from 0.315 to 0.667
from the Baseline case. The functional exception probability increased by 0.282 points.
This is because the error rate increased when the EC allocated and retrieved inaccurate
solutions for his exceptions from lesser skilled members. In VDT’s organizational design
recommendation, the high FRI, coupled with a large increase in the functional exceptions
probability rate highlights that the complete environmental report task requires urgent
149
attention and action by the project manager or the assigned member responsible for the
task. The task also reflects the reduction in rework (-19%) due to the lesser K-skilled
member’s propensity to ignore exceptions. Coordination volume increased (20%)
because more exceptions were generated that required coordination attention by the EC,
while the wait volume reflects a big increase of 59% due to the increase of error rate
made by more frequent occurrence of exceptions.
At the position level (see Table 4-4), we highlight the effects of inaccurate
Knetwork on selected members. First is the environmental consultant (EC) who passed
the exception to others in his Knetwork. The other two are the landscape architect (LA)
and general contractor (GC) who were perceived as having higher K-skill levels, but did
not. For the EC, his total exception volume increased by 24% due to the increasing error
rate caused by passing exceptions to lesser skilled members in his Knetwork. His rework
volume was reduced by 19% since the lesser skilled member advised or undertook to
ignore the exceptions. His coordination volume increased by 10% since he is required to
coordinate more due to the increasing number of exceptions. The EC has the biggest wait
volume increase (59%) because he now has to wait for decisions on an expected larger
number of exceptions. For the LA, his total exception volume had a small increase of 4%.
His rework volume was only 1% since he was mostly likely to ‘solve’ the problem
although it may be incorrect. However, his coordination volume increased by 22% since
he had higher work volume due to the increasing number of exceptions being passed to
him. In addition, the LA reduced his wait volume by 31% because he would ‘solve’ the
exceptions himself, albeit incorrectly. For the GC, his total exception volume had no
change. His rework volume was 10% because he did handle some exceptions correctly.
150
Like the LA, the CG increased his coordination volume by 31% due to higher work
volume caused by increasing passing of exceptions to him. The GC had no wait volume.
Table 4-4. Comparison of Organizational Performance Measures for Selected Members PERFORMANCE
MEASURES
BASELINE (SD)
X-BASELINE
(SD)
PERFORMANCE CHANGE
Environmental Consultant Total Exceptions Volume (days)
12.328 (2.331)*
15.315 (3.591)*
24%
Maximum Backlog (days)
7.929 9.368 18%
Rework Volume (days)
7.084 5.713 -19%
Coordination Volume (days)
3.384 3.730 10%
Wait Volume (days)
2.865 4.554 59%
Landscape Architect Total Exceptions Volume (days)
1.480 (0.662)*
1.534 (0.677)*
4%
Maximum Backlog (days)
1.714 1.692 -1%
Rework Volume (days)
0.784 0.793 1%
Coordination Volume (days)
1.755 2.147 22%
Wait Volume (days)
0.150 0.104 -31%
Geotech Consultant Total Exceptions Volume (days)
2.100 (1.446)*
2.105 (1.459)*
0%
Maximum Backlog (days)
2.176 2.292 5%
Rework Volume (days)
1.24 1.36 10%
Coordination Volume (days)
1.349 1.761 31%
Wait Volume (days)
0.000 0.000 0%
Note: * p < 0.05 (two-tail t-test); N=500
151
4.7 VALIDATION
We conducted an intellective model comparison (Thomsen, Levitt, et al. 1999) of our
simulated model with an earlier ethnographic study (Ibrahim and Paulson 2005) to
compare the outcome. A two-tail t-test at 95% significance level was applied to the
sample of 50 trials from N=500 trials. We found that H1 is positive, where a continuous
member in a dynamic organization, who possesses an accurate cognition of his other
team members’ K-skill, is able to utilize his knowledge network to maintain the
organization’s performance when he turns to other team members to complement his
incomplete cognitive skills. H2 is also positive, where a discontinuous member, who does
not have an accurate cognition of his other team members’ K-skills, can put the
organization at risk when he turns to other team members to complement his incomplete
cognitive skills.
We used a computational simulation model to test our hypotheses. The VDT
model we used has been validated by many previous researchers (e.g., Thomsen et al.
1999), and fulfills the three key criteria for being used as a “theorem prover” (Burton and
Obel 1995) - reality, content, and structure - to examine hypotheses. Further validation is
provided by cross-validating Ibrahim and Paulson’s (2005) ethnographic study on
discontinuity in organizations. Their study found discontinuous membership contributing
to knowledge-loss in the facility development projects when an enterprise works on a
complex process (i.e., which has multiple concurrent phases, high task interdependencies,
etc.) in a dynamic environment (i.e., high on uncertainty, complexity, and equivocality).
152
4.8 DISCUSSION AND RECOMMENDATIONS
Our study demonstrates that discontinuous membership negatively affects the overall
performance of an organization. Several implications emerge from this study. We present
these implications, and recommend future studies in this discussion section.
First, the study supports another study on how discontinuous membership is a
source for the well-known knowledge loss phenomenon in the facility development life
cycle (Ibrahim and Paulson 2005). A new member could cause the task he is handling to
incur higher functional risk index, and in the long run could put the whole project at risk.
The incomplete knowledge of prior history of a task or project can trigger an escalation
of schedule delays and cost overruns. It is unfortunate that, during the pre-construction
phases in the facility development, any missing knowledge may force facility owners to
decide not to proceed with the construction.
Second, the discontinuous attribute undermines Cohen and Levinthal’s (1990)
finding about the absorptive capacity of a firm, in which an organization builds upon its
prior knowledge. Instead, the progressive build-up of a discontinuous organization’s
knowledge is weakened because former members would bring out the organization’s
knowledge while the rest of the team builds up its knowledge. Hence, the process
emphasis that Scott (2003) recommends in an open system draws our attention and
support to the organization as a system persisting over time. An open system exhibits an
organizational structure that stresses the complexity and variability of the individual
parts—individual participants and subgroups—as well as the looseness of connections
among them. Multiple parts are viewed as capable of semiautonomous action, and are
loosely coupled to other parts. In the open system, individuals and subgroups form and
153
leave the coalitions, emphasizing the earlier contingency theory of Lawrence and
Lorsch’s (1967) arguments about the fluid movement between people and process while
the process is still on-going.
Third, the study saw how regularities in the organization (Grant 1996; Kogut and
Zander 1992) do help the organization to overcome a position’s dynamism. The project
could still move forward despite having a new position on board or after the position is
omitted. This is evidenced by the lack of significant changes to the overall total work
volume and the duration of project for both test cases. However, our intellective model
reflects how a subtly incomplete Knetwork can impact negatively on the performance of
the overall project as a well-established organization moves forward.
Fourth, there is a need to consider the integration of non-hierarchical (Wegner
1987) with hierarchical information-processing (Galbraith 1974 and 1977) for
contingency fit in organization design (Burton and Obel 2003). The VDT-KN extension
integrated Wegner’s (1987) transactive memory theory into the Virtual Design Team’s
(VDT) computational organization simulation model to handle all the functional
exceptions a task generates. VDT is a well-validated computational organization theory
tool (Jin and Levitt 1996; Burton and Obel 1995) that uses Galbraith’s (1974 and 1977)
hierarchical information-processing process to design an organization, and the intellective
model proves that such non-hierarchical information-processing affects organizational
performance.
Our findings recommend that future organizational design must consider the
knowledge flow factor as posed by Nissen (in review). This study supports his view that
an enterprise could obtain or dominate a new organization through a focus on dynamic
154
knowledge flows. Nissen’s work provides discrete qualitative categories for potential
operationalization of knowledge flow in an enterprise. His four knowledge flow
dimensions are type of knowledge (tacit versus explicit), level of socialization associated
with the knowledge (individual, group, organization, and inter-organization), activities of
knowledge work (create, share, apply, etc.), and flow time. Further study is recommended
to establish numerical measurements for these constructs. Moreover, another study by
Ibrahim et al. (2005) affirms that tacit-dominant knowledge area shows different
knowledge flow behaviors compared to explicit-dominant knowledge area. They found
that transactive memory (Wegner (1987) is not happening in tacit-dominant knowledge
areas, but is very much supported in explicit-dominant knowledge areas. They suggest
that knowledge flows in a tacit-dominant environment is more via socialization and
internalization (Nonaka 1994). Ibrahim et al. (2005) also conclude with a similar
recommendation to consider knowledge flow in the design of an organization.
Further literature search in the communications field in sociology reveals that
organizational communication scholars (Monge and Contractor 2003) have in recent
years established a new area to study emergent communication networks in their study of
information flow in organizations. The traditional interpretation of a “formal” network
(Weber (1947) in Monge and Contractor (2003)) was presumed to represent the channels
of communication through which orders were transmitted downward and information was
transmitted upward10 (p. 8). On the other hand, “emergent” networks represent other than
the vertical hierarchical characteristics of information-processing. It includes the
10 See Scott’s (2003) explanation of a rational system, referring to the technical or functional purpose of the organization to meet the implementation goals of the work process.
155
development of hierarchy through socialization outside the formal structure.11 We equate
the “emergent” network as synonymous to our non-hierarchical knowledge network. The
rationale for studying this emergent communication networks is due to the fact that it:
“……has evolved out of the inconclusive findings relating formal organizational
structure to organizational behavior (Johnson, 1992, 1993; also see McPhee &
Pole, 2001). Jablin’s (1987) review of the empirical research on formal
organizational structure variables such as hierarchy, size, differentiation, and
formalization. More recently, a series of meta-analytic studies have concluded
that the relationships between formal structure, organizational effectiveness
(Doty, Glick, & Huber, 1993; Huber, Miller, & Glick, 1990), and technology
(Miller, Glick, Wang, & Huber, 1991) are largely an artifact of methodological
designs. The fact that formal structural variables have failed to provide much
explanatory power has led several scholars to question the utility of further
research on formal structures. Rather, they have argued that it is preferable to
study emergent structures because they better contribute to our understanding of
organizational behavior (Bacharach & Lawler, 1980; Krackhardt & Hanson,
1993; Krikorian, Seibold, & Goode, 1997; Roberts & O’Reilly, 1978;
Roethlisberger & Dickson, 1939).
- Monge and Contractor (2003, p. 9)
11 Also see Scott’s (2003) explanation of natural system, referring to the social collectivities of the organization that try to survive the situational circumstances.
156
In summary, since discontinuity in organizations affects organizational
performance, several implications lead us to recommend two major areas for further
research. First is in organization theory, where we recommend the consideration of
knowledge flows in organization design. Second is in knowledge management systems,
where we recommend the consideration of different knowledge type for better knowledge
transfer mechanisms.
4.9 CONCLUSIONS
We used a computational organizational theory tool to understand how the flow of
knowledge among team members impacts organizational performance in organizations
with discontinuous membership. The study establishes that a new member in a
discontinuous membership organization who uses his inaccurate cognition of his other
team members to complement his incomplete cognitive skill, can expose his task to
higher risk of failure, which in turn can be detrimental to his overall team’s
organizational performance in the long run. It illustrates the need to consider non-
hierarchical knowledge networks as part of organization design for better accuracy in
predicting organizational performance. This study contributes by merging Wegner’s
(1987) non-hierarchical transactive memory with Galbraith’s (1974 and 1977) vertical
information-processing approach to the design of organization. It extends the virtual
design team tool by integrating exception handling through non-hierarchical knowledge
networks among peers, while establishing how knowledge flows can affect organizational
performance when the organization has discontinuous membership.
157
4.10 REFERENCES
Alavi, M. and D. E. Leidner. 2001. Review: Knowledge management and knowledge
management systems: Conceptual foundations and research issues. MIS Quarterly 25 (1): 107-136.
Bacharach, S. B., P. Bamberger, et al. 1996. The organizational transformation process:
The micropolitics of dissonance reduction and the alignment of logics of action. Administrative Science Quarterly 41 (3): 477-506.
Burton, R. M., and B. Obel. 1995. The validity of computational models in organization
science: From model realism to purpose of the model. Computational and Mathematical Organization Theory (1): 57-71.
Burton, R. M., and B. Obel. 2003. Strategic organizational diagnosis and design:
Developing theory for application. Boston: Kluwer Academic Publishers. Carillo, P., H. Robinson, A. Al-Ghassani, and A. Chimay. 2004. Knowledge management
in UK: Construction, strategies, resources and barriers. Project Management Journal 35 (1): 46-56.
Carlile, P. R. and E. S. Rebentisch, E. S. 2003. Into the black box: The knowledge
transformation cycle. Management Science 49 (9): 1180-1195. Christiansen, T. 1993. Modeling efficiency and effectiveness of coordination in
engineering design teams: VDT-the Virtual Design Team. Ph.D. diss., Department of Civil and Environmental Engineering, Stanford University.
Cohen, G.P. 1992. The Virtual Design Team: An object-oriented model of information
sharing in project team. Ph.D. Diss., Department of Civil and Environmental Engineering, Stanford University.
Cohen, W. M. and D. A. Levinthal. 1990. Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly 35 (1, Special Issue: Technology, Organizations, and Innovation): 128-152.
Contractor, N., D. Brandon, M. Huang, E. T. Palazzolo, D. Steinley, C. Su, V. R. Suri, S.
A. Swarbrick, J. Templin, and S. Wasserman (in review). Multi-theoretical multilevel models of information retrieval in knowledge networks.
Cyert, R. M., and J. G. March. 1963. A behavioral theory of the firm. Englewood Cliffs:
Prentice-Hall. Davenport, T. H., and L. Prusak. 1998. Successful knowledge management projects. Sloan
Management Review 37 (4): 53-65.
158
Depress, C., and D. Chauvel. 1999. Mastering information management: Part six- knowledge management. Financial Times (8 March): 4-6.
Doty, D. H., W. H. Glick, and G. P. Huber. 1993. Fit, equifinality, and organizational
effectiveness: A test of two configurational theories. Academy of Management Journal 36: 1196-1250.
Galbraith, J. R. 1974. Organization design: An information processing view. Interfaces 4
(3): 28-36. Galbraith, J. R. 1977. Organization design. Reading, Massachusetts: Addison-Wesley. Hollingshead, A. B. 1998a. Cognitive interdependence and convergent expectations.
Journal of Personality and Social Psychology 81 (6): 1080-1089. Hollingshead, A. B. 1998b. Retrieval processes in transactive memory systems. Journal
of Personality and Social Psychology 74 (3): 659-671. Huber, G. P., C. C. Miller, and W. H. Glick. 1990. Developing more encompassing
theories about organizations: The centralization-effectiveness relationship as an example. Organization Science 1: 11-40.
Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for
affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Ibrahim, R., and M. E. Nissen. 2003. “Emerging technology to model dynamic
knowledge creation and flow among construction industry stakeholders during the critical feasibility-entitlements phase,” Proceedings of The ASCE 4th Joint Symposium on IT in Civil Engineering in Nashville, Tennessee, November 15-16 by the American Society of Civil Engineers.
Ibrahim, R., and M. E. Nissen. 2004. “Simulating environmental contingencies using
SimVision®,” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pensylvannia, June 27-29 by the North American Association for Computational Social and Organization Science.
Ibrahim, R., and B. C. Paulson, Jr. 2005. Discontinuity in organizations: How
environmental characteristics contribute to the project's knowledge loss phenomenon. CRGP Working Paper No. 12, Stanford University, CA.
Ibrahim, R., and M. E. Nissen. 2005. “Developing a knowledge-based organizational
performance model for discontinuous participatory enterprises,” Proceedings of the Thirty-Eighth Annual Hawaii International Conference on System Sciences in Hawaii January 3-6 by College of Business Administration, University of Hawaii at Manoa.
159
Ibrahim, R., M. Shumate, R. Levitt, and N. Contractor. 2005. Discontinuity in organizations: Knowledge flow behaviors in sequential workflow processes. CRGP Working Paper No. 17, Stanford University.
Jablin, F., and K. J. Krone. 1987. Organizational assimilation. Newbury Park, California:
Sage. Jin, Y., and R. E. Levitt. 1996. The virtual design team: A computational model of
project organizations. Computational and Mathematical Organization Theory 2 (3): 171-196.
Johnson, J. D. 1992. Approaches to organizational communication structure. Journal of
Business Research 25: 99-113. Johnson, J. D. 1993. Organizational communication structure. Norwood, NJ: Ablex. Kogut, B., and U. Zander. 1992. Knowledge of the firm, combinative capabilities, and the
replication of technology. Organization Science 3 (3, Focused Issue: Management of Technology): 383-397.
Krackhardt, D., and J. R. Hanson. 1993. Informal networks: The company behind the
chart. Harvard Business Review 71: 104. Krikorian, D. D., D. R. Seibold, and P. L. Goode. 1997. Reengineering at LAC: A case
study of emergent network processes. Edited by B. D. Sypher. Case studies in organizational communication 2: Perspectives on contemporary work life (129-144). New York: Guilford.
Kunz, J.C., R. E. Levitt, and Y. Jin. 1998. The virtual design team: A computational
simulation model of project organizations. Communications of the Association for Computing Machinery 41 (11): 84-92.
Lambert, M. H., and B. Shaw. 2002. Transactive memory and exception handling in
high-performance project teams. CIFE Technical Report No. 137, Stanford University.
Lawrence, P. R., and J. W. Lorsch 1967. Organization and environment: Managing differentiation and integration. Boston: Graduate School of Business Administration, Harvard University.
Levitt, R.E., J. Thomsen, T. R. Christiansen, J. C. Kunz, Y. Jin, and C. Nass. 1999. Simulating project work processes and organizations: Toward a micro-contingency theory of organizational design. Management Science 45 (11): 1479-1495.
160
March, J. G., and H. A. Simon. 1958. Organizations. New York: John Wiley & Sons, Inc.
McPhee, R. D., and M. S. Poole. 2001. Organizational structures and configurations.
Edited by F. M. Jablin and L. L. Putnam. The new handbook of organizational communication: Advances in theory, research, and methods (503-543). Thousand Oaks, CA: Sage.
Miller, C. C., W. H. Glick, Y. D. Wang, and G. P. Huber. 1991. Understanding
technology-structure relationships: Theory development and meta-analytic theory testing. Academy of Management Journal 34: 370-399.
Mintzberg, H., and F. Westley. 1992. Cycles of organizational change. Strategic
Management Journal 13 (Special Issue: Fundamental Themes in Strategy Process Research): 39-59.
Monge, P. E., and N. S. Contractor. 2003. Theories of communication networks. Oxford,
UK: Oxford University Press. Moreland, R. 1999. Transactive memory: Learning who knows what in work groups and
organizations. In J/ Levine (Ed.) Shared cognition in organizations: The management of knowledge. Lawrence Erlbaum Associates, Hillsdale, NJ.
Neighborhood Reinvestment Corporation. 1994. How to be your own developer: Making
the development decision. Washington, D.C.: Neighborhood Reinvestment Corporation.
Nissen, M. E. (in review). Understanding knowledge-flow patterns for design. Nissen, M. E., and R. E. Levitt. 2002. Dynamic models of knowledge-flow dynamics.
CIFE Working Paper No. 76, Stanford University. Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization
Science 5 (1): 14-37. Palazzolo, E. T., D. Serb, Y. She, C. Su, and N. S. Contractor (in review). Co-Evolution
of communications and knowledge networks as transactive memory systems: Using computational models for theoretical integration and extensions.
Polanyi, M. (1967). The tacit dimension. London, Routledge and Keoan Paul. Roberts, K. H., and C. A. O’Reilly. 1979. Some correlates of communication roles in
organizations. Academy of Management Journal 22: 42-57. Roethlisberger, F., W. Dickson (1939). Management and the worker. New York: Wiley.
161
Scott, W. R. 2003. Organizations: rational, natural, and open systems. New Jersey: Prentice Hall.
Szulanski, G. 2000. The process of knowledge transfer: A diachronic analysis of
stickiness. Organizational Behavior and Human Decision Processes 82 (1): 9-27. Thompson, J. D. 1967. Organization in action: Social science bases in administrative
theory. New York, NY: McGraw-Hill. Thomsen, J. 1998. Virtual Team Alliance (VDA): Modeling the effects of goal
incongruence in semi-routine, fast-paced project organizations. Ph. D. Diss., Department of Civil and Environmental Engineering, Stanford University.
Thomsen, J., R. E. Levitt, J. Kunz, C. Nass, and D. Fridsma. 1999. A trajectory for
validating computational emulation models of organizations. Computational and Mathematical Organization Theory 5 (4): 385-401.
Von Hippel, E. 1994. "Sticky information" and the locus of problem solving: Implications
for innovation. Management Science 40 (4): 429-439. Weber, M. 1947. The theory of social and economic organization (A. H. Henderson & T.
Parsons, Eds. & Trans.). Glencoe, IL: Free Press. Wegner, D. M. 1987. Transactive memory: A contemporary analysis of the group mind.
Edited by B. Mullen, and G. R. Goethals. Theories of group behavior. New York: Springer-Verlag, 185-208.
Wegner, D. M. 1995. A computer network model of human transactive memory. Social
Cognition 13 (3): 319-339. Yuan, Y., J. Fulk, M. Shumate, P. Monge, J. A. Bryant, and M. Matsaganis (in review).
Individual participation in organizational information commons: the impact of team level social influence and technology-specific competence.
162
CHAPTER 5
DISCONTINUITY IN ORGANIZATIONS:
DEVELOPING A KNOWLEDGE-BASED ORGANIZATIONAL
PERFORMANCE MODEL FOR DISCONTINUOUS MEMBERSHIP
Rahinah Ibrahim12 and Mark Nissen13
5.1 ABSTRACT
Knowledge management research needs consistent and empirically supported theory for
sound foundations. Our research seeks to understand how to extend established
organization theory and emerging knowledge-flow theory to inform the design of
organizations through development of empirically grounded theory. Because knowledge
flows enable workflows, and workflows drive performance, theory suggests the
organization of knowledge—particularly tacit knowledge—is critical for competitive
advantage. However, tacit knowledge does not flow well through the enterprise. It
attenuates particularly quickly in organizations that experience discontinuous
membership. The research described in this article builds upon an ethnographic study of
facility development to understand the knowledge-flow patterns and pathologies in an
organization that experiences severe discontinuous membership. Informed by
ethnography, we employ methods and tools of computational organization theory to
12 Stanford University 13 Naval Postgraduate School, Monterey, California
163
assess alternate organizational designs and knowledge-flow patterns. This work extends
organization theory to address the dynamics of knowledge flows. And it informs practice
through new theory on designing organizations with discontinuous membership.
Key Words: Discontinuous membership, dynamics, knowledge flows, knowledge
management, information processing, contingency theory, organization design.
5.2 INTRODUCTION
Knowledge management research needs a consistent and cohesive theory supported by
empirical evidence to provide sound and stable foundations for the field (Edwards et al.
2003). Our research seeks to understand how to extend established organization theory
and emerging knowledge-flow theory to inform the design of organizations through
development of empirical theory. Because knowledge flows enable workflows, and
workflows drive performance, they are essential to organizational performance wherever
knowledge and information work are involved. Moreover, theory suggests the
organization of knowledge—particularly tacit knowledge—is critical for competitive
advantage (Nonaka 1994; Carlile and Rebentisch 2003). This is the case in particular
where different types of knowledge (e.g., tacit, explicit) exhibit different knowledge-flow
behaviors (Ibrahim et al 2005a).
The problem is, tacit knowledge does not flow well through the enterprise (Nissen
2002). Rather, it “sticks” (von Hippel 1994) predictably in specific people, organizations,
places and times (Nissen 2005b). Moreover, tacit knowledge flows attenuate particularly
quickly in organizations that experience discontinuous membership. More debilitating
164
than personnel turnover—which has its own impact on impeding knowledge flows—
discontinuous membership involves the coming and going of organizational roles (i.e., in
addition to the people who play them) or positions during work performance (Ibrahim
and Paulson 2005). Both tacit and explicit knowledge are lost routinely in such
environments. And organizations that experience discontinuous membership find it
difficult to learn over time and across projects. Even more insidious, however, such
organizations find it difficult to remember even what they have learned before or what
they know already. Plus, far from a rare occurrence or freak phenomenon, discontinuous
membership affects a broad cross section of organizations, particularly those engaged in
project work (e.g., automotive design, building construction, military operations).
Unfortunately, little is known today about how to design organizations in a manner that
mitigates or obviates the impact of discontinuous membership and that amplifies
knowledge flows despite a dynamic mix of participating organizational roles.
The research described in this article builds upon an ethnographic study of facility
development to understand the knowledge-flow patterns and pathologies in an
organization that experiences severe discontinuous membership. Previous research has
shown facility development to be a rich domain for the study of knowledge-flow patterns
(Ibrahim and Nissen 2003). In particular, the creation, sharing and use of tacit knowledge
are critical during the early phases (e.g., feasibility and entitlements) of the facility
development life cycle process (Ibrahim et al. 2005a). Ibrahim (2001) describes this life
cycle as the sequence of process activities associated with developing a new facility, and
is comprised of five sequential phases: feasibility, entitlements, building permit,
construction, and property management. The feasibility phase starts as soon as a
165
landowner reviews a parcel with the idea of building. This first life-cycle phase ends
when the landowner formally applies for a development permit. The succeeding
entitlements phase begins with the formal application for development permit and ends
when construction commences at the site. Knowledge flows are particularly important
during these two early phases, when the majority of influential decisions are made
(Paulson 1976). Poor knowledge flows can cause severe project setbacks in terms of
rework and delays (Jin and Levitt 1996). Such severe impacts are common in the
construction industry among others.
For instance, Ibrahim and Paulson (2005) describe how a facility developer, using
his accumulated tacit knowledge and experience, agreed with a city’s authority to
maintain an oak grove at one corner of a property site planned for new development. A
few months into the facility development process, his building permit was rejected,
because the mechanical engineer had submitted a building plan that routed the water
piping system through this oak grove. From the perspective of the mechanical engineer
(ME), using his accumulated tacit knowledge and experience, this pipe-routing plan was
logical and reflected high professional competence. That corner was the location for all
major water in-take points to the site, and that route would be the cheapest since it was
the shortest.
In terms of knowledge flows, the organization had no process through which
discontinuous members such as the ME—that started participating later in the facility
development life cycle—could learn easily about commitments made by the developer—
who was both knowledgeable and experienced—in an earlier phase of the project. Notice
this knowledge-flow error involves more than a simple mistake of missed
166
communication. The developer, relying upon the architect as the main architectural-
engineering design coordinator and the ME’s professional competence and experience,
assumed that the oak grove preservation would be integrated automatically. The architect,
who provided the architectural documents to the ME, was present during the city
council’s meeting. Neither did the ME think to inquire about or read through a file of
development commitments. The architect’s document is usually taken as ‘the basis’ for
his professional services. Here, knowledge in one part of the life-cycle process (i.e.,
known by the developer) failed to flow to another (i.e., to the ME).
Even more disturbing than this costly mistake is the fact that similar knowledge-
flow problems are pervasive, even when the facility development organization has
explicit information or maintains one development project manager throughout the
facility development life cycle process (Ibrahim and Nissen 2005). Given the knowledge
possessed in different parts of the organization, at different times, by discontinuous
members, the developer could have known about the pipe-routing problem, and the ME
could have known about the oak grove commitment. But the ME did not. The
organizational design and knowledge-flow patterns precluded him from knowing. This
phenomenon causes us to question some tenets of absorptive capacity theory (Cohen and
Levinthal 1990; e.g., that an organization’s current knowledge builds upon its prior
knowledge). In our facility development example, the prior knowledge was possessed
and known in the organization (e.g., separately by the developer and the ME), and
explicit knowledge of both the oak grove commitment and main water connection point
existed (Ibrahim and Paulson 2005). But knowledge possession is insufficient apparently
167
to guarantee knowledge flows and appropriate action. And the existence of explicit
knowledge is insufficient apparently to guarantee that it will be used.
Through the balance of this article, we describe theoretical foundations for a
model of knowledge-based organizational performance where discontinuous membership
affects the organization. The following Section 5.3 reviews a focused segment of the
literature and presents how this study departs from present organization and knowledge-
flow theories. Section 5.4 explains how we can extend a computational organizational
modeling tool to emulate the dynamic behaviors of knowledge flows discovered through
investigation of the facility development life cycle process. We develop a computational
organization theory (COT) case study model in Section 5.5; and present the results,
analysis, and validation methodology in Section 5.6. Then, we discuss how we can
further develop additional contingency parameters for the design of organizational fit. We
conclude with key insights from the study, practical implications for the manager, and
recommendations for future research along the lines of this investigation.
5.3 Literature Review
Two areas of the literature are particularly important in this study of organizational
design with discontinuous membership: 1) information processing in organizations, and
2) knowledge-flow dynamics. We summarize each in turn.
5.3.1 Information Processing in Organizations
From the organization literature, we find much concentration on organization formation
and behavior (March and Simon 1958; Cyert and March 1963; Galbraith 1974 and 1977;
Mintzberg and Westley 1992; Scott 2003; Burton and Obel 2003). Most researchers focus
168
upon hierarchy as the basic structure for organizing complex social activity. Cooperation
among members is achieved through vertically imposed bureaucratic processes (Grant
1996; Weber 1947). Rules and programs to coordinate behavior between interdependent
subtasks are used (March and Simon 1958).
Galbraith’s (1974 and 1977) information-processing model is an extension of
contingency theory (Lawrence and Lorsch 1967). The efficacy of an organization
depends upon its fit with the environment and other contextual factors. Scott (2003)
explains that people and tasks are organized to attain specific goals reflecting a rational
view. The natural view perceives organizations first and foremost as “collectivities” with
many seemingly irrational aspects.
In this study, we focus in particular on the rational information-processing model
of organization (Galbraith 1974 and 1977). But we incorporate some irrational inputs of
natural systems. Ibrahim and Paulson’s (2005) ethnographic study demonstrates that the
complex facility development life cycle (assumed as impossible to standardize in Carillo,
et al 2004) can be rationalized with the understanding of several sequential or concurrent
workflows combined with the identification of critical convergent points. The key to
integrating the natural irrational environmental factors is by ensuring the linkages of
interdependent tasks among the multiple workflows (Grant 1996).
Galbraith (1974 and 1977) proposes that decision-makers need to process
information well during exception-handling for the organization to perform well.
Galbraith argues that the greater the task uncertainty, the greater the amount of
information that participants in an organization must process. Similarly, when an
organization faces greater uncertainty—such as in the facility development’s operating
169
environment—its members face increasingly frequent situations for which they have no
sufficient organizational routines (Nelson and Winter 1982 in Scott 2003) or rules to
guide their decision-making or to inform their work performance. In such an
environment, and when the structural hierarchical information-processing is in place, the
norm is for lower-ranked staff to seek guidance or information from their supervisors to
handle these exceptions. The ME in the oak grove case, for example, had to refer to the
architect who was the design leader to inform him of the building approval’s rejection.
But, he was instructed instead to revise the ME’s documents to meet the city’s
requirement.
Burton and Obel (2003) extend this contingency perspective through development
and articulation of six contingency factors to guide organizational design: management
style, climate, size/ownership, environment, technology, and strategy. This extension of
contingency theory has been formalized via a scholarship-based expert system (Nissen
2005c) called Organization Consultant (Burton and Obel 2003) that employs automated
inference to support theory-based design of organizations. Despite years of successful
description and explanation, however, the structural emphasis during information-
processing is becoming inadequate to explain the emerging knowledge network
workforce (Monge and Contractor 2003; Ibrahim et al. 2005a; Ibrahim et al. 2005b).
Monge and Contractor (2003) rationalize studying emergent communication networks
when findings are inconclusive in relating formal organizational structure to
organizational behavior. In addition, Ibrahim et al. (2005a) find different knowledge
types affecting knowledge-flow behaviors differently. Ibrahim et al. (2005b) provide
early empirical evidence that informal knowledge networks do affect the organizational
170
performance of an enterprise. Hence, integrating horizontal (or non-hierarchical)
information-processing structure for decision-making is of particular interest to us.
Recent scholars such as Lambert and Shaw (2003) have turned to Wegner’s
(1987) transactive memory theory to explain the existence of this non-hierarchical
information-processing structure. Using methods and tools of computational organization
theory (COT), the Ibrahim et al. (2005b) study shows how such non-hierarchical
information processing affects organizational performance. Their study establishes that a
new member in a discontinuous membership organization, who uses his inaccurate
cognition of his other team members to complement his incomplete cognitive skill, can
expose his task to higher risk of failure, which in turn can be detrimental to his overall
team’s organizational performance in the long run. It illustrates the need to consider non-
hierarchical knowledge networks as part of organization design for better organizational
performance.
Wegner (1987) also describes transactive memory as a shared system for
encoding, storing, and retrieving information. The three key processes of a transactive
memory system are (a) directory updating, where people learn what others are likely to
know; (b) information allocation, where new information is communicated to the person
whose expertise will facilitate its storage; and (c) retrieval coordination, which is a plan
for retrieving needed information on any topic based on knowledge of the relative
expertise of the individuals in the memory system.
Wegner (1987) and later Moreland (1999) pose that organizational teams may act
like large brains, where individuals store information to be combined with others’
information. When individuals in that team need information, they go to the “expert
171
node” or expert member. In other words, exceptions are referred to the perceived
“expert” in the organization, regardless of such member’s hierarchical position in the
organization.
In summary, there is an apparent lack of consideration for non-hierarchical
information-processing during organization design. Moreover, scholars have yet to
consider the discontinuous membership characteristic of an organization.
5.3.2 Knowledge-flow Dynamics
Knowledge management is an increasing concern in the design of organizations. Alavi
and Leidner (2001) note an abundance of literature on knowledge creation, storage and
retrieval. Other scholars (e.g., Carlile and Rebentisch 2003) similarly note the emphasis
on knowledge storage and retrieval in the literature. Additionally, knowledge transfer and
sharing are becoming increasingly important for scholars attempting to explain dynamic
flows of knowledge that enable workflow processes (and hence organizational
performance; e.g., see Carlile and Rebentisch 2003, Nonaka and Takeuchi 1995, Nissen
2002). Some scholars also look at knowledge transfer in organization learning (such as
Nadler, Thompson, and van Boven 2003), and how knowledge exchange affects the
social and expert status of individuals (Thomas-Hunt, Ogden, and Neale 2003).
Kogut and Zander (1992) pose that firms are more successful in transferring
knowledge within organizations than between organizations. Although individuals hold
knowledge, it is also expressed in regularities by which members cooperate in a social
community. Building upon this concept, Nonaka (1994) argues that organizational
membership plays a critical role in articulating and amplifying knowledge. He proposes
four modes of knowledge transfer—socialization, externalization, combination, and
172
internalization (SECI)—in a dynamic spiral of interactions between knowledge type
(termed epistemological; e.g., tacit, explicit) and organizational reach (termed
ontological; e.g., individual, inter-organizational).
Nissen (2002) extends Nonaka’s work by integrating six stages of a knowledge-
flow life cycle process: 1) creation or acquisition, 2) organization, 3) formalization, 4)
distribution or sharing, 5) application, and 6) evolution or refinement. This six-step
knowledge life cycle is an amalgamation of earlier views of knowledge life cycle (e.g., as
proposed by Davenport and Prusak 1998, Depress and Chauvel 1999).
Von Hippel (1994) uses the term ‘stickiness’ to describe how enabling tacit
knowledge can ‘stick’ with problem-solving capabilities in different locations. Stickiness
connotes the difficulty experienced with transferring tacit knowledge, for example, in
which an organization recreates and maintains a complex, causally ambiguous set of
routines in a new setting (Szulanski 2000).
In Nissen’s later work (i.e., Nissen 2005a), he states that new organizational
forms may obtain and even dominate through a focus on dynamic knowledge flows.
Nissen provides discrete qualitative categories for potential operationalization of
knowledge flows in the enterprise. Building further upon Nonaka (1994), Nissen’s four
knowledge flow dimensions are type of knowledge (e.g., tacit, explicit), level of
socialization associated with the knowledge (e.g., individual, group, organization, inter-
organization), life cycle activities of knowledge work (e.g., create, share, apply), and flow
time (e.g., lengthy, brief).
In order to understand knowledge creation by individuals, Grant (1996)
conceptualizes that the firm is an institution for integrating knowledge at the next
173
organization level. Grant attempts to devise mechanisms for integrating individuals’
specialized knowledge. He proposes four mechanisms to coordinate the integration of
knowledge within an enterprise: (a) having rules and directives to enable the conversion
of tacit knowledge to explicit knowledge; (b) sequencing of workflow processes that
minimize communication but ensure the input of expertise at different times; (c) creating
routines to support complex patterns of interactions between individuals in the absence of
rules, directives, or even significant verbal communication; and (d) establishing group
problem-solving and decision-making routines.
Grant’s resulting knowledge-based theory of the firm has implications for the
basis of organizational capability, the principles of organization design (in particular, the
analysis of hierarchy and the distribution of decision-making authority), and the
determinants of horizontal and vertical boundaries of the firm. His knowledge-based view
perceives interdependence as an element of organizational design and the subject of
managerial choice rather than exogenously driven by the prevailing production
technology. Grant emphasizes knowledge application and the role of the individual as the
primary actor in knowledge creation and the principal repository of knowledge. However,
he also calls for further research needed on the knowledge-based theory of the firm that
can embrace knowledge creation and application.
Cohen and Levinthal (1990) argue that the ability of a firm to recognize the value
of new, external information, to assimilate it, and to apply it is critical to its innovative
capabilities. They term this capability a firm’s absorptive capacity and suggest that it is
largely a function of a firm’s level of prior related knowledge. Cohen and Levinthal
describe the basis for an individual’s absorptive capacity to include: 1) development of
174
specific knowledge cognition, 2) development of experience and skills, and 3)
development of problem-solving skills. They refer to the prior possession of knowledge
before any type of individual development. However, there is a distinction between an
individual’s absorptive capacity versus that of an organization. The absorptive capacity of
an organization is not simply the sum of every employee’s knowledge. Rather, it includes
the organization’s ability to exploit such knowledge. Absorptive capacity at the
organizational level depends on the organization’s direct interface with the external
environment. It depends also on transfers of knowledge across and within subunits that
may be quite removed from the original point of knowledge entry or creation. This aspect
refers to how the organization—a group of individuals—perceives the inherent tacit
knowledge from its environment and gains from that experience.
To study a firm’s absorptive capacity, Cohen and Levinthal focus on the structure
of communication between the external environment and the organization, as well as
among the subunits of the organization. They focus also on the character and distribution
of expertise within the organization. Both scholars argue that the development of
absorptive capacity, and in turn innovative performance, are history- or path-dependent.
They argue how lack of investment in an area of expertise early on may foreclose the
future development of a technical capability in that area. Such is the case with our oak
grove preservation example cited above. The facility development life cycle process
faltered in part because the facility owner did not invest in a mechanical engineer during
the feasibility-entitlements phase.
In summary, we note an interest in the role of knowledge transfer in
organizational learning, an increasing need for literature in the area of tacit knowledge
175
transfer, and a call for empirical, theoretical development in the study of knowledge
management. This study contributes towards the development of theory in knowledge
transfer within the tacit realm involving discontinuous organization.
5.4 Modeling Knowledge Flows Computationally
The utilization of computational models to represent and simulate dynamic knowledge
flows through and between organizations is an emerging trend in knowledge management
research. For instances, Nissen and Levitt (2004) and Ibrahim and Nissen (2004) conduct
such studies linking an organization with its processes to investigate the dynamics of
knowledge flow. Prior computational organizational theory (COT) research has examined
the dynamic work processes and information flows associated with project- or task-based
organizations only (such as Carley and Prietula 1994, Levitt et al. 1994). Utilizing such
dynamic computational models can provide insight into unique operating environments
that are difficult (or even impossible) to test in real life. In addition, computational
techniques can be used early in the facility development life cycle, to examine and
compare alternate workflow approaches and organizational designs before committing
time and money to any specific option. This can serve to reduce a project’s risk while
increasing its knowledge. In this section, we describe the complex facility development
process we are studying, explain how we theoretically integrate knowledge-flow theory
in a COT model, and explain the choice of COT tool for our study.
176
5.4.1 Facility Development Characteristics
The facility development process is characterized as 1) having multiple concurrent and
sequential workflows, 2) having discontinuous membership, 3) having multiple task
interdependencies, and 4) displaying different knowledge forms (i.e., tacit- or explicit-
dominant knowledge areas). Recall from above how Ibrahim (2001) illustrates the division
of the facility development process into five sequential phases: 1) feasibility, 2)
entitlements, 3) building permit, 4) construction, and 5) property management phases.
Ibrahim and Paulson (2005) later combine the two early phases into a single feasibility-
entitlements phase for better clarity. Our study focuses on this combined feasibility-
entitlements phase and includes the activities associated with obtaining a building permit.
Together these three, early stage life cycle phases comprise the pre-construction
workflow.
Our computational modeling effort is guided by both empirical and theoretical
research. Specifically, we compare the ethnographic results from Ibrahim and Paulson’s
(2005) study with Burton and Obel’s (2003) contingency theory factors. The former
ethnographic results provide a current, empirical case for application of the latter,
theoretical prescriptions about organizational design. For instance, drawing from theory,
our facility development workflow and organizational characteristics are prescribed to be
set up in conformance with a set of organizational design propositions: if the
organizational environment has high complexity, high uncertainty, and high equivocality,
then the organizational design should reflect low formalization (e.g., few rules for
coordination and control), low organizational complexity (e.g., low horizontal and
177
vertical differentiation), and low centralization (e.g., infrequent direct involvement by top
management in decision making).
Several aspects of the facility development process are consistent with such
theory. Because the facility development environment demands multiple, concurrent and
sequential workflows—i.e., five sequential processes plus two concurrent processes—it
illustrates a high complexity environment with multiple interdependency links.
Examination of this process in the field reveals many uncertainties, stemming principally
from non-controlling, decision-making processes that involve extra-organizational, public
and governing authorities (such as, whether or not the development project will obtain a
funding program or obtain a permit approval). Decision processes such as these can
postpone progress through the facility development life cycle or render it infeasible to
continue a particular project. Equivocality is evident too, often caused by ad-hoc,
external, random and unpredictable requests to accommodate certain public and authority
conditions that influence the design and its process.
For instance, consider a funding program that requires a play structure but was not
called out in the original project-planning brief. If no space for such structure is allocated
in the design, then the design team will be required to revise the site plan to include this
new element. Environmental characteristics such as these make facility development a
high-risk endeavor. The non-profit developer in our case project successfully developed
and completed only fourteen percent of its projects annually.
In terms of the theoretical prescription above (i.e., for low formalization,
organizational complexity and centralization), our fieldwork suggests a mix of consistent
and inconsistent aspects of the organizational design propositions. Due diligence, an
178
analytical process, provides a good example, because project managers attempt to
formalize this process but have yet to come up with a successful model of formalization.
Project managers studied by Ibrahim (2001) would go through a long checklist to come
up with the best development option for a property with a view of purchasing it for
development. Major items on the due-diligence checklist include specific legal entities of
the property, all governmental requirements, details of development schedule involving
all major tasks duration and procedures, determination of the target market, program brief
for design, required governmental fees, and review of available documents.
Theory predicts also that the facility development will be better off with low
vertical differentiation and centralization. With top management’s accumulated tacit
knowledge, the due-diligence items would eventually influence the design, finance, and
target market decisions for the development project. The process can take as little as two
weeks to as much as three months to complete. Experience suggests it is effective to
allow the project managers unlimited authority to ‘create’ the best development approach
for the developer. In such a case, the development proposal gives good financial return, is
realistically feasible to construct, and has community acceptance. Low vertical
differentiation and centralization contribute to this effectiveness. The project manager
orchestrates all consultants under him, while he only reports to one executive manager.
Theory predicts further a tendency by managers to get overloaded in such
organizations. However, Ibrahim and Paulson (2005) note that the more experience
project managers have, the more efficient and creative they become. To mitigate the
effects of management overloading as suggested by theory, a team of managers and
others—from multiple, matrix organizations—tend to work together on several work
179
processes concurrently (e.g., design, financing, and asset management). Some members
of one functional team may serve on another functional team, such as the project
managers or the architects. But they have to divide their allocated time commitments
between several functional teams to achieve different functional goals.
In this field study, we find discontinuous membership through various positions
and organizational roles that come and go at different phases of the facility development
life cycle. Certain organizational roles participate—for some periods of time—in some
matrix organizations and workflow processes but do not participate—or participate at
other times—in other organizations and processes. Current theory has little to say about
organizing for discontinuous membership as such. Instead, our current understanding of
discontinuous membership derives principally from empirical work.
For instance, Ibrahim and Paulson (2005) discuss how discontinuous membership
occurs when a particular workflow process requires a specific mix of skills for task
performance. In a matrix organization with discontinuous membership, the same person
(or role) may contribute different skills when joining different organizational matrices.
The effect of joining several matrices within a larger workflow process forces members
to divide their total time commitments across the various, different matrices in which
they participate—either concurrently or often at different times. Hence from the
perspective of any, individual matrix organization, such person (or role) participates
intermittently (i.e., membership is discontinuous).
Additionally, theory suggests use of rich media. The study finds a mixture of both
rich and poor media utilization. Media richness indicates the form, amount, and kind of
information, where information richness is defined by Daft (1992 in Burton and Obel
180
2003, p.286) as the information-carrying capacity of data in the following order (from
richest to poorest): 1) face-to-face, 2) telephone and other personal electronic media, 3)
letters, notes, and memos, and 4) bulletins, computer reports, and data reports.
Through the fieldwork referenced above, we find abundant media-rich sources
during the feasibility and entitlements phases. For instance, project managers have
numerous, repeated meetings to gauge and obtain accurate understanding of certain
public or authority requirements. The goal is obtaining informal consent prior to
investing further into the project. For instance, through fieldwork we identify several
occasions when a facility developer would abandon the project after failing to obtain the
majority of a city council’s vote prior to submitting his proposal.
However, as the project progresses, the study finds less and less media richness,
even as more experts participate—via discontinuous membership—on the design team.
For instance, a mechanical engineer will provide heating, ventilation, and air-
conditioning data, while an electrical engineer will provide lighting and energy data to
the design workflow. Indeed, more and more low-richness media are used as the facility
development life cycle progresses. By the final, property management phase, facility
developers tend to use only computational databases for report making. This kind of
longitudinal regression of media richness through the course of a facility development
process provides novel insight into how choices of communication media impact
knowledge flows.
Moreover, theory suggests providing result-based incentives. However, Ibrahim
and Paulson (2005) find differing incentive schemes (and hence different goals) among
the design team versus the finance team, for example. On the one hand, the design team
181
aims for completion of the design documentation for their professional fees. Once the
documents are complete and their fees have been collected, design team members have
little incentive to be concerned about seemingly unrelated goals such as project funding;
they get paid for professional services whether the project is funded or not. On the other
hand, the finance team aims at obtaining project funding for the sake of the facility
project’s survival. Finance is crucial because, without funding, the project will be
abandoned. It is willing to comply with—at times very costly—additional conditions that
increase the overall project’s cost by complicating the design and requiring rework from
the design team.
Overall, we find mixed empirical support for the theoretical prescriptions noted
above. Although the facility development process takes place within an environment of
high complexity, uncertainty and equivocality—and reflects the kinds of low vertical
differentiation and centralization prescribed theoretically also—we find in contrast: an
attempt at relatively high formalization; project management experience ameliorating
overload; multiple, concurrent matrix organizations; a longitudinal regression of media
richness; and a mixture of incentive schemes (and hence goals) across various process
participants. Such latter empirical findings run counter to the theoretical prescriptions
summarized above. Further, we find an empirical case of discontinuous membership,
about which extant theory has little to say at present.
5.4.2 Integrating Knowledge-Flow Theory in COT
Extant COT tools do not model the behaviors of knowledge flows well. To study
dynamic knowledge flows in the facility development process, we need to identify and
select a COT tool that can accommodate adaptation to represent dynamic knowledge in
182
an organizational context. To accomplish this, we draw from and build upon Nissen’s
(2002) Vertical and Horizontal Processes Model to provide a foundation for such
adaptation. This model characterizes the powerful interaction between flows of work and
the enabling flows of knowledge in an enterprise. But this model is focused internally and
does not examine explicitly the environment within which an organization operates. In
particular, it lacks the concepts necessary to characterize an organization that experiences
discontinuous membership (details to follow below).
Ibrahim and Nissen (2005) extend Nissen’s model to conceptualize a new
analytical unit: Knowledge Group Set (KGS). A KGS unit is a group of interrelated,
concurrent, horizontal processes that are required to accomplish common goals along
vertical processes. Here, horizontal processes refer to workflows, whereas vertical
processes refer to knowledge flows. See Figure 5-1 for illustration. From an internal
perspective, each horizontal workflow process may appear to be independent. But when
viewed as a group, the set of workflows contains one or more, interdependent tasks that
link and interrelate concurrent processes within a common time frame. Hence an
organization manager—or set of managers—must be responsible for the performance of
multiple, interrelated processes, which are all linked directly to accomplishment of
common goals. This represents a complex case for organizational design. With multiple,
interrelated processes and managers, one may find multiple, unrelated organizational
structures that require integration.
For example (refer to Figure 5-1), consider a workflow for pre-construction
(labeled “Workflow Process 1”), the goal of which (labeled “G1”) is obtaining a building
permit to commence construction on a property. Another, concurrent workflow for
183
finance (labeled “Workflow Process 2”) has a different but related goal (labeled “G2”) of
obtaining the necessary construction loan to finance the building construction. Notice
how some tasks in each workflow require input from one another.
Figure 5-1: A Knowledge Group Set (KGS) unit consisting of two or more horizontal
workflow processes with task interdependency links during knowledge life cycle period
(Adapted and revised from Nissen’s Horizontal and Vertical Processes Model
(Nissen 2002, Fig. 3)).
For instance, knowledge obtained through performance of Task 1 (labeled T1) at
the top workflow process provides a critical input to performance of the first task of the
bottom workflow; knowledge from Task 2 of the second workflow is required for
performance of Task 3 in the first workflow; Task 4 knowledge from the top feeds Task 4
work on the bottom; and so forth. Notice also, the goals associated with each workflow
are linked in the figure. For instance, Goal 1(labeled G1) is required to achieve Goal 2.
184
The KGS concept captures the kinds of interdependent and concurrent knowledge flows
and workflows depicted in Figure 5-1 and identified in the facility development process.
More specifically in the facility development process, upon obtaining site control
from a property owner, the finance team sets out to seek potential investors in the project.
Meanwhile, the concurrent design team proceeds with its design work based on the initial
project brief provided by the facility developer. In negotiating the financing support, for
instance, an educational funding program may require a computer cluster within a family
housing project to qualify. Hence, the facility developer would instruct the design team to
integrate a computer cluster in the community center. In Workflow Process 1, the design
team succeeds in obtaining the building permit (i.e., accomplishing G1), after integrating
the additional computer cluster. In Workflow Process 2, the finance team succeeds in
obtaining the construction loan (i.e., accomplishing G2) after the design team includes a
computer cluster in the design proposal. G1 has to be completed first before G2.
Notice that there are different skill sets requirements for the different teams. In the
design team, there is a strong component of building-related professionals such as
architects, engineers, and builders, whereas the finance team requires complementary
knowledge such as accounting, finance and law, along with familiarity with the building
process.
Further, Ibrahim and Paulson’s (2005) study highlights that some members of one
team may be concurrent members of one or more other teams, too. Such members, who
participate concurrently in multiple matrix organizations, have to divide and allocate their
time across the various matrix organizations in which they participate. This division and
allocation of members’ time across different matrix organizations illustrates clearly an
185
effect of discontinuous membership. And by capturing the relative percentages of time
that various members spend participating in the different organizations, we discover an
approach to representing discontinuous membership—along with its knowledge effects—
via computational models.
We can further combine several KGS units to form a larger knowledge flow
framework as depicted in Figure 5-2. Here we illustrate six, concurrent, interdependent
workflows (i.e., horizontal processes) that feed into three, sequential KGS units (i.e.,
vertical arrows labeled “KGS 1,” 2, & 3). Each KGS corresponds to one phase of the
facility development life cycle. The diamond shapes in the figure represent (possibly)
different matrix teams; that is, the membership (i.e., organizational positions or roles) of
the various matrix teams changes—both along the flows of work (i.e., horizontally) and
across the life cycle phases (i.e., vertically)—through both space and time (e.g., they may
be geographically distributed and temporally distinct). We refer to such compositions of
interrelated KGS units as Knowledge Group Set (KGS) Flow Models.
More specifically in the facility development process, we further break the facility
development life cycle process into smaller phases. One KGS unit represents a phase in
the facility development process. Based on Ibrahim’s (2001) facility development life
cycle, KGS1 in the figure represents the feasibility phase, KGS2 represents the
entitlements phase, and KGS3 represents the building permit phase. Knowledge from the
feasibility phase (i.e., KGS1) flows into the succeeding entitlements phase (i.e., KGS2).
The goal for KGS1 is submitting the application for the development permit, while the
goal for KGS2 is obtaining the development permit. When it transits into the next phase,
186
the matrix team combination may change because of different skill sets required in the
different workflows.
Figure 5-2. The Knowledge Group Set (KGS) Flow Model during
a facility development process.
The mechanical engineer and the electrical engineer, for instance, usually join
during the building permit phase (i.e., KGS3) because the building permit submission
requires a document that illustrates compliance of providing a healthy and safe facility
project. Such new, engineering experts provide inputs on heating, ventilation, air-
conditioning, electrical needs, and like knowledge-based tasks. The goal at the end of this
KGS3 unit is start of construction. Similarly, the flow of knowledge from one phase to
the next is represented by the downward arrow. Such knowledge flows require
considerable time—normally several years—to complete the transit across all life cycle
187
phases. Knowledge flows also cross considerable geographical space, as multiple matrix
team members participate, often from many different firms. The KGS enables us to
model such phenomena identified through fieldwork.
5.4.3 A COT Tool
We utilize the Virtual Design Team (VDT) computational tool to model and emulate
knowledge flows. The VDT Research Program is well established as a planned
accumulation of collaborative research over two decades to develop rich, theory-based
models of organizational processes (Nissen and Levitt 2004). VDT uses an agent-based
representation (Cohen 1992; Jin and Levitt 1996; Kunz et al. 1998) that incorporates into
the computational tool research on micro-level organizational behaviors, which were
formalized to reflect well-accepted organization theory (Levitt et al. 1999). This
contributes toward VDT external validity and generalizability. Extensive empirical
validation projects (e.g., Christiansen 1993) contribute further toward such validity and
generalizability.
However, no extant computational tool is perfect for our task. And Nissen and
Levitt’s (2004) study highlights some limitations of VDT. Unlike mathematically
representable and analyzable micro-behaviors of physical systems, the dynamics of
organizations are influenced by a variety of social, technical and cultural factors. They
are difficult to verify experimentally, and are not amenable to numerical representation,
mathematical analysis or precise measurement. Nissen and Levitt expect ambiguity when
people and social interactions—distinct from physical systems—drive the behavior of
organizations. Nonetheless, VDT provides a capability to represent—statically and
dynamically—the key structures and behaviors of a knowledge group set flow model. We
188
use SimVision®—the commercial version of VDT—to model, at a relatively high level,
the complex facility development process described above.
5.5 A COT Case Study Model
This section explains the representation of the feasibility, entitlements, and building
permit phases in a KGS Flow Model in SimVision®. This KGS Flow Model represents
high-level activities in two concurrent workflow processes: the three sequential phases—
feasibility, entitlements, and building permit—that comprise the pre-construction
workflow, and the finance workflow that is accomplished concurrently. Our COT case
study is a 43-unit affordable family housing development for farm workers located in the
San Francisco Bay Area. This facility development has been in operation since June
2001, but has been plagued with civil and wastewater problems since construction began.
This discussion begins with specification of a baseline model, which depicts the
organization and process under conditions of continuous membership. Recall from above,
such conditions are consistent with theory but inconsistent with our empirical
observations. We then specify an Alternate model, which depicts the organization and
process under conditions of discontinuous membership. Comparing the relative behavior
and performance of these alternate, computational models enable us to isolate the effects
of discontinuous membership and to examine how it impacts the performance of a
complex process. A brief description of the simulation approach follows and completes a
transition to results and analysis in Section 5.6.
189
Baseline Model. Our baseline model represents high-level activities of the COT case
during pre-construction. As described above, there are two process workflows running
concurrently: the three sequential phases of the pre-construction workflow, and the
concurrent finance workflow. Together these two workflow processes consist of 39 tasks
with twelve milestones. The pre-construction workflow has three company staff members
and six external consultants. The finance workflow has four company staff members and
two external consultants. The start date for both workflow processes is June 2, 1997,
when the developer first obtained control of the property.
Figure 5-3 presents a screenshot from the SimVision® modeling tool showing the
finance organization and workflow as represented in terms of the KGS unit discussed
above. The finance matrix team consists of four owner’s representatives and two external
consultants. Tasks are represented by the rectangular boxes, which are linked sequentially
to other tasks or milestones (symbolized by the diamond boxes) by precedence links
(solid lines in the figure). Figure 5-4 shows similarly the pre-construction organization
and workflow in KGS terms.
To represent knowledge flows, we establish task interdependencies between the
pre-construction and finance workflows by having four ghost connectors between them.
This applies to both the Baseline and Alternate Model. Ghost connectors provide
modeling connections or constraints between the workflows by mimicking tasks and
milestones from one workflow to another. They allow linkages to multiple, other
milestones and tasks. The ghost connectors are used to represent flows of explicit
knowledge from one concurrent workflow to another. There is also a ghost
190
communication link that reflects communication between the pre-construction stage and
the finance teams. It is used to represent tacit knowledge flows.
End Entitlements
Phase End BuPermit
Start Feasibility
Phase
Figure 5-3. Network diagram of the concurrent f
191
Owner’s
ilding Phase
inance phas
External Consultant
e in Baseline Model.
Feasibility Phase
Entitlements Phase
Building Permit Phase
Owner’s
ConsultantExternal
Figure 5-4. Network diagram of the sequential feasibility, entitlements, and
building permit phases representing the pre-construction stage in Baseline Model.
We draw from our fieldwork to parameterize the SimVision® model. Model
variables and parameter settings are summarized in Table 5-1. These parameter settings
reflect well-established norms for specifying SimVision® models for construction
projects (see Jin and Levitt 1996). We describe the variables and parameters based on
192
SimVision®’s definitions here. Centralization reflects whether decisions are made by
senior project positions or decentralized to individual responsible positions. With low
centralization, responsible positions tend to make their own decisions and there is thus
less communication required. Team experience is a measure of how successfully the team
has performed related projects. The team experience value contributes (inversely) to the
amount of formal communication that is required between team members with
interdependent tasks. Other factors are the position's own application experience, skill
set, skill levels, and the task's requirement complexity. Formalization is a measure of
how formal the communication is in an organization. High formalization means
communication tends to occur in formal meetings. With low formalization, it's more
common for communication to occur informally between positions. Matrix strength is the
extent to which positions are located in skill-based functional departments and supervised
directly by functional managers (Low) or co-located with other skill specialists in
dedicated project teams and have project supervision from a project manager (High).
Medium matrix strength means that workers make approximately equal amounts of
formal and informal communications.
Table 5-1: Variables and Parameter Settings for Baseline Model
VARIABLES PARAMETER Centralization LOW Team Experience MEDIUM Formalization MEDIUM Matrix Strength MEDIUM Information Exchange Probability 0.7 Noise Probability 0.2 Functional Error Probability 0.05 Project Error Probability 0.05 Work Volume per Full Time Equivalent 8 hours/ FTE Work Days per Week 5 days/ week
193
Information exchange probability measures the level of communication in the
project between positions that are responsible for interdependent tasks with
communications links. Noise probability is a way to measure the effect of interruptions in
the ordinary working day that take time away from doing the project tasks. Functional
error probability is the probability that a task will fail and require rework. Functional
errors are errors that are localized to a task and cause rework only in that task by the
responsible position. Project error probability is the probability that a task will fail and
generate rework for all dependent tasks connected to it by rework links. Work volume is
the predicted time that all positions on a project spend doing direct work. We set the FTE
to an 8-hour per day of direct work duration.
Table 5-2 summarizes the staffing positions we build into the Baseline Model
based on our fieldwork. In Column 1, the stakeholders are divided into two categories:
owner and consultants-builder. The position title is listed for each stakeholder in Column
2. Each position listed in Column 3 is assigned a full-time equivalent (FTE) value in
Column 4. All of these positions and FTE values reflect empirical data collected through
fieldwork. Hence the COT model matches the field organization exactly in this regard.
For example, the Executive Director devotes 0.30 FTE to the project workflow. This is
the FTE level reported by the Executive Director in the field.
194
Table 5-2: Total Staffing and Position FTE’s for Baseline Model
STAKE-HOLDER
TYPE OF STAKEHOLDER
POSITION TITLE IN SIMVISION®
TOTAL POSITION
FTE's EXEC DIRECTOR 0.30 PROJECT MANAGER 0.40 DESIGN-CONSTRUCTION MANAGER 0.50 PROPERTY DIRECTOR 0.10
OWNE
R
DEVELOPER OWNER
SERVICE DIRECTOR 0.10 FINANCE CONSULTANT FINANCE ADVISOR 1.00 LEGAL CONSULTANT LEGAL ADVISOR 1.00 ENVIRON MENTAL CONSULTANT ENV STAFF 1.00 GEOTECH CONSULTANT GEOTECH STAFF 1.00 CIVIL ENGINEER CIVIL ENGINEER 1.00
PROJECT ARCHITECT 1 PROJECT ARCHITECT 2
ARCHITECT
CONCEPT ARCHITECT
1.00
GEN CONTRACTOR 1 0.15
CONS
ULTA
NTS-
BUILD
ER
GENERAL CONTRACTOR
GEN CONTRACTOR 2 1.00
Alternate Model. Our Alternate model represents a discontinuous organization. It is
derived directly from the baseline above. Indeed, we duplicate the SimVision® variables
and parameter settings from the Baseline Model. The key differences are in the
distribution of FTEs for staffing each position across different organizational matrices in
the Alternate Model. The matrix combination reflects the different skill sets each
workflow requires to perform its tasks. This is how we represent discontinuous
membership. The three matrices based on our field observation are City, Building, and
195
Owner. The skill set of the City Matrix is expertise to consolidate public and financing
support from the local jurisdictions. The skill set of the Building Matrix is the expertise to
consolidate the planning, design, and technical aspects regarding the facility proposal in
order to ensure compliance to build. The skill set of the Owner Matrix is the expertise to
coordinate the developer’s activities pertaining to the development proposal.
The FTE’s distribution is summarized in Table 5-3. The positions, roles and total
FTE levels are identical to those reported above in Table 5-2 for the Baseline Model. But
note the middle columns in Table 5-3 that summarize the distribution of FTEs across the
three, different matrices. This is how we allocate staffing to the respective positions
according to involvement in the different organizations. For instance, the Executive
Director staff role is listed in Table 5-3 with a total FTE of 0.30. This is the same FTE
level listed for this role in Table 5-2 above. However, notice the distribution of the
Executive Director’s time (in FTEs) across two, different organizations: 0.20 in the City
matrix, and 0.10 in the Owner matrix. While the Executive Director is participating in the
City matrix, such an organization receives the benefits of his or her knowledge as well as
the work it enables. But such knowledge and work are not available continuously. The
same applies to the other positions and staff roles.
196
Table 5-3: Distribution of FTE’s for Team Members in City, Building,
and Owner Matrices
MATRIX FTE's STAKE-HOLDER
TYPE OF STAKEHOLDER
POSITION TITLE IN SIMVISION® CITY BUILDING OWNER
TOTAL STAFFING
FTE's
TOTAL POSITION
FTE's EXEC DIRECTOR 0.20 0.10 0.30 0.30 PROJECT MANAGER 0.20 0.10 0.10 0.40 0.40 PRE-CONSTRUCTION PHASES MANAGER 0.10 0.40 0.50 0.50 PROPERTY DIRECTOR 0.10 0.10 0.10
OWNE
R
DEVELOPER OWNER
SERVICE DIRECTOR 0.10 0.10 0.10 FINANCE CONSULTANT FINANCE ADVISOR 1.00 1.00 1.00 LEGAL CONSULTANT LEGAL ADVISOR 1.00 1.00 1.00 ENVIRON MENTAL CONSULTANT ENV STAFF 1.00 1.00 1.00 GEOTECH CONSULTANT GEOTECH STAFF 1.00 1.00 1.00 CIVIL ENGINEER CIVIL ENGINEER 0.20 0.80 1.00 1.00
PROJECT ARCHITECT 1 0.20 0.20 PROJECT ARCHITECT 2 0.20 0.20
ARCHITECT
CONCEPT ARCHITECT 0.80 0.80
1.00
GEN CONTRACTOR 1 0.05 0.10 0.15 0.15
CONS
ULTA
NTS-
BUILD
ER
GENERAL CONTRACTOR
GEN CONTRACTOR 2 0.10 0.90 1.00 1.00
Simulation. We run a Monte Carlo simulation of 100 cases for both the Baseline and
Alternate Models in SimVision® and compare their organizational performance results.
Performance measures include simulated duration, critical path method (CPM) duration,
total work volume, cost, functional risk index, project risk index, and communication risk.
The steps in building the COT model provide considerable accuracy in terms of
197
emulating the facility development organization and process behaviors. It can provide us
retrospective validation, which Thomsen et al. (1999) recommend for validating
computational emulation models for organizations. To obtain a retrospective validation
for both models, we require the two models to maintain the construct characteristics from
the ethnographic study, i.e., having multiple concurrent or sequential phases, having task
interdependencies, having discontinuous memberships, and displaying different
knowledge forms during the process.
5.6 Results, Analysis, and COT Validation
This section presents key simulation results for both Baseline and Alternate Models. We
summarize the results of the Baseline and Alternate Models in Table 5-4. The first
column includes the seven performance measures used to summarize and report the
results. The second column includes values for the Baseline Model that represents an
organization with continuous membership. The last column includes values for the
Alternate Model that represents an organization with discontinuous membership. The
first three measures depict simulated and CPM durations, and total work volume in days.
The latter three measures depict different dimensions of risk (and hence process quality).
A cost measure is included.
198
Table 5-4: Statistics of selected simulated values for Baseline and Alternate Models
VALUE NAMES BASELINE ALTERNATE Simulated Duration (days) 695 760 Total Work Volume (days) 735 715 CPM Duration (days) 669 674 Cost $418K $442K Functional Risk Index (FRI) 0.54 0.75 Project Risk Index (PRI) 0.59 0.48 Communications Risk 0.38 0.38
Note: Bold values represent comparatively better outcomes in SimVision®.
The simulated duration of the Baseline Model is shorter than that of the Alternate
Model (i.e., 695 vs. 760 days). This measures a schedule penalty associated with
discontinuous membership in the Alternate Model. When specific knowledge and skills
are needed but unavailable in a particular organizational matrix, progress on the facility
development project as a whole can lag behind. But the Baseline Model has greater total
work volume than the Alternate Model does, (i.e., 735 vs. 715 days). This measures an
economic benefit associated with discontinuous membership in the Alternate Model.
With several people (or roles) participating discontinuously in one or more organizational
matrices, it reflects leaner staffing overall for the project. We note that both Baseline and
Alternate Models have almost identical CPM duration values (i.e., 669 vs. 674 days).
This measures negligible effect by discontinuous membership on the overall schedule.
In terms of cost, the Baseline Model incurs lesser cost than that of the Alternate
Model (i.e., $418K vs. $442K). This measures an economic penalty associated with
discontinuous membership in the Alternate Model. The costs involved are those
pertaining to the professional fees, and the salaries of the developer’s employees during
the pre-construction stage.
199
In terms of risk measures, the Baseline Model has a lower Functional Risk Index
(FRI) than the Alternate Model does (i.e., 0.54 vs. 0.75). This measures a functional
quality penalty associated with discontinuous membership in the Alternate Model. Nearly
half again as many functional exceptions are left ignored or addressed incompletely in the
Alternate Model than in the Baseline. Indeed, SimVision® states that a FRI or PRI value
above 0.7 reflects an unacceptable level of risk for any organization. Alternatively, the
Alternate Model has a lower Project Risk Index (PRI) than the Baseline Model does (i.e.,
0.48 vs. 0.59). This measures a project-quality benefit associated with discontinuous
membership in the Alternate Model. Even with discontinuous membership, the Alternate
Matrix organization leaves relatively fewer project-level tasks incomplete at the end of its
workflows than the Baseline does. Finally, both Baseline and Alternate Models have
identical communication risk values of 0.38. Apparently discontinuous membership has
negligible overall effects on communication quality. Nonetheless, this value is higher
than 0.2, which SimVision® suggests is worrisome in terms of missed communications.
From the above COT simulation results, the continuous membership organization
has more advantages than the discontinuous membership organization. It is superior in
terms of offering a lower cost and a faster delivery schedule—two performance measures
that facility developers prioritize. Hence, discontinuous membership would make the
Alternate Model a less desirable organization design for facility developers. As a
contribution to organization design, we compare the COT results with extant theories and
seek insights to develop empirical prescriptions for designing an organization with
discontinuous membership.
200
First, we would like to ascertain if there are any benefits when operating a
discontinuous organization. As noted above, a discontinuous organization is an
organization that has discontinuous membership due to the different skill set
requirements to complete tasks in certain workflows (Ibrahim and Paulson 2005). The
COT results show that when discontinuous membership operates during a complex
process, such as facility development, there are some relative advantages, as well as some
disadvantages. Among its benefits is that it reduces the total work volume required to
perform the tasks in a workflow process. This gives cost savings to the facility developer
(in terms of paying lower compensation because it has lower total work volume) without
incurring additional delivery schedule (in terms of having almost identical CPM duration
values). It is common practice by facility developers to negotiate for better professional
contracts individually while imposing an independent window of time to deliver the task.
This is because the professional service provider is expert enough to perform his work
independent of others. This assumption is supported by a separate study by Ibrahim et al.
(2005a), which states that experts are capable of working independently and are
comfortable operating in a discontinuous membership organization.
Another benefit is that discontinuous membership encourages greater project
integration (e.g., fewer incomplete reworked tasks, fewer tasks that do not meet adequate
quality) when compared to a continuous organization. One reason for this is that facility
developers hire experts to provide their services only as and when they are needed.
Therefore, fewer project-level mistakes occur, lowering the PRI value comparatively. As
mentioned before, expert members have the ability to deliver their tasks independently,
hence providing better PRI in a discontinuous organization. However, even the PRI value
201
of 0.48 is relatively high (e.g., according to SimVision user guidelines), and should be
worrisome to the management of either continuous or discontinuous organizations. This
will require extraordinary coordination to ensure efficient knowledge flows between
different workflows and different team members.
On the other hand, discontinuous membership incurs more disadvantages in terms
of having greater schedule delay due to the additional coordination and wait volumes—
imposed indirectly on the supervisor. The low centralization forces the supervisor to wait
longer for feedback from his functional team leaders. As the supervisor, the project
manager has to coordinate the functional team leaders (i.e., the hired experts) who have
no direct contacts with other functional team leaders except for those being called upon
by the project manager. And as noted above, an organization with discontinuous
membership incurs an economic penalty in terms of higher project cost. Hence the
discontinuous organization does offer some relative advantages over its continuous
counterpart, but it is disadvantaged in terms of cost and schedule—arguably the most
important measures of project performance.
Second, we review our techniques for representing knowledge flows via COT
models. Drawing from emerging knowledge-flow theory, we would like our model to
represent both explicit and tacit knowledge flows. Using ghost connectors, we model
explicit links between one workflow (i.e., KGS) to the next, which enable accumulated
knowledge to flow to the succeeding workflow (and indirectly to its responsible
organization). In addition, we model tacit knowledge flows between different workflows
by adding ghost communication links between identified interdependent tasks.
202
Further, our KGS Flow Model—grounded upon dynamic knowledge-flow
theories—enables us to represent the essential elements of an organization reflecting
discontinuous membership. This provides an advance in terms of computational
modeling. Moreover, through this advance, our computational models of both continuous
and discontinuous membership point to similar problems (i.e., the civil survey task failure
and an overloaded project manager) that were observed in corresponding real
organizations in practice. This provides a convincing element of retrospective empirical
validation (Thomsen et al. 1999) and promotes our claims regarding external validity of
the KGS Flow Model.
Using a computational model provides some noteworthy learning in terms of
organization studies. Having a semi-formal representation of two, alternate organizations
(e.g., continuous and discontinuous) enables us to examine their similarities and
differences closely, to start and stop the action at critical times, and to repeat the
simulation many times under controlled conditions. This leads to insights that would be
very difficult to capture through fieldwork alone.
For instance, a serendipitous insight occurred when we observed critical path
changes across model runs. Once a set of workflows have been established for a project
organization, the critical path is assumed broadly to be quite stable over time. But in our
models, multiple interdependency links cause the critical path to deviate drastically and
to affect multiple workflows. Since we made no changes to the modeled workflow
processes themselves, such critical path changes reveal how the shift of priorities in
handling emerging issues is responsible. In one instance, the failure to obtain funding
from a program caused the critical path to shift from pre-construction to the finance
203
workflow. This insight may give scholars hope that they may, after all, formalize the
early “shaping” phase of the facility development process, and develop the much needed
knowledge management system that caters to such dynamic operating environment.
Finally, the analysis of the COT results shows that our knowledge-based model is
reliable in explaining how discontinuous membership affects organizational performance.
As noted above, our model “predicted” project problems (e.g., functional failure of the
civil survey work due to the overloaded project manager) that occurred in practice. In
seems increasingly clear that organization theory is enriched through our examination of
discontinuous membership. Discontinuous membership affects the structural information
processing in an organization. But current theory lacks adequate prescriptions to handle
discontinuous membership in organizational design. The inconsistencies of the
contingency fit’s prescriptions to the ethnographic study findings, and the support by
evidence from our knowledge-based COT models, suggest a need to revisit contingency
theory in terms of looking at designing an organization based on its knowledge flow
attributes. We are guided by Nissen (in review) who posits that scholars may eventually
design organizations based on their knowledge flow properties. We discuss in detail in
the following section how we can further support this position.
5.7 Discussion
Our knowledge-based COT models provide evidence that discontinuous membership
does affect the current structural information processing of an organization. Hence,
organization theory needs to consider the discontinuity membership attribute during the
design of organization. Our focused literature review shows current theory also lacks
204
relevant prescriptions to handle discontinuous membership in an organization. We refer
to recent studies that link knowledge attributes to organizational performance. Results
from the VDT-KN extension (Ibrahim et al. 2005b) provides cross-validation for an
earlier ethnographic study by Ibrahim and Paulson (2005), which states that discontinuity
in organizations is a factor in the knowledge loss phenomenon that occurs in facility
development. Ibrahim et al. (2005a) also find that knowledge flow behaviors in a
discontinuous organization are qualitatively different from those of a stable organization
depending on the “expertise” and “continuous” nature of the members. Moreover, their
study finds that knowledge flow behaviors depend on the knowledge type—tacit- or
explicit-dominant—of the knowledge areas in a workflow process.
Burton and Obel (2003) developed six contingency factors corresponding to the
situational fit for the design of organizational structure for an information-processing
organization. They are management style, climate, size/ownership, environment,
technology, and strategy. The situation fit requires that the organization’s context or
situation is internally consistent. An instance is that an equivocal environment and a
routine technology do not fit, but it exists for some organizations. Facility development
organizations are among this group. There is no recommended design, in fact, for this
situation.
On the contrary, a discontinuous organization changes throughout the process, as
and when different skill sets are required in order for the team to respond to the changing
requirements of the life cycle process (Ibrahim and Paulson 2005). The organization may
change for a particular phase as it progresses to the next sequential phase. Based on
Ibrahim and Paulson’s (2005) study, the facility development life cycle operates on a
205
combined complex, uncertain, and equivocal environment. On the other hand, the
organization structure for the concurrent phases remains constant through a parallel phase
and multiple, sequential phases. This observation relates very well to Scott’s (2003) open
system perspective. In an open system, the organizational structure stresses the
complexity and variability of the individual parts—individual participants and
subgroups—as well as the looseness of connections among them. Multiple parts are
viewed as being capable of semiautonomous action, which are loosely coupled to other
parts. In the open system, individuals and subgroups form and leave the coalition,
emphasizing the earlier contingency theory of Lawrence and Lorsch (1967 in Scott
(2003)) on the fluid movement between people and process while the process is still on-
going.
Lawrence and Lorsch (1967 in Scott (2003)) were the first to coin the contingency
term when they argue that different environments place differing requirements on
organizations: specifically, environments characterized by uncertainty and rapid rates of
change in market conditions or technologies. These changes present different demands—
both constraints and opportunities—on organizations than do placid and stable
environments. In order for the organizations to cope with these various environments,
they create specialized subunits with differing structural features. Both researchers found
that the more differentiated the organizational structure, the more difficult it will be to
coordinate the activities of the various subunits and the more bases for conflict will exist
among participants. Hence, more resources and effort must be devoted to coordinating
the various activities and to resolving conflicts among members if the organization is to
206
perform effectively. This concept would later influence Galbraith (1974 and 1977 in
Scott (2003)) for his famed information-processing theory of the organization.
Maintaining the production flows and feedback loops of input, throughput, and
output production flows in such an organization is problematic. In fact, Burton and
Obel’s (2003) closest category for such dynamic organization configuration is ad hoc. Ad
hoc configuration has no hierarchy at all (p. 64), but Ibrahim and Paulson’s (2005)
ethnographic study indicates that the project manager is very much in ‘control’ although
she or he may be overwhelmed by an overload of information processing tasks. Their
study observed an organization that is consistently evolving between the major milestone
events they identified.
The dilemma in managing knowledge flows in a discontinuous organization is
that the organization continues to evolve during the sequential process while another is
maintained in another concurrent process. The interdependent, but loose, connection is
the only link for knowledge flows to happen between the different organizations. It is
easy for knowledge loss to occur when the connectivity between the organizations is
loose, and made worse with discontinuous membership as evidenced by this study. By
having discontinuous membership throughout various phases of the process, the
enterprise is at risk of not being efficient and effective as demonstrated by the test cases
in the VDT-KN simulation (Ibrahim et al. 2005b). New team members contribute new
knowledge, and members who leave bring out some knowledge with them. Therefore, we
are proposing the emergence of a new structural configuration of an organization called
discontinuous.
207
In lieu of this unique discontinuous structural configuration, this study finds that
low centralization and formalization requirements also contribute to the knowledge loss
phenomenon in the facility development enterprise. According to Proposition 6.6 (Burton
and Obel 2003, p. 184), when an organization’s environment is high on equivocality,
high on complexity, and high on uncertainty, then its formalization should be low,
organizational complexity should be low, and centralization should be low. An
organization may be able to provide routine technology to coordinate and monitor a
complex environment so long as it integrates all the task interdependencies between
critical tasks in its workflow processes (ibid.). The contingency fit in Burton and Obel’s
contingency theory requires that the organizational design situation is internally
consistent (ibid., p. 17), which, in the facility developer’s case, is not a good fit. In fact,
Burton and Obel’s contingency theory points out that it would be most difficult to have a
routine technology for uncertain and equivocal environments—the dilemma facing the
design and development of a much needed, yet suitable, knowledge management system
for facility development.
The contingency fit is also unable to recommend any means for formal
integration, but instead recommends an appropriate incentive system to coordinate the
various activities. In this regard, Ibrahim and Paulson (2005) find evidence of conflicting
incentives to various organizations within the facility development enterprise. For
instance, the need for some kind of formalization (having standard operating rules) is
required by the facility developer because of the need to integrate the financing
requirements with the project’s design. It promotes good credibility standing for for-
profit facility developers’ investors while ensuring the success of obtaining competitive
208
funds for the non-profit facility developers. Prudent knowledge flow management is good
practice for long-term property management. On the other hand, members of the design
team are working towards financial rewards as agreed in their professional contracts. The
mechanical engineer in the oak grove preservation example was working towards
completing his task for which he relied on the architect’s early documents while the
facility developer was still negotiating with the local authority. The incentive systems for
such an enterprise may as well be based on the level of its reach (i.e., individual, group,
organization, and inter-organization) as an additional design parameter property of the
organization. We foresee that in a tacit-dominant knowledge area, the higher the
continuous attribute and the higher presence within one life-cycle phase, the organization
could be more effective and efficient (Ibrahim et al. 2005a).
Given the present status quo, we claim that the ethnographic study findings and
the contingency misfits suggest that knowledge loss will be a never-ending problem to
facility developers unless researchers and practitioners acknowledge the discontinuity
attribute in the facility development’s operating environment. This claim is substantiated
by Ibrahim et al.’s (2005b) study, which affirms that a new member in a discontinuous
organization who uses his inaccurate cognition of his other team members to complement
his incomplete cognitive skill, can expose his task to a higher risk of failure that, in turn,
is detrimental to his overall team’s organizational performance in the long run.
Knowledge flows in different phases—sequential or concurrent—depend on the
dominant knowledge type most likely to transpire within the phases. Ibrahim et al.
(2005a) find that experts who are continuous members of a discontinuous organization
facilitate the flow of knowledge in that organization. They also find that the functional
209
experts are self-sufficient in explicit-dominant knowledge areas, but, in tacit-dominant
knowledge areas, the continuous presence of the experts is critical for knowledge flow.
With support from Ibrahim et al. (2005b) that discontinuous membership could be
detrimental to an organization, we pose that the explicitness level of knowledge is key to
determine how effective and efficient an organization would be in various properties and
structural configuration fit.
Furthermore, the knowledge network analysis by Ibrahim et al. (2005a) illustrates
that within a single life-cycle phase, there could be a combination of different knowledge
types—i.e., tacit and explicit—depending on which knowledge area is required to
perform the task. Their study illustrates that different dominant knowledge types have
different knowledge flow characteristics. Being continuous encourages expert members
to contribute generously to other members of the organization during the highest risk
period of the facility development life cycle, i.e., the feasibility and entitlements phases.
Although Galbraith (1974 and 1977) views the organization as an information-processing
entity, and that exception-handling is the responsibility of the supervisor in the
organization structure, the study found a non-hierarchical communication patterns in all
the knowledge areas it identified for the Knowledge Areas Mapping Exercise (KAME).
Monge and Contractor (2003) called this informal network an “emergent” network,
which represents other than the vertical hierarchical characteristics of information-
processing. It includes the development of hierarchy through socialization outside the
formal structure. The VDT-KN extension study (Ibrahim et al. 2005b) also demonstrates
that non-hierarchical information-processing can affect organizational performance. The
regressing media richness that occurs as a facility development process progresses
210
provides additional support for this claim. We claim that tacit knowledge within the
organization flows principally through socialization and internalization (Nonaka 1994)
especially during the high media rich period—i.e., during the feasibility and entitlements
phases. However, instead of moving up and down the organizational hierarchy,
knowledge flows through the social hierarchy where perceived expertise by individual
members is the determinant (Ibrahim et al. 2005a).
Based on these earlier studies conveying the impacts of knowledge flows on
organizational performance (Ibrahim and Paulson 2005; Ibrahim et al. 2005a; Ibrahim et
al. 2005b), we posit that knowledge flow behaviors affect organizational performance in
a substantial way. Moreover, this study illustrates that discontinuous membership affects
organizational performance substantially as well. In summarizing our discussion, our
arguments would support a proposal to add knowledge as a seventh contingency factor
(i.e., as articulated by Burton and Obel 2003) for the design of organizational structure,
where tacit and explicit are its measures. We also propose discontinuity as another
structural configuration measure, and reach as another design parameter property
measure. Our proposal supports Nissen’s (2005b) claim that future organization design
can be based on knowledge flows.
5.8 CONCLUSION
There are several noteworthy contributions and implications from this research. A major
contribution of this work is that the knowledge-based organizational performance model
can justify and support the addition of knowledge as a seventh contingency factor in
Burton and Obel’s (2003) contingency theory for organization design. Our claim is
211
supported by a separate study by Ibrahim et al. (2005b) illustrating via a proof-of-concept
computational model that inaccurate expertise cognition by a new member in a
discontinuous organization can negatively affect the overall organizational performance
of an enterprise. Future research is in line to develop knowledge contingency fit
propositions for inclusion in the well-established diagnosis and design of organizations.
The utilization of knowledge flows in organization design provides a second
contribution. Through this research, we have grounded Nissen’s (2004) knowledge-flow
trajectory model and Nonaka’s (1994) dynamics of knowledge creation and flow theories
as important factors for consideration in contingency theory. More research in
developing descriptive and measurable knowledge-flow constructs is also in line for the
future.
Another contribution and implication in a product development process stems
form our new understanding of how discontinuous membership affects knowledge flows
in an enterprise. The knowledge-based organizational performance model takes into
consideration the practical aspects of knowledge transfer among temporal members. It
has implications on future methods of transferring knowledge in temporal organizations.
The extension of transactive memory theory to include knowledge access to another
member who is not present in the current team can help improve how organizations
manage and train their knowledge resources. Future research is in line to review how
organizations create, store, retrieve, and transfer knowledge, especially tacit knowledge
that mainly belongs to individuals of the enterprise.
In conclusion, the knowledge-based organizational performance model extends
organization theory to address the dynamics of knowledge flows. It is grounded in
212
established organization theories—by virtue of a multi-disciplinary case study research
methodology—such as Galbraith’s (1974 and 1977) information-processing theory,
Grant’s (1996) knowledge of the firm theory, Cohen and Levinthal’s (1990) absorptive
capacity theory, Burton and Obel’s (2003) contingency theory, and Wegner’s transactive
memory theory (1987; in Hollingshead 1998a and 1998b). The model is further grounded
in Nonaka’s (1994) dynamics of knowledge creation and flows theory, and Nissen’s
(2002) knowledge flow trajectory of knowledge flow dynamics theories. The knowledge-
based organizational performance model will inform practice through new theory on
designing organizations with discontinuous membership.
5.9 ACKNOWLEDGEMENTS
This paper is part of the first author’s doctoral research at Stanford University,
which is sponsored by the Ministry of Science, Technology, and Innovation of Malaysia
in affiliation with Universiti Putra Malaysia. The UPS Foundation Endowment at
Stanford University provided additional support. We acknowledge the contributions of
Professors Boyd C. Paulson and Raymond E. Levitt of Stanford University, Palo Alto,
California. This is an extended version of the authors’ paper presented at the 38th
Hawaiian International Conference on System Sciences, Hawaii on January 3-6, 2005.
5.10 REFERENCES
Alavi M., and D. E. Leidner. 2001. Review: Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly 25 (1): 107-136.
Burton, Richard M., and B. Obel. 2003. Strategic organizational diagnosis and design:
Developing theory for application. Boston: Kluwer Academic Publishers.
213
Carillo, P., H. Robinson, A. Al-Ghassani, and A. Chimay. 2004. Knowledge management
in UK: Construction, strategies, resources and barriers. Project Management Journal 35 (1): 46-56.
Carley, K., and M. Prietula. 1994. Computational organization theory. Hillsdale, NJ:
Erlbaum. Carlile, P. R., and E. S. Rebentisch, E. S. 2003. Into the black box: The knowledge
transformation cycle. Management Science 49 (9): 1180-1195. Christiansen, T. 1993. Modeling efficiency and effectiveness of coordination in
engineering design teams: VDT-the Virtual Design Team. Ph.D. diss., Department of Civil and Environmental Engineering, Stanford University.
Cohen, G.P. 1992. The Virtual Design Team: An object-oriented model of information
sharing in project team. Ph.D. Diss., Department of Civil and Environmental Engineering, Stanford University.
Cohen, W. M., and D. A. Levinthal. 1990. Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly 35 (1, Special Issue: Technology, Organizations, and Innovation): 128-152.
Cyert, R. M., and J. G. March. 1963. A behavioral theory of the firm. Englewood Cliffs:
Prentice-Hall. Daft, R. 1992 in Burton and Obel 2003. Organization theory and design (4th Ed.) St.
Paul, MN: West. (2nd ed., 1986; 3rd ed., 1989): p. 286. Davenport, T.H., and L. Prusak. 1998. Successful knowledge management projects.
Sloan Management Review 37 (4): 53-65. Depress, C., and D. Chauvel. 1999. Mastering information management: Part six-
knowledge management. Financial Times (8 March): 4-6. Edwards, J., H. Handzic, S. Carlsson, and M. Nissen. 2003. Knowledge management
research and practice: Visions and directions. Knowledge Management Research & Practice 1 (1): 49-60.
Galbraith, J. R. 1974. Organization design: An information processing view. Interfaces 4
(3): 28-36. Galbraith, J. R. 1977. Organization design. Reading, Massachusetts: Addison-Wesley. Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management
Journal 17 (Special Issue: Knowledge and the Firm): 109-122.
214
Hollingshead, A. B. 1998a. Retrieval processes in transactive memory systems. Journal
of Personality and Social Psychology 74 (3): 659-671. Hollingshead, A. B. 1998b. Communication, learning, and retrieval in transactive
memory systems. Journal of Experimental Social Psychology 34: 423-442. Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for
affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Ibrahim, R., and B. C. Paulson, Jr. 2005. Discontinuity in organizations: How
environmental characteristics contribute to the project's knowledge loss phenomenon. CRGP Working Paper No. 12, Stanford University.
Ibrahim, R., and M. E. Nissen. 2003. “Emerging technology to model dynamic
knowledge creation and flow among construction industry stakeholders during the critical feasibility-entitlements phase,” Proceedings of The ASCE 4th Joint Symposium on IT in Civil Engineering in Nashville, Tennessee, November 15-16 by the American Society of Civil Engineers.
Ibrahim, R., and M. E. Nissen. 2004. “Simulating environmental contingencies using
SimVision®,” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pennsylvania, June 27-29 by the North American Association for Computational Social and Organization Science.
Ibrahim, R., and M. E. Nissen. 2005. “Developing a knowledge-based organizational
performance model for discontinuous participatory enterprises,” Proceedings of the Thirty-Eighth Annual Hawaii International Conference on System Sciences in Hawaii January 3-6 by College of Business Administration, University of Hawaii at Manoa.
Ibrahim, R., M. Shumate, R. Levitt, and N. Contractor. 2005a. Discontinuity in
organizations: Knowledge flow behaviors in sequential workflow processes. CRGP Working Paper No. 17, Stanford University.
Ibrahim, R., R. Levitt, and M. Ramsey. 2005b. Discontinuity in organizations: Impacts of
knowledge flows on organizational performance. CRGP Working Paper No. 18, Stanford University.
Jin, Y., and R. E. Levitt. 1996. The virtual design team: A computational model of
project organizations. Computational and Mathematical Organization Theory 2 (3): 171-196.
215
Kogut, B., and U. Zander. 1992. Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science 3 (3, Focused Issue: Management of Technology): 383-397.
Kunz, J.C., R. E. Levitt, and Y. Jin. 1998. The virtual design team: A computational
simulation model of project organizations. Communications of the Association for Computing Machinery 41 (11): 84-92.
Lambert, M. H., and B. Shaw. 2002. Transactive memory and exception handling in
high-performance project teams. CIFE Technical Report No. 137, Stanford University.
Lawrence, P. R., and J. W. Lorsch. 1967. Organization and environment: Managing
differentiation and integration. Boston: Graduate School of Business Administration, Harvard University.
Levitt, R. E., G. P. Cohen, J. C. Kunz, C. I. Nass, T. R. Christiansen, and Y. Jin. 1994.
The Virtual Design Team: Simulating how organization structure and information processing tools affect team performance. Edited by Carley, K. M. and M. J. Prietula. Computational and Mathematical Organization Theory. Hillsdale, NJ: Lawrence Erlbaum.
Levitt, R. E., J. Thomsen, T. R. Christiansen, et al. 1999. Simulating project work
processes and organizations: Towards a micro-contingency theory of organizational design. Management Science 45: 1479-1495.
March, J. G., and H. A. Simon. 1958. Organizations. New York: John Wiley & Sons,
Inc. Mintzberg, H., and F. Westley. 1992. Cycles of organizational change. Strategic
Management Journal 13 (Special Issue: Fundamental Themes in Strategy Process Research): 39-59.
Monge, P. E., and N. S. Contractor. 2003. Theories of communication networks. Oxford,
UK: Oxford University Press. Moreland, R. 1999. Transactive memory: Learning who knows what in work groups and
organizations. In J. Levine (Ed.) Shared Cognition in Organizations: The Management of Knowledge. Hillsdale, NJ: Lawrence Erlbaum Associates.
Nadler, J., L. Thompson, and L. van Boven. 2003. Learning negotiation skills: Four
models of knowledge creation and transfer. Management Science 49 (4): 529-540. Nelson, R. R., and S. G. Winter. 1982 (in Scott 2003). An evolutionary theory of
economic change. Cambridge, MA: Belknap Press of Harvard University Press.
216
Nissen, M. E. 2002. An extended model of knowledge-flow dynamics. Communications of the Association for Information Systems 8: 251-266.
Nissen, M. E. 2005a. Harnessing knowledge dynamics: Principled organizational
knowing and learning. Hershey, PA: Idea Group Publishing (forthcoming). Nissen, M. E. 2005b. Toward designing organizations around knowledge flows. In: K.
Desouza (Ed.), New frontiers in knowledge management (forthcoming). Nissen, M. E. 2005c. “A computational approach to diagnosing misfits, inducing
requirements, and delineating transformations for Edge Organizations.” Proceedings International Command and Control Research and Technology Symposium, McLean, VA (June).
Nissen, M. E. (in review). Understanding knowledge-flow patterns for design. Nissen, M. E., and R. E. Levitt. 2004. Agent-based modeling of knowledge dynamics.
Knowledge Management Research and Practice 2 (3). Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization
Science 5 (1): 14-37. Nonaka, I., and H. Takeuchi. 1995. The knowledge-creating company- How Japanese
companies create the dynamics of innovation? New York: Oxford University Press.
Paulson, B. C., Jr. 1976. Designing to reduce construction cost. Journal of the
Construction Division 102 (CO4), Proc. Paper 12600, December: 587-592. Scott, W. R. 2003. Organizations: rational, natural, and open systems. New Jersey:
Prentice Hall. Szulanski, G. 2000. The process of knowledge transfer: A diachronic analysis of
stickiness. Organizational Behavior and Human Decision Processes 82 (1): 9-27. Thomas-Hunt, M. C., T. Y. Ogden, and M. A. Neale. 2003. Who's really Sharing? Effects
of social and expert status on knowledge exchange within groups. Management Science 49 (4): 464-477.
Thomsen, J., R. E. Levitt, J. C. Kunz, C. I. Nass, and D. B. Fridsma. 1999. A trajectory
for validating computational emulation models of organizations.” Journal of Computational & Mathematical Organization Theory 5 (4) Dec: 385-401.
Von Hippel, E. 1994. "Sticky information" and the locus of problem solving: Implications
for innovation. Management Science 40 (4): 429-439.
217
Weber, M. 1947. The theory of social and economic organization (A. H. Henderson and T. Parsons, Eds. and trans.). Glencoe, IL: Free Press.
Wegner, D. M. 1987. Transactive memory: A contemporary analysis of the group mind.
Edited by B. Mullen, and G. R. Goethals. Theories of Group Behavior. New York: Springer-Verlag, 185-208.
218
APPENDIX 1
RIVERWOOD APARTMENTS
1-1 Riverwood Apartments Fact Sheet 1-2 Riverwood Workflows and Organizations for Test Cases
1-3 Project Summary Tasks List
1-4 Property Development Documents and Schedule Tracer Form
1-5 Operation and Warranty Manual Content List
219
APPENDIX 1-1: RIVERWOOD APARTMENTS FACT SHEET An affordable Housing Development by Mid-Peninsula Housing Coalition
General
• Riverwood, taking its name from the nearby Guadalupe River, defines the corner of Tasman Drive and Lick Mill Boulevard and completes the missing piece of the residential neighborhood to the south. Riverwood seeks to provide critically needed affordable housing in two separate housing projects on one 4.3 acre site. Residents of Riverwood will be within walking distance of transit, restaurants, a school, and parks.
• The Riverwood Apartment development is comprised of two separate projects: -Riverwood Grove – 71 units of large family housing -Riverwood Place – 148 units of efficiency studio (SRO) housing
• Creating desirable affordable housing for two distinct yet similar populations: single adults and families is not a simple problem. Each group has its own service requirements and spatial needs, yet each group needs safe, decent affordable housing in Santa Clara. In response, Mid-Peninsula Housing Coalition (MPHC) has approached Riverwood as two separate projects. The two uses will be adjacent on the parking, common space, and management.
History of Development
• City of Santa Clara identified a strong need for affordable family and individual housing development. The City’s 1990 – 2005 general plans projected nearly 1,000 units of housing affordable to house holds earning less than 50% of the area median income are needed. The City has a legal obligation to meet housing needs established in the General Plan.
• Recent economic expansion in Santa Clara County has pushed rents skyward, and economic projection predicts steady graph and steady pressure on housing in the foreseeable future. This pressure will only magnify the difference between market and affordable rent. According to projections ’98 by the Silicon Valley Manufacturing Group, rents increased 30% from 1996 – 1998 in Santa Clara County. With market rate rents for apartments reaching well over $2,000 per month, many working families simply cannot make ends meet.
• Due to lack of affordable housing in Santa Clara, hundreds of working families are experiencing severe housing cost burdens or are on waiting list for affordable housing in Santa Clara.
• Getting in an affordable housing development is difficult as well. The waiting list for the Santa Clara Country Housing Authorities is more than 3,000 names long. Waiting lists for MPHC affordable housing projects in northern Santa Clara County is so long that they remain closed to new applicants during most of the years.
• In April 1998 the City of Santa Clara, the Redevelopment Agency of the City of Santa Clara, and the Sports and Open Space Authority of the City of Santa Clara
220
designated this site for developments of affordable housing and authorized the City manager and Redevelopment Agency Executive Director to take the necessary steps to develop this site.
• On November 20, 1998 a Request for Proposals (RFP) was sent to area developers. The RFP stated that the City Council had designated this site as appropriate for affordable housing.
• MPHC responded to the RFP with a proposal and as a result of that competitive process was selected as the developer of developer of the site on March 23, 1999 by the city Council. Since this date MPHC has been working with City staff towards the completion of the development.
• The Redevelopment Agency has included this development in the Agency’s Five-year Implementation Plan adopted on July 31, 1999.
City Approvals
• This development is required to go through the same entitlement process as any other development.
• This development must comply with all zoning, building code, and accessibility requirements.
• This development must pay all planning, building, and impact fees as any other project would be required to.
• MPHC must pay school impact fees as required by law. Design and Amenities
• Development is planned and designed by the award-winning firms of Backen Arrigoni &Ross Architects and Berger-Detmer Architects.
• The development is master planned as two separate and independent components, the family housing development at 2150 Tasman Drive and the efficiency studio housing development at 5090 Lick Mill Boulevard. Each development will have its own amenities and management.
• The Pedestrian Pathway from Calle de Escuela to Tasman Drive will be maintained.
• Building massing was thoroughly studied to minimize impacts on surrounding residential neighbors. To that end, the corners of each building are stepped down from three to two stories and the entire southeastern side of the building four is held to two stories as well.
• 176 out of 258 parking spaces are in garages. 2/3 of the cars of this development will be out of sight. Ample space will be provided for visitor parking.
• Laundry facilities will be provided on site for each development. • Each development will have its own courtyards and open space. The family
development will be provided with a tot lot and play areas. • The family development will have an after school educational enrichment
program, summer youth enrichment program, and computer educational program for school aged youth.
• Light-rail and bus routes are available adjacent to the site.
221
Approximate Apartment and Community Area Sizes and Rents Component Approx. Size (sf) 40% Rent 50% 60% Family Development 1-bedroom 648-700 $652 $815 $978 2-bedroom 1,042 $783 $978 $1,174 3-bedroom 1,088-1,206 $905 $1,131 $1,357 4-bedroom 1,312 $1,009 $1,261 $1,513 Community Area 2430 N/A
Approx. Size (sf) 40% Rent 50% 60%
SRO Standard unit 324 to 401 $532 $761 $913 Loft Unit 400 to 450 $532 $761 $913 Manager’s Unit 831 N/A Common Areas 23,000 N/A Income Levels HH Size 35% AMI 40% 50% 60% 100% 1 $21,315 $24,360 $30,450 $36,540 $60,900 2 $24,360 $27,840 $34,800 $41,760 $69,600 3 $27,405 $31,320 $39,150 $46,980 $78,300 4 $30,450 $34,800 $43,500 $52,200 $87,000 5 $32,900 $37,600 $47,000 $56,400 $94,000 Property Management
• These developments will professionally managed by Mid-Peninsula Housing Management Cooperation (MPHMC), a nonprofit affiliate of MPHC. MPHC has twenty years of experiences and a track record of excellent management of affordable housing developments
• MPHMC manages nearly 4,200 apartments in six Bay Area counties. All developments are available for your inspection.
• MPHMC must comply with all state and federal fair housing law and regulations. • Two on-site management employees will be supported by additional staff
including a full time maintenance crew. • Landscaping maintenance will be contracted out to a professional service. • Operating budgets are being designed to allow ample funds for the necessary and
appropriate management of this development. • A standard “no-pet” policy will apply to all residents of the development. • MPHC properties average about 1.3 persons per bedroom. • MPHMC has policy of no extended guests in apartments.
Resident Election
• Efficiency studios will be occupied by one individual only. • Rental preference will be given to households who live or work locally. • MPHMC will market these developments to local businesses, local employers,
teachers, and local public employees.
222
• Potential residents must have sufficient income, good credit, and good rental references.
• MPHMC conducts personal visit to potential tenants’ existing homes to determine if they can care for an apartment. It’s in MPHMC’s best interest to allow only responsible tenants to reside in our developments. We do not allow residents to destroy property.
• It is MPHMC’s policy to assign households to appropriately size apartments. MPHMC is bound by regulatory agreements and cannot allow overcrowding
Parking
• MPHMC employs a managed parking plan to assign parking spaces and enforce parking regulations.
• In the proposed Family Housing Development 1.85 parking spaces per dwelling unit are being provided on site. In the comparable MPHC properties an average of 1.5 parking spaces per dwelling unit provide adequate parking to meet residents needs. Below is a breakdown of how parking will be provided for the family housing development. More than ½ of the parking spaces are in garages hidden from view.
Family Housing Development Parking Type Approximate # of Spaces Surface 54 Tuck Under Garages 38 Below Grade Garage 40 Total 132
• in the proposed Efficiency Studio Development .85 parking spaces per dwelling unit are being provided onsite. In comparable MPHC properties an average of less than .5 parking spaces per dwelling unit provide adequate parking to meet resident needs. Below is a breakdown of how parking will be provided for the efficiency studio development. More than ¾ of the parking is provided in a garage hidden from view.
Efficiency Studio Development ParkingType Approximate # of Spaces Surface 28 Below Grade Garage 96 Total 126 Development Schedule Riverwood Place Riverwood Grove Appoint Architect October, 1999 October, 1999 Start Site Plan Review February, 2000 February, 2000 Receipt of Development Approval October, 2000 October, 2000 Start Construction October, 2001 October, 2001 Complete Construction March, 2003 January, 2003
223
APPENDIX 1-2: RIVERWOOD WORKFLOWS AND ORGANIZATIONS FOR TEST CASES
224
225
Figu
re A
-1. R
iver
woo
d A
partm
ents
aff
orda
ble
hous
ing
deve
lopm
ent p
rogr
am in
Sim
Vis
ion®
Figu
re A
-2. T
est c
ases
for s
eque
ntia
l kno
wle
dge
flow
s.
226
227
Figu
re A
-3. W
orkf
low
pro
cess
and
org
aniz
atio
n fo
r fea
sibi
lity-
entit
lem
ents
pha
se o
f Riv
erw
ood
Apa
rtmen
ts.
228
Figu
re A
-4. W
orkf
low
pro
cess
and
org
aniz
atio
n fo
r bui
ldin
g pe
rmit
phas
e of
Riv
erw
ood
Apa
rtmen
ts.
229
Figu
re A
-5. W
orkf
low
pro
cess
and
org
aniz
atio
n fo
r de
velo
pmen
t pro
ject
fina
ncin
g ph
ase
of R
iver
woo
d A
partm
ents
.
230
NO
TE: A
ssig
ned
acto
r’s s
kill
in b
old.
Tabl
e A
-1. M
ultip
le S
kill
Sets
of t
he R
iver
woo
d En
terp
rise
Tabl
e A
-1. M
ultip
le S
kill
Sets
of t
he R
iver
woo
d A
partm
ents
Ent
erpr
ise
Table A-2: Riverwood Actors’ Fulltime Equivalent (FTE) Distribution
STAFFING PERSONELS
Feas
ibility
-Enti
tleme
nts
Build
ing P
ermi
t Pha
se
FAM
Cons
tructi
on
FAM
Finan
ce
FAM
Asse
t Man
agem
ent
FAM
Prop
erty
Mana
geme
nt
SRO
Cons
tructi
on
SRO
Finan
ce
SRO
Asse
t Man
agem
ent
SRO
Prop
erty
Mana
geme
nt
OWNER Executive Director 0.40 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 Project Manager 0.50 0.20 0.20 0.25 0.15 0.20 0.20 0.25 0.15 0.20 Services Director 0.10 0.10 0.10 0.10 Accounting Department 0.50 0.50 Chief Operating Officer 0.30 0.30 Public Relations Executive 1.00 1.00 Regional Manager 0.30 0.30 Compliance Specialist 1.00 1.00 Property Manager 0.30 0.30 Site Manager 1.00 1.00 CONSULTANTS/BUILDER Title Company 1.00 Environmental Engineer 1.00 Surveyor 1.00 1.00 1.00 Architect 1.00 4.00 0.50 0.50 Civil Engineer 0.50 1.00 0.10 0.10 Landscape Architect 0.50 1.00 0.10 0.10 Geotech Engineer 1.00 Financial Consultant 1.00 1.00 1.00 1.00 1.00 General Contractor 0.10 1.00 2.00 2.00 Value Engineer 1.00 1.00 Wood Structural Engineer 0.25 0.10 0.10 Concrete Structural Engineer 0.25 0.10 0.10 MEP Engineer 0.50 0.10 0.10 3rd Party Inspector 0.10 0.10 Geotech Inspector 0.10 0.10 Legal Advisor 0.50 0.15 0.50 0.15 Auditor 1.00 1.00
NOTE: 1FTE = 8 hour per day
231
Table A-3: Riverwood Owner Personels’ Roles and Application Experiences
OWNER STAFFING PERSONELS
Feas
ibility
-Enti
tleme
nts
Build
ing P
ermi
t Pha
se
FAM
Cons
tructi
on
FAM
Finan
ce
FAM
Asse
t Man
agem
ent
FAM
Prop
erty
Mana
geme
nt
SRO
Cons
tructi
on
SRO
Finan
ce
SRO
Asse
t Man
agem
ent
SRO
Prop
erty
Mana
geme
nt
Executive Director PM (H)
PM (M)
PM (M)
PM (H)
PM (H)
PM (H)
PM (M)
PM (H)
PM (H)
PM (H)
Project Manager PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
PM (M)
Services Director SL (M)
SL (M)
SL (M)
SL (M)
Accounting Department SL (M)
SL (M)
Chief Operating Officer SL (M)
SL (M)
Public Relations Executive ST (M)
ST (M)
Regional Manager SL (M)
SL (M)
Compliance Specialist SL (M)
SL (M)
Property Manager ST (M)
ST (M)
Site Manager ST (M)
ST (M)
NOTE: 1FTE = 8 hour per day PM = Project Manager; SL = Subteam Leader; ST = Subteam Member; (H) = High; (M) = Medium; (L) = Low
232
Table A-4: Riverwood Consultants’ and Builder’s Roles and Application Experiences
CONSULTANTS/BUILDER STAFFING PERSONELS Fe
asibi
lity-E
ntitle
ments
Build
ing P
ermi
t Pha
se
FAM
Cons
tructi
on
FAM
Finan
ce
FAM
Asse
t Man
agem
ent
FAM
Prop
erty
Mana
geme
nt
SRO
Cons
tructi
on
SRO
Finan
ce
SRO
Asse
t Man
agem
ent
SRO
Prop
erty
Mana
geme
nt
Title Company ST (M)
Environmental Engineer ST (M)
Surveyor ST (M)
ST (M)
ST (M)
Architect SL (H)
SL (H)
SL (H)
SL (H)
Civil Engineer ST (M)
ST (M)
ST (M)
ST (M)
Landscape Architect ST (M)
ST (M)
ST (M)
ST (M)
Geotech Engineer ST (M)
Financial Consultant ST (H)
ST (H)
ST (H)
ST (H)
ST (H)
General Contractor SL (H)
SL (H)
SL (H)
SL (H)
Value Engineer ST (H)
ST (H)
Wood Structural Engineer ST (M)
ST (M)
ST (M)
Concrete Structural Engineer ST (M)
ST (M)
ST (M)
MEP Engineer ST (M)
ST (M)
ST (M)
3rd Party Inspector ST (H)
ST (H)
Geotech Inspector ST (M)
ST (M)
Legal Advisor ST (H)
ST (H)
ST (H)
ST (H)
Auditor ST (M)
ST (M)
NOTE: 1FTE = 8 hour per day PM = Project Manager; SL = Subteam Leader; ST = Subteam Member; (H) = High; (M) = Medium; (L) = Low
233
APPENDIX 1-3: PROJECT SUMMARY TASKS LIST
1. GENERAL 1.1 Property Information 1.2 Property Profile
1.2.1 Residential units profile according to initial rent up. 1.2.2 Building profile 1.2.3 Construction profile 1.2.4 Development schedule
1.3 Project Highlights (not limited to the following issues) 1.3.1 Acquisition issues 1.3.2 Ownership issues 1.3.3 Governmental review issues 1.3.4 Design issues 1.3.5 Construction issues 1.3.6 Regulatory issues 1.3.7 Potential legal implications 1.3.8 Other issues invisible on hard documents
2. PROJECT DESCRIPTION 2.1 General 2.2 Build-up Area Summary 2.3 Parking 2.4 Development Cost Summary 2.5 Environmental Reports and Documents
2.5.1 Environmental reports and documents schedule 2.5.2 Description of required environmental mitigation
2.6 Government Approvals Schedule 2.7 Design-Construction Schedule 2.8 Residential Units Schedule 2.9 Direct Contracts 2.10 Building Documents Log 2.11 Amenities Provision
2.11.1 Fire protection provision 2.11.2 Appliances provision 2.11.3 Utilities metering 2.11.4 Other amenities
2.12 Selected Drawings (Minimum site plan, building plans, and all different unit plans)
3. PROJECT FINANCING 3.1 Funding Sources 3.1.1 Permanent sources summary
3.1.2 Permanent lenders 3.1.3 Grants 3.1.4 Guarantees/Letter of Credit
3.2 Finance documents log 3.3 Finance schedule
234
4. PROPERTY & ASSET MANAGEMENT 4.1 General 4.2 Compliance Requirements 4.2.1 Income restrictions
4.2.2 Operating constraints 4.2.3 Initial rent roll
4.3 Reporting Requirements 4.4 Fees and Distribution
4.4.1 Management fees 4.4.2 Order of priority for spending operating income
4.5 Reserves 4.5.1 Replacement reserves 4.5.2 Operating reserves
4.6 Capital Management Plan 4.6.1 Latest capital improvement plan 4.6.2 Preventive maintenance plan 4.6.3 Long-term capital improvement plan 4.6.4 Capital improvement procedure
4.7 Property Management Documents Log 4.8 Property Management Schedule 5. SERVICES 5.1 TCAC Regulatory commitments 5.2 AHP Regulatory commitments 5.3 Services Documents Log 5.4 Services Schedule 6. OWNERSHIP 6.1 General 6.2 LP Buyout 6.3 Required approvals and notifications 6.4 Taxes 6.5 Insurance 6.6 Guarantee
6.6.1 Development Deficit Guaranty 6.6.2 Operating Deficit Guaranty
6.7 Ownership Documents Log 6.8 Ownership Schedule 7. APPENDIXES (Minimum, but not limited to) Appendix 1: Contacts Schedule Appendix 2: TCAC Proforma (Application, then replaced by Final Account) Appendix 3: Site and Location Plan Appendix 4: Preventive Maintenance Annual Schedule
235
APPENDIX 1-4: PROPERTY DEVELOPMENT DOCUMENTS AND SCHEDULE TRACER FORM
1. PROJECT NAME: ______________________________________________________ 2a. ENVIRONMENTAL REPORTS AND DOCUMENTS
DESCRIPTION PREPARED BY REFERENCE/ PROJ. NO.
DATE OF DOCUMENT
OFF-SITE?
Survey Plan Y
Soil/ Geotechnical Report
Y
Site Topo Survey Y ALTA Site Survey N Phase 1 Report N Phase II Report N Draft EIR Report N Asbestos Analysis N* Archeological Y Biotic Review Y Traffic Report Y Acoustic Report Y Plumbing Report Y Electrical Report Y Termite Report Y Flood Mitigation Report
Y
NOTE: * Depends on whether new or rehab type of projects. 2b. GOVERNMENTAL APPROVALS SCHEDULE
DESCRIPTION PREPARED BY
REFERENCE/ PROJ. NO.
DATE OF DOCUMENT
OFF-SITE?
Negative Declaration Y Coastal Comm. Approval Y Article 34 State Const. Y Environmental Review (CEQA)
Y
Conditional Use Permit N Variance Approval N Planning Approval N Development Approval N Grading Permit Y Building Permit Y Certificate of Final Occupancy
N
236
2b. DESIGN AND CONSTRUCTION SCHEDULE
DESCRIPTION DATE Appointment of Architect Site Control Environmental Review Submission
Planning Submission Building Permit Submission Bidding Building Contract Award Site Hand-over/ Notice to Proceed
Notice of Construction Completion
Certificate of Final Occupancy
Placed-in service
Move-in date 100% Leased-up
2c. BUILDING DOCUMENTS LOG
DESCRIPTION REFERENCE LOCATION DATE As-built drawings* (include RFIs and COs)
Original- Central Copy- Property
Operation and Warranty Binder
Original- Central Copy - Property
NOTE: Drawings include one 11”x17” set for Central
237
3a. FINANCING DOCUMENTS LOG
PROG./ SOURCE
DESCRIPTION OF REFERENCED DOCUMENTS
REFERENCE DATE OF DOC.
DATE RECORDED
AHP Deed of Trust, Assgn. of Rents & Security Agreement
Promissory Note Affordable Housing Direct Subsidy Agreement
HCD/ HOME
Deed of Trust, Assgn. of Rents, Security Agrmt & Fixture Filing
Promissory Note HCD Regulatory Agrmt HCD Development Agrmt HCD Standard Agrmt RDA Promissory Note Deed of Trust Financing Agrmt Agrmt re Covenants, Conditions,
and Restrictions Assignment & Assumption Agrmt
HUD Promissory Note Deed of Trust with Owner Financing Agrmt Agrmt re Covenants, Conditions,
and Restrictions
CHFA Promissory Note CHFA Regulatory Agrmt
Deed of Trust
TCAC Regulatory Agrmt LP Partnership Agrmt Partnership Agrmt Amendment*
238
3b. FINANCE SCHEDULE
DESCRIPTION DATE Letter of intent to purchase Site acquisition closing Construction loan application Construction loan commitment Construction loan closing RDA application RDA closing RDA maturity date AHP application AHP closing AHP maturity date HCD/HOME application HCD/HOME closing HCD/HOME maturity date CHFA application CHFA closing CHFA maturity date HUD application HUD closing HUD maturity date TCAC application TCAC approval Partnership Closing Partnership Funding End of Partnership date
239
4a. PROPERTY MANAGEMENT DOCUMENTS LOG
DESCRIPTION REFERENCE DATE Management Budget & Plan Management Agreement Affirmative Marketing Plan Approved First Year Budget Applicants Database Property Inventory List Preventive Maintenance Schedule Initial Rent Roll Long-term Capital Improvement Plan
4b. PROPERTY MANAGEMENT SCHEDULE
DESCRIPTION DATE Property Management Plan Proposal Management Budget & Plan Approval Management Agreement Closing Lease-up Process Start Date Affirmative Marketing Plan Operating Budget Approval Advertising Start Date Advertising End Date Application Distribution Start Date Application Distribution Finish Date Application Deadline Lottery Draw Date Complete Certification Process Issuance of Congratulation Letter Start Date Complete Unit Inventory Checklist Complete Property Inventory Checklist Complete Preventive Maintenance Schedule Complete Long-term Capital Improvement Plan Tenants Move-in 100% Leased-up Effective Date of Property Management Services 90-day Notice of Termination by Mgmt Agent 30-day Notice of Termination by Owner Notice for Management Agent Renewal End of Property Management
240
5a. SERVICES DOCUMENTS LOG
DESCRIPTION REFERENCE DATE Latest Demographic Report Prelim Services Requirements Report Services Agreement Third Party Services Agreement MOU MOU
5b. SERVICES SCHEDULE
DESCRIPTION DATE Preliminary Services Requirements Report Services Consultancy Agreement Services Agreement Closing Facility Handing Over Effective Services Start Date Termination Notice Date Services Agreement End Date
241
6a. OWNERSHIP DOCUMENTS LOG
DESCRIPTION REFERENCE DATE Initial GP Incorporation Development Approval Partnership Agreement Grant Deed Tax ID Number of LP Amendment to Partnership Agreement Certificate of 501(c)3 Status Land Leases Final Title Report Final ALTA Survey Welfare Exemption Insurance Guarantees Subsidy contracts Form 8609 Appraisal for Option Notice
6b. OWNERSHIP SCHEDULE
DESCRIPTION DATE Letter of Intent to Purchase Property Initial GP Incorporation Site Acquisition Development Approval TCAC Approval Partnership Closing Amendment of LP Certificate of Occupancy Major Partnership Funding Form 8609 Notice of Option to Purchase End of Option to Purchase Appraisal for Notice Option to Purchase Close Option to Purchase (90 days after notice of option to purchase)
End of Partnership Agreement
242
APPENDIX 1-5: OPERATION AND WARRANTY MANUAL CONTENT LIST
SUBCONTRACTOR INFORMATION
1. SUBCONTRACTOR’S OFFICE/ EMERGENCY PHONE LIST WARRANTY
1. SUBCONTRACTORS WARRANTY SUMMARY 2. PRODUCTS WARRANTY SUMMARY 3. WARRANTY PROCEDURE 4. GENERAL CONTRACTOR WARRANTY 5. SUBCONTRACTORS WARRANTY
PROJECT APPROVAL/ DOCUMENTS
1. SUBMITTAL LOG 2. BUILDING PERMIT CARDS 3. INSULATION CERTIFICATE 4. COLOR SELECTIONS
PRODUCT INFORMATION & OPERATING/MAINTENANCE MANUALS
1. 01000- GENERAL Not applicable 2. 02000- SITE CONSTRUCTION
02825- Fences, gates, and hardware 02870- Site, street and mall furnishings
3. 03000- CONCRETE 4. 04000- MASONRY 5. 05000- METALS 6. 06000- WOOD & PLASTICS 7. 07000- THERMAL & MOISTURE PROTECTION 8. 08000- DOORS & WINDOWS 9. 09000- FINISHES 10. 10000- SPECIALTIES 11. 11000- EQUIPMENT 12. 12000- FURNISHINGS 13. 13000- SPECIAL CONSTRUCTION 14. 14000- CONVEYING SYSTEM 15. 15000- MECHANICAL 16. 16000- ELECTRICAL
243
AS-BUILT DRAWINGS 1. ARCHITECTURAL 2. STRUCTURAL 3. CIVIL 4. LANDSCAPE 5. PLUMBING 6. MECHANICAL 7. ELECTRICAL 8. FIRE SPRINKLER
a) Sample Heading for Subcontractors Warranty Summary Table
Subcontractor Name
Scope of Work Warranty Start Date
Warranty End Date
Notice Period before Action
by Owner Corp Pavers,
Inc. Paving Jan 18, 2001 Jan 17, 2002 7 days
b) Sample Heading for Products Warranty Summary Table
CSI Number
Manufacturer Installed By Warranty Start Date
Warranty End Date
02825 Metal Gates, Inc
ABC Subcontractor
Jan 18, 2001 Jan 17, 2005
244
APPENDIX 2
VIRTUAL DESIGN TEAM – KNOWLEDGE NETWORK
245
246
Figu
re A
-6. W
orkf
low
pro
cess
and
org
aniz
atio
n w
ithou
t kno
wle
dge
netw
ork
247
Figu
re A
-7. W
orkf
low
pro
cess
and
org
aniz
atio
n w
ith k
now
ledg
e ne
twor
k
APPENDIX 3
KNOWLEDGE ASSET MAPPING EXERCISE (KAME)
3-1 Pre-Knowledge Asset Mapping Interview Protocol 3-2 Riverwood Code Book
248
APPENDIX 3-1: PRE-KNOWLEDGE ASSET MAPPING INTERVIEW PROTOCOL Thank you for agreeing to participate in this research. We are interested in learning more about the group you are associated with. To help us gain a better understanding of your group we have questions in four categories: group membership demographics, group’s core competencies, group’s core tasks, and group’s information infrastructure. Please provide as much information as you can by expanding on the questions as you see fit, and feel free to interrupt me at any point if you need further clarification about the question.
I. GROUP MEMBER DEMOGRAPHICS
How many members are on this project group? The table provided can be used to collect information to the following questions:
What are the names of the departments that members of this project group belong to?
What are the group members’ names? (Please identify first and last names) PLEASE SEE CHART BELOW. Which members are associated with what departments? PLEASE SEE CHART BELOW. For each group member, what title do they hold, what is their position within the group, and what level are they within the organization?
DEPT./PHASE GROUP MEMBER NAME TITLE POSITION AND LEVEL FIRST LAST
249
Is there a directory or organizational chart that identifies the role/position of each group member, and is it possible to get a copy for future reference?
What are the names of the projects this group is currently working on? Could you provide a brief description of each project?
What is the typical time frame for each project (e.g., a day, week, month, year)? What other departments (project phases) does this project group collaborate with to complete their work? What are the names of people outside this project group that these members frequently collaborate with to complete their work? What other projects in the company are interdependent with this group’s work? What are some of the jargon words of this group, and what do they mean or refer to? What acronyms might we need to know when working with this group, and what do they mean?
II. CORE COMPENTENCIES
Can you identify approximately five macro-level knowledge areas that this group deals with on a daily/weekly basis that are used to support project development? In other words, what knowledge areas must project group members possess to get their work done? On average, how frequently do group members receive new information on each of the five macro-level knowledge areas?
1. ONCE EVERY WEEK 2. ONCE EVERY MONTH 3. ONCE EVERY WEEK
250
4. TWICE EVERY WEEK 5. TWICE EVERY MONTH
How frequently is each of the five macro-level knowledge areas used?
1. ONCE EVERY WEEK 2. TWICE EVERY MONTH 3. TWICE EVERY WEEK 4. EVERY SECOND DAY 5. TWICE EVERY MONTH
Is there a handbook or organizational chart identifying who is responsible for what knowledge areas and is it possible to get a copy for future reference?
III. CORE TASKS
Can you identify approximately five macro-level tasks that the group deals with on a daily/weekly basis that are completed to support project development?
1. OBTAIN PROGRAM (FINANCIAL/SERVICE) FEEDBACKS 2. REVIEW FINANCIAL FEASIBILITY 3. REVIEW ARCHITECTURAL-ENGINEERING-CONSTRUCTION
DOCUMENTS 4. COORDINATE COMPLIANCE REQUIREMENTS 5. PROCESS PROGRAM APPLICATIONS
On average, how frequently do members receive new tasks in each of the five macro-level task areas that you just mentioned?
1. ONCE EVERY 2 WEEKS 2. ONCE EVERY WEEK 3. TWICE EVERY WEEK 4. ONCE EVERY 2 WEEKS 5. ONCE A MONTH
IV. INFORMATION INFRASTRUCTURE AND DATABASES (electronic or
paper):
What kind of technology infrastructure do you have for the group?
Are there any semi-private group spaces? Is there any limitation based on firewalls to outside sources?
251
What types of tools, if any, does the technology provide (e.g., video, push technologies, directory assistance)?
What databases, either electronic or paper, do group members have that contain work related information?
What are the names of these databases and which of these are digital and which are hard copies? What types of information does each database hold?
Is there a difference in who has access to each of these databases? If so, who has read/write/access permissions to each database?
Do you have searching capabilities for these databases? (i.e., a search engine) Are there other referral databases group members have access to?
How much information does each database posses on the five macro-knowledge areas you identified above? How much information does each database posses that is related to the five macro-tasks you identified above? _____________________________________________________________________ Thank you for your time in answering these questions. Your responses will be very useful in our research and survey development. We will contact you if we have any further questions. Additionally, please feel free to contact us at any point if you have questions about this interview. Contact name: Rahinah Ibrahim Phone number: (650) 799-9321 Email: [email protected]
252
APPENDIX 3-2: RIVERWOOD CODE BOOK Codebook information for Knowledge Areas Required for Each Task for the "Concept Design-Development Approval" Period (tk1) Question 1 (of 36) Here is a matrix containing a list of knowledge areas (the columns) and tasks (the rows) that were identified by people in your team for the "concept design-development approval" period. We would like to know the degree to which each knowledge area is needed to complete each task. Please indicate the degree of required knowledge for each task. Please use the scrollbar at the bottom of the screen to view the last column if necessary. This datatype has 15 columns, labeled tk1_x_y, where x is: 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents 3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications and y is 0 -- Regulatory/Authority Requirements 1 -- Development project finance 2 -- AEC issues The value for each entry is: 0 -- No Response 1 -- I Don't Know 2 -- None 3 -- A Little Amount 4 -- A Moderate Amount 5 -- An Enormous Amount Codebook information for Knowledge Areas Required for Each Task for the "Development Approval-Start Construction" Period (tk2) Question 2 (of 36) Here is a matrix containing a list of knowledge areas (the columns) and tasks (the rows) that were identified by people in your team for the "development approval-start construction" period. We would like to know the degree to which each knowledge area is needed to complete each task. Please indicate the degree
253
of required knowledge for each task. Please use the scrollbar at the bottom of the screen to view the last column if necessary. This datatype has 10 columns, labeled tk2_x_y, where x is: 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents 3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications and y is 0 -- AEC Documentation process 1 -- Bidding process The value for each entry is: 0 -- No Response 1 -- I Don't Know 2 -- None 3 -- A Little Amount 4 -- A Moderate Amount 5 -- An Enormous Amount Codebook information for Knowledge Areas Required for Each Task for Project Financing (tk3) Question 3 (of 36) Here is a matrix containing a list of knowledge areas (the columns) and tasks (the rows) that were identified by people in your team for project financing. We would like to know the degree to which each knowledge area is needed to complete each task. Please indicate the degree of required knowledge for each task. Please use the scrollbar at the bottom of the screen to view the last column if necessary. This datatype has 10 columns, labeled tk3_x_y, where x is: 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents 3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications
254
and y is 0 -- Regulatory/Authority Requirements 1 -- AEC (Architecture- Engineering- Construction) The value for each entry is: 0 -- No Response 1 -- I Don't Know 2 -- None 3 -- A Little Amount 4 -- A Moderate Amount 5 -- An Enormous Amount Codebook information for Interrelated Tasks (ti) Question 4 (of 36) Here is a matrix containing the list of tasks that were identified by people in your team. We would like to know which tasks are interrelated. Please check the box for each pair of interrelated tasks. For example, if Obtaining Financial/Service Program Feedbacks is related to Review Architectural- Engineering- Construction (AEC) Documents, you would check the box in row 3, column 1. You may check more than one box in each row. This data type has 25 columns, labeled ti_x_y, where x is: 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents 3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications and y is 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents 3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications The value for each entry is:
255
Codebook information for Group Use of Information (tm) Question 5 (of 36) The following items concern how knowledge is utilized among your team's members: There are 12 columns in matrix, labeled 'tm_x', where x is: 0 -- Most of my work is done independently. 1 -- Members of my team have a lot of overlapping knowledge. 2 -- Each member has unique knowledge that they bring to our team. 3 -- I depend very much on the expertise of other members of my team in order to do my job. 4 -- I depend very much on the expertise of other people outside my team in order to do my job. 5 -- I work very closely with other team members. 6 -- I know a lot about the expertise of my team members. 7 -- My team members know a lot about my expertise. 8 -- My team members know a lot about one another's expertise. 9 -- My team coordinates knowledge well. 10 -- Each member of my team has a specialized role. 11 -- Members of my team have interchangeable roles. The range of each of these columns is: 0 -- No Response 1 -- I Don't Know 2 -- Strongly Disagree 3 -- Disagree 4 -- Neither 5 -- Agree 6 -- Strongly Agree Codebook information for Responsibility (tr) Question 6 (of 36) We would like to know what level of responsibility you think the members of your team (including yourself) has in each of the tasks listed at the bottom of the screen. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how much responsibility you think each member in your team has in each task. Click on the box you think represents that member's amount of responsibility. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear.
256
After this is completed, you will have the opportunity to change your answers. To change a response, click on the token (small shape) under the person's name and the pop-up will re-appear. Select a new value and the pop-up will disappear. There are 95 columns in the matrix, labeled 'tr_x_y', where x corresponds to the id of the actor being ranked, and y is: 0 -- Obtaining financial and service program evaluations 1 -- Reviewing financial feasibility 2 -- Reviewing AEC documents 3 -- Coordinating regulatory compliance 4 -- Preparing funding program applications The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Not Responsible 3 -- Secondary 4 -- Primary Codebook information for k1 (ko1) Question 7 (of 36) We would like to know what level of knowledge you think the members of your team (including yourself) has in each of the knowledge areas listed at the bottom of the screen during the "concept design-development approval" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select what level of knowledge each member in your team has in each knowledge area. Click on the box you think represents that member's level of knowledge. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed, you will have the opportunity to change your answers. To change a response, click on the token (small shape) under the person's name and the pop-up will re-appear. Select a new value and the pop-up will disappear. Due to technical reasons, the shape for "Expert" and "None" are the same. This will not interfere with the information you provide.
257
There are 57 columns in the matrix, labeled 'ko1_x_y', where x corresponds to the id of the actor being ranked, and y is: 0 -- Regulatory/Authority Requirements 1 -- Development project finance 2 -- AEC issues The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- None 3 -- Beginner 4 -- Intermediate 5 -- Expert Codebook information for k1 (ko2) Question 8 (of 36) We would like to know what level of knowledge you think the members of your team (including yourself) has in each of the knowledge areas listed at the bottom of the screen during the "development approval-start construction" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select what level of knowledge each member in your team has in each knowledge area. Click on the box you think represents that member's level of knowledge. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed, you will have the opportunity to change your answers. To change a response, click on the token (small shape) under the person's name and the pop-up will re-appear. Select a new value and the pop-up will disappear. Due to technical reasons, the shape for "Expert" and "None" are the same. This will not interfere with the information you provide. There are 38 columns in the matrix, labeled 'ko2_x_y', where x corresponds to the id of the actor being ranked, and y is: 0 -- AEC Documentation process 1 -- Bidding process
258
The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- None 3 -- Beginner 4 -- Intermediate 5 -- Expert Codebook information for k1 (ko3) Question 9 (of 36) We would like to know what level of knowledge you think the members of your team (including yourself) has in each of the knowledge areas listed at the bottom of the screen during project financing. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select what level of knowledge each member in your team has in each knowledge area. Click on the box you think represents that member's level of knowledge. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed, you will have the opportunity to change your answers. To change a response, click on the token (small shape) under the person's name and the pop-up will re-appear. Select a new value and the pop-up will disappear. Due to technical reasons, the shape for "Expert" and "None" are the same. This will not interfere with the information you provide. There are 38 columns in the matrix, labeled 'ko3_x_y', where x corresponds to the id of the actor being ranked, and y is: 0 -- Regulatory/Authority Requirements 1 -- AEC (Architecture- Engineering- Construction) The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- None 3 -- Beginner 4 -- Intermediate 5 -- Expert
259
Codebook information for Prior Collaboration (pcol) Question 10 (of 36) Prior Collaboration Using the adjacent screen, please indicate any team member(s) with whom you have collaborated prior to joining the Riverwood Place Project by clicking on their name(s). Start with the name at the top and move clockwise around the circle so that you do not miss anyone. If you need to delete an entry, select the diamond on the line and press the delete key. There are 19 columns in matrix, labeled 'pcol_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- Collaborate Codebook information for Current Collaboration (ccol) Question 11 (of 36) Current Collaboration Using the adjacent screen, please indicate any team member(s) with whom you have collaborated since joining the Riverwood Place Project by clicking on their name(s). Start with the name at the top and move clockwise around the circle so that you do not miss anyone. To delete an entry, select the diamond on the line and press the delete key. There are 19 columns in matrix, labeled 'ccol_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- Collaborate Codebook information for Frequency of Communication (fc) Question 12 (of 36) Communication This is the last question for the first section of this exercise. We would like to know how often you think the members of your team (including yourself) communicate (either via telephone, email, or face-to-face) with one another.
260
1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to evaluate the frequency of communication between each pair of members in your team. Click on the box corresponding closest to the frequency of communication between the two members. 3. Repeat step 2 for each pair of members in your team. 4. When all pairs have been answered, the pop-up window will disappear. At that point, you may either click on "Next Question >>" to continue the exercise, or you may click here to to change your response(s). There are 361 columns in matrix, labeled 'fc_x_y', where x is the id of actor 1 and y the id of actor 2. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Less than once per year 4 -- Less than once every 6 months 5 -- Less than once every month 6 -- Less than once a week 7 -- Less than once per day 8 -- Once per day Codebook information for Getting Information About Regulatory/Authority Requirements for the "Concept Design-Development Approval" Period (cra1_a) Question 13a (of 36) Getting Information About Regulatory/Authority Requirements during the "Concept Design-Development Approval" Period In your work, you may need information about Regulatory/Authority Requirements that you do not possess. Using the adjacent screen, please indicate how often you have retrieved information about Regulatory/Authority Requirements from your colleagues during the "Concept Design-Development Approval" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have retrieved information from other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear.
261
After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 13a-13c. When you are completing questions 13b-13c, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cra1_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Getting Information About Development project finance for the "Concept Design-Development Approval" Period (cra1_b) Question 13b (of 36) Getting Information About Development project finance during the "Concept Design-Development Approval" Period Please indicate how often you have retrieved information about development project finance from your colleagues during the "Concept Design-Development Approval" period. There are 19 columns in matrix, labeled 'cra1_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
262
Codebook information for Getting Information About AEC issues for the "Concept Design-Development Approval" Period (cra1_c) Question 13c (of 36) Getting Information About AEC issues during the "Concept Design-Development Approval" Period Please indicate how often you have retrieved information about development project finance from your colleagues during the "Concept Design-Development Approval" period. There are 19 columns in matrix, labeled 'cra1_c_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Getting Information About AEC Documentation process during "Development Approval-Start Construction" Period (cra2_a) Question 14a (of 36) Getting Information About AEC Documentation process during the "Development Approval-Start Construction" Period In your work, you may need information about AEC Documentation process that you do not possess. Using the adjacent screen, please indicate how often you have retrieved information about AEC Documentation process from your colleagues during the "Development Approval-Start Construction" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have retrieved information from other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear.
263
After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 14a-14b. When you are completing question 14b, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cra2_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Getting Information About Bidding process during "Development Approval-Start Construction" Period (cra2_b) Question 14b (of 36) Getting Information About Bidding process during the "Development Approval-Start Construction" Period Please indicate how often you have retrieved information about Bidding process during the "Development Approval-Start Construction" period. There are 19 columns in matrix, labeled 'cra2_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
264
Codebook information for Getting Information About Regulatory/Authority Requirements for Project Financing (cra3_a) Question 15a (of 36) Getting Information About Regulatory/Authority Requirements for Project Financing Activities In your work, you may need information about Regulatory/Authority Requirements that you do not possess. Using the adjacent screen, please indicate how often you have retrieved information about Regulatory/Authority Requirements from your colleagues for project financing activities. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have retrieved information from other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 15a-15b. When you are completing question 15b, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cra3_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
265
Codebook information for Getting Information About AEC (Architecture- Engineering- Construction) for Project Financing (cra3_b) Question 15b (of 36) Getting Information About AEC (Architecture- Engineering- Construction) for Project Financing Activities Please how often you have retrieved information about AEC (Architecture- Engineering- Construction) during project financing activities. There are 19 columns in matrix, labeled 'cra3_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Information Needs (amtinf) Question 16 (of 36) In responding to the following items, keep in mind activities that you have worked on in a month. There are 6 columns in matrix, labeled 'amtinf_x', where x is: 0 -- In most cases, I have the knowledge I need to make decisions. 1 -- I have enough knowledge to meet my professional objectives effectively. 2 -- The amount of information available to me is sufficient for me to make good decisions. 3 -- Most information I receive is very valuable. 4 -- I have found that information is generally complete enough for me to make good decisions. 5 -- I have full confidence that I make decisions based on accurate information. The range of each of these columns is: 0 -- No Response 1 -- I Don't Know 2 -- Strongly Disagree 3 -- Disagree
266
4 -- Neither 5 -- Agree 6 -- Strongly Agree Codebook information for Information Available (reffic) Question 17 (of 36) Everything you need to do at work takes time and energy. In responding to the following statements, think about the time and energy it takes to do the job well. There are 3 columns in matrix, labeled 'reffic_x', where x is: 0 -- Gathering all the information I need for my work takes too much time. 1 -- I am able to accomplish my work goals easily with the resources I have. 2 -- It is difficult for me to do my job well given the information usually available to me. The range of each of these columns is: 0 -- No Response 1 -- I Don't Know 2 -- Strongly Disagree 3 -- Disagree 4 -- Neither 5 -- Agree 6 -- Strongly Agree Codebook information for Generated Information (reshet) Question 18 (of 36) In general, the information I generate is... There are 3 columns in matrix, labeled 'reshet_x', where x is: 0 -- too specific to my projects to be useful to anyone else in the organization. 1 -- useful to other departments/teams. 2 -- relevant to projects that colleagues are working on. The range of each of these columns is: 0 -- No Response 1 -- I Don't Know 2 -- Strongly Disagree 3 -- Disagree
267
4 -- Neither 5 -- Agree 6 -- Strongly Agree Codebook information for Providing Information About Regulatory/Authority Requirements during the "Concept Design-Development Approval" Period (cai1_a) Question 19a (of 36) Providing Information About Regulatory/Authority Requirements during the "Concept Design-Development Approval" Period In your work, you may receive of create information about Regulatory/Authority Requirements. Using the adjacent screen, please indicate how often you have provided unsolicited information about Regulatory/Authority Requirements to your colleagues during the "Concept Design-Development Approval" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have provided information to other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 19a-19c. When you are completing questions 19b-19c, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cai1_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
268
Codebook information for Providing Information About Development project finance during the "Concept Design-Development Approval" Period (cai1_b) Question 19b (of 36) Providing Information About Development project finance during the "Concept Design-Development Approval" Period Please indicate how often you have provided information about Development project finance during the "Concept Design-Development Approval" period. There are 19 columns in matrix, labeled 'cai1_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Providing Information About AEC issues during the "Concept Design-Development Approval" Period (cai1_c) Question 19c (of 36) Providing Information About AEC issues during the "Concept Design-Development Approval" Period Please indicate how often you have provided information about AEC issues during the "Concept Design-Development Approval" period. There are 19 columns in matrix, labeled 'cai1_c_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
269
Codebook information for Providing Information About AEC Documentation process during the "Development Approval-Start Construction" Period (cai2_a) Question 20a (of 36) Providing Information About AEC Documentation process during the "Development Approval-Start Construction" period In your work, you may receive of create information about AEC Documentation process. Using the adjacent screen, please indicate how often you have provided unsolicited information about AEC Documentation process to your colleagues during the "Development Approval-Start Construction" period. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have provided information to other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear. After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 20a-20b. When you are completing question 20b, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cai2_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often
270
Codebook information for Providing Information About Bidding process during the "Development Approval-Start Construction" Period (cai2_b) Question 20b (of 36) Providing Information About Bidding process during the "Development Approval-Start Construction" period Please indicate how often you have provided information about Bidding process during the "Development Approval-Start Construction" period. There are 19 columns in matrix, labeled 'cai2_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Providing Information About Regulatory/Authority Requirements for Project Financing (cai3_a) Question 21a (of 36) Providing Information About Regulatory/Authority Requirements for Project Financing Activities In your work, you may receive of create information about Regulatory/Authority Requirements. Using the adjacent screen, please indicate how often you have provided unsolicited information about Regulatory/Authority Requirements to your colleagues for project financing activities. 1. Click in the middle of the circle on the adjacent screen and a small window will pop up. 2. In the small window, you are asked to select how often you have provided information to other colleagues during the project phase(s) that pertain(s) to your work. Click on the box you think represents your best estimate. 3. Repeat step 2 for each member in your team. 4. When you have responded for all members, the pop-up window will disappear.
271
After this is completed you will have the opportunity to change your answers. To change a response, click on the colleague's name and the pop-up will re-appear. Select the value and the pop-up will disappear. These are the instructions for questions 21a-21b. When you are completing question 21b, if you need the entire instructions you may return to this question by using the "<< Previous Question" button. There are 19 columns in matrix, labeled 'cai3_a_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know 2 -- Never 3 -- Seldom 4 -- Sometimes 5 -- Often 6 -- Very Often Codebook information for Providing Information About AEC (Architecture- Engineering- Construction) for Project Financing (cai3_b) Question 21b (of 36) Providing Information About AEC (Architecture- Engineering- Construction) for Project Financing Activities Please indicate how often you have provided information about AEC (Architecture- Engineering- Construction) for project financing activities. There are 19 columns in matrix, labeled 'cai3_b_x', where x is the id of the user this relates to. The range of each of these columns is: 0 -- - 1 -- I don't know
3 -- Seldom 2 -- Never
4 -- Sometimes 5 -- Often 6 -- Very Often
272
Codebook information for Work Products (ptii)
In general, my work products...
2 -- require input from someone else in the team.
4 -- are completed in a team approach with others in the team.
0 -- No Response 1 -- I Don't Know
4 -- Neither
6 -- Strongly Agree
Codebook information for Departmental Projects (ptid)
0 -- is independent of other MidPen's project' teams. 1 -- utilizes input from other projects in my own company.
0 -- No Response
2 -- Strongly Disagree 3 -- Disagree 4 -- Neither
Question 22 (of 36)
There are 6 columns in matrix, labeled 'ptii_x', where x is: 0 -- feed into work done by someone else in the team. 1 -- feed into work done by someone else outside the team.
3 -- require input from someone else outside the team.
5 -- are completed in a team approach with others outside the team. The range of each of these columns is:
2 -- Strongly Disagree 3 -- Disagree
5 -- Agree
Question 23 (of 36)
The Riverwood Place Project team... There are 4 columns in matrix, labeled 'ptid_x', where x is:
2 -- completes the project together with other MidPen project teams. 3 -- completes its individual member tasks within the prescribed deadline.
The range of each of these columns is:
1 -- I Don't Know
5 -- Agree 6 -- Strongly Agree
273
Codebook information for tc (tc)
On the following scale, rate the amount of time that it takes your team to
There are 5 columns in matrix, labeled 'tc_x', where x is: 0 -- Obtaining Financial/Service Program Feedbacks
2 -- Review Architectural- Engineering- Construction (AEC) Documents
The range of each of these columns is:
3 -- Longer Than Anticipated
5 -- Shorter Than Anticipated 6 -- Much Shorter Than Anticipated
Irrespective of the time you spent on them, rate the quality with which your team accomplishes each of the following tasks.
3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications The range of each of these columns is:
4 -- Average Quality
Question 24 (of 36)
accomplish the following tasks.
1 -- Review Financial Feasibility
3 -- Coordinate Regulatory Compliance 4 -- Prepare Funding Program Applications
0 -- No Response 1 -- I Don't Know 2 -- Much Longer Than Anticipated
4 -- As Anticipated
Codebook information for qc (qc) Question 25 (of 36)
There are 5 columns in matrix, labeled 'qc_x', where x is: 0 -- Obtaining Financial/Service Program Feedbacks 1 -- Review Financial Feasibility 2 -- Review Architectural- Engineering- Construction (AEC) Documents
0 -- No Response 1 -- I Don't Know 2 -- Very Poor Quality 3 -- Below Average Quality
5 -- Above Average Quality 6 -- Excellent Quality
274
Codebook information for Satisfaction (satprg)
and with your work in general.
projects?
2 -- Very Dissatisfied
1 -- 1-20% 2 -- 21-40% 3 -- 41-60% 4 -- 61-80%
Question 26 (of 36) The following items concern how satisfied you were with team project decisions,
There are 6 columns in matrix, labeled 'satprg_x', where x is: 0 -- How satisfied are you with the quality of your team's decisions? 1 -- How satisfied are you with your work? 2 -- How satisfied are you with the work of your team? 3 -- How satisfied are you with how well your team coordinates who will do what? 4 -- How satisfied are you with how well your team works together as a unit? 5 -- How satisfied are you with the efficiency of your team in completing
The range of each of these columns is: 0 -- No Response 1 -- I Don't Know
3 -- Dissatisfied 4 -- Neither 5 -- Satisfied 6 -- Very Satisfied Codebook information for Remote Work (remowork) Question 27 (of 36) What percentage of your work is done remotely (e.g., at home, in the car, while traveling)?
There is a single column 'remowork' in the matrix. Values are: 0 -- 0% of My Work
5 -- 81-100%
275
Codebook information for Job Type (jobtyp1)
Values are: 0 -- Overall Team Leader 1 -- Section Leader - I coordinated several members' tasks 2 -- Team Member - No member reports to me
There is a single column 'jobtyp2' in the matrix.
1 -- Section Leader - I coordinated several members' tasks 2 -- Team Member - No member reports to me
There is a single column 'jobtyp3' in the matrix.
Values are:
2 -- Team Member - No member reports to me
Question 28 (of 36) During the "concept design-development approval" period, which category best describes your responsibilities?
There is a single column 'jobtyp1' in the matrix.
Codebook information for Job Type (jobtyp2) Question 29 (of 36)
During the "development approval-start construction" period, which category best describes your responsibilities?
Values are: 0 -- Overall Team Leader
Codebook information for Job Type (jobtyp3) Question 30 (of 36) During "project financing activities", which category best describes your responsibilities?
0 -- Overall Team Leader 1 -- Section Leader - I coordinated several members' tasks
276
Codebook information for Years Worked, Company (tpr)
0 -- Years
2 -- Weeks
-3 -- No Response
0 -- 0
10 -- 10
Question 31 (of 36) To the best of your knowledge, how long have you worked with the Riverwood Place team? There are 3 columns in matrix, labeled 'tpr_x', where x is:
1 -- Months
The range of each of these columns is:
-2 -- I Don't Know -1 --
1 -- 1 2 -- 2 3 -- 3 4 -- 4 5 -- 5 6 -- 6 7 -- 7 8 -- 8 9 -- 9
11 -- 11 12 -- 12 13 -- 13 14 -- 14 15 -- 15 16 -- 16 17 -- 17 18 -- 18 19 -- 19 20 -- 20 21 -- 21 22 -- 22 23 -- 23 24 -- 24 25 -- 25 26 -- 26 27 -- 27 28 -- 28 29 -- 29
277
30 -- 30 31 -- 31 32 -- 32 33 -- 33 34 -- 34
-3 -- No Response
0 -- 0
10 -- 10
12 -- 12
14 -- 14
35 -- 35 36 -- 36 37 -- 37 38 -- 38 39 -- 39 40 -- 40 Codebook information for Years Worked, Position (tp) Question 32 (of 36)
To the best of your knowledge, how long have you worked for Mid-Peninsula Housing Coalition? There are 3 columns in matrix, labeled 'tp_x', where x is: 0 -- Years 1 -- Months 2 -- Weeks
The range of each of these columns is:
-2 -- I Don't Know -1 --
1 -- 1 2 -- 2 3 -- 3 4 -- 4 5 -- 5 6 -- 6 7 -- 7 8 -- 8 9 -- 9
11 -- 11
13 -- 13
15 -- 15 16 -- 16
278
17 -- 17 18 -- 18 19 -- 19 20 -- 20 21 -- 21 22 -- 22 23 -- 23 24 -- 24
26 -- 26
30 -- 30
39 -- 39 40 -- 40
Question 33 (of 36)
In what year were you born?
There are 2 columns in matrix, labeled 'birth_x', where x is:
-3 -- No Response -2 -- I Don't Know -1 --
2 -- 2 3 -- 3
5 -- 5
25 -- 25
27 -- 27 28 -- 28 29 -- 29
31 -- 31 32 -- 32 33 -- 33 34 -- 34 35 -- 35 36 -- 36 37 -- 37 38 -- 38
Codebook information for Year (birth)
0 -- Decades 1 -- Years The range of each of these columns is:
0 -- 0 1 -- 1
4 -- 4
279
6 -- 6 7 -- 7 8 -- 8
Codebook information for Education (educ)
9 -- 9 10 -- 10 Codebook information for Gender (gender) Question 34 (of 36) What is your gender? There is a single column 'gender' in the matrix. Values are: 0 -- Male 1 -- Female
Question 35 (of 36) What is the highest level of education you received? There is a single column 'educ' in the matrix.
Values are: 0 -- Some high school 1 -- High school diploma 2 -- Some college 3 -- Associate's degree 4 -- Bachelor's degree 5 -- Master's degree 6 -- Doctorate degree Codebook information for Comments Question 36 (of 36)
280
BIBLIOGRAPHY
Burton, R. M., and B. Obel. 2003. Strategic organizational diagnosis and design: Developing theory for application. Boston: Kluwer Academic Publishers.
Carley, K., and M. Prietula. 1994. Computational organization theory. Hillsdale, NJ: Erlbaum.
Alavi M., and D. E. Leidner. 2001. Review: Knowledge management and knowledge
management systems: Conceptual foundations and research issues. MIS Quarterly 25 (1): 107-136.
Allmendinger, J., and J. R. Hackman. 1996. Organizations in changing environments:
The case of East German symphony orchestras. Administrative Science Quarterly 41 (3): 337-369.
American Institute of Architects. 1997. Abbreviated standard form of agreement between
owner and architect. AIA document B151-1997. New York: The American Institute of Architects.
Bacharach, S. B., P. Bamberger, et al. 1996. The organizational transformation process:
The micropolitics of dissonance reduction and the alignment of logics of action. Administrative Science Quarterly 41 (3): 477-506.
Bookout, L. W. Jr. 1990. Residential development handbook. Washington, D.C.: ULI-
Urban Land Institutes. Burton, R. M., and B. Obel. 1995. The validity of computational models in organization
science: From model realism to purpose of the model. Computational and Mathematical Organization Theory (1): 57-71.
California Redevelopment Association. 1998. California affordable housing handbook:
Strategies for planning and development. Sacramento, CA: California Redevelopment Association.
Carillo, P., H. Robinson, A. Al-Ghassani, and A. Chimay. 2004. Knowledge management
in UK: Construction, strategies, resources and barriers. Project Management Journal 35 (1): 46-56.
Carlile, P. R., and E. S. Rebentisch, E. S. 2003. Into the black box: The knowledge
transformation cycle. Management Science 49 (9): 1180-1195. Christiansen, T. 1993. Modeling efficiency and effectiveness of coordination in
engineering design teams: VDT-the Virtual Design Team. Ph.D. diss., Department of Civil and Environmental Engineering, Stanford University.
281
Cohen, G.P. 1992. The Virtual Design Team: An object-oriented model of information sharing in project team. Ph.D. Diss., Department of Civil and Environmental Engineering, Stanford University.
Cohen, W. M., and D. A. Levinthal. 1990. Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly 35 (1, Special Issue: Technology, Organizations, and Innovation): 128-152.
Contractor, N., D. Brandon, M. Huang, E. T. Palazzolo, D. Steinley, C. Su, V. R. Suri, S.
A. Swarbrick, J. Templin, and S. Wasserman (in review). Multi-theoretical multilevel models of information retrieval in knowledge networks.
Cresswell, J. W. 2003. Research design: Qualitative, quantitative, and mixed methods
approach. Thousand Oaks: Sage Publications, Inc. Cyert, R. M., and J. G. March. 1963. A behavioral theory of the firm. Englewood Cliffs:
Prentice-Hall. Daft, R. 1992 in Burton and Obel 2003. Organization Theory and Design (4th Ed.) St.
Paul, MN: West. (2nd ed., 1986; 3rd ed., 1989): p. 286. Davenport, T.H., and L. Prusak. 1998. Successful knowledge management projects.
Sloan Management Review 37 (4): 53-65. Depress C., and D. Chauvel. 1999. Mastering information management: Part six-
knowledge management. Financial Times (8 March): 4-6. Doty, D. H., W. H. Glick, and G. P. Huber. 1993. Fit, equifinality, and organizational
effectiveness: A test of two configurational theories. Academy of Management Journal 36: 1196-1250.
Dyck, B., and F. A. Starke. 1999. The formation of breakaway organizations:
observations and a process model. Administrative Science Quarterly 44 (4): 792-822.
Edwards, J., H. Handzic, S. Carlsson, and M. Nissen. 2003. Knowledge management
research and practice: Visions and directions. Knowledge Management Research & Practice 1 (1): 49-60.
Fulton, W. 1999. Guide to California planning. Point Arena, CA: Solano Press Books. Galbraith, J. R. 1974. Organization design: An information processing view. Interfaces 4
(3): 28-36. Galbraith, J. R. 1977. Organization Design. Reading, Massachusetts: Addison-Wesley.
282
Garrett, James H., I. J. Flood, et al. 2004. Information technology in civil engineering- Future trends. Journal of Computing in Civil Engineering 18 (3): 185-186.
Glasser, B. G., and A. L. Strauss. 1999. The discovery of grounded theory: Strategies for
qualitative research. New York: Aldine De Gruyter. Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management
Journal 17 (Special Issue: Knowledge and the Firm): 109-122. Hollingshead, A. B. 1998. Cognitive interdependence and convergent expectations.
Journal of Personality and Social Psychology 81 (6): 1080-1089. Hollingshead, A. B. 1998a. Retrieval processes in transactive memory systems. Journal
of Personality and Social Psychology 74 (3): 659-671. Hollingshead, A. B. 1998b. Communication, learning, and retrieval in transactive
memory systems. Journal of Experimental Social Psychology 34: 423-442. Huber, G. P., C. C. Miller, and W. H. Glick. 1990. Developing more encompassing
theories about organizations: The centralization-effectiveness relationship as an example. Organization Science 1: 11-40.
Ibrahim, R. 2001. Feasibility of 4D CAD in design development and approval process for
affordable housing. Engineer Diss., Department of Civil and Environmental Engineering, Stanford University.
Ibrahim, R., and B. C. Paulson, Jr. 2005. Discontinuity in organizations: How
environmental characteristics contribute to the project's knowledge loss phenomenon. CRGP Working Paper No. 12, Stanford University.
Ibrahim, R., and M. E. Nissen. 2003. “Emerging technology to model dynamic
knowledge creation and flow among construction industry stakeholders during the critical feasibility-entitlements phase,” Proceedings of The ASCE 4th Joint Symposium on IT in Civil Engineering in Nashville, Tennessee, November 15-16 by the Association of Civil Engineers.
Ibrahim, R., and M. E. Nissen. 2004. “Simulating environmental contingencies using
SimVision®,” Proceedings of the NAACSOS Conference 2004 in Pittsburgh, Pennsylvania, June 27-29 by the North American Association for Computational Social and Organization Science.
Ibrahim, R., and M. E. Nissen. 2005. “Developing a knowledge-based organizational
performance model for discontinuous participatory enterprises,” Proceedings of the Thirty-Eighth Annual Hawaii International Conference on System Sciences in Big Island, Hawaii January 3-6 by College of Business Administration, University of Hawaii at Manoa.
283
Ibrahim, R., M. Shumate, R. Levitt, and N. Contractor. 2005. Discontinuity in
organizations: Knowledge flow behaviors in sequential workflow processes. CRGP Working Paper No. 17, Stanford University.
Ibrahim, R., R. Levitt, and M. Ramsey. 2005b. Discontinuity in organizations: Impacts of
knowledge flows on organizational performance. CRGP Working Paper No. 18, Stanford University.
Jablin, F., and K. J. Krone. 1987. Organizational assimilation. Newbury Park, California:
Sage. Jin, Y., and R. E. Levitt. 1996. The virtual design team: A computational model of
project organizations. Computational and Mathematical Organization Theory 2 (3): 171-196.
Johnson, J. D. 1992. Approaches to organizational communication structure. Journal of
Business Research 25: 99-113. Johnson, J. D. 1993. Organizational communication structure. Norwood, NJ: Ablex. Kogut, B., and U. Zander. 1992. Knowledge of the firm, combinative capabilities, and the
replication of technology. Organization Science 3 (3, Focused Issue: Management of Technology): 383-397.
Kone, L. D. 1994. Land development. Washington, D.C.: National Association of Home
Builders. Krikorian, D. D., D. R. Seibold, and P. L. Goode. 1997. Reengineering at LAC: A case
study of emergent network processes. Edited by B. D. Sypher. Case studies in organizational communication 2: Perspectives on contemporary work life (129-144). New York: Guilford.
Kunz, J.C., R. E. Levitt, and Y. Jin. 1998. The virtual design team: A computational
simulation model of project organizations. Communications of the Association for Computing Machinery 41 (11): 84-92.
Lambert, M. H., and B. Shaw. 2002. Transactive memory and exception handling in
high-performance project teams. CIFE Technical Report No. 137, Stanford University.
Landis, J. (ed.). 2001. Pay to play: Residential development fees in California cities and
counties, 1999. Department of Housing and Community Development, Division of Housing Policy Development, California.
284
Lawrence, P. R., and J. W. Lorsch 1967. Organization and environment: Managing differentiation and integration. Boston: Graduate School of Business Administration, Harvard University.
Levinthal, D. A. 1991. Organizational adaptation and environmental selection-
interrelated processes of change. Organization Science 2 (1, Special Issue: Organizational Learning: Papers in Honor of (and by) James G. March): 140-145.
Levitt, R. E., J. Thomsen, T. R. Christiansen, et al. 1999. Simulating project work processes and organizations: Towards a micro-contingency theory of organizational design. Management Science 45: 1479-1495.
Levitt, B., and J. G. March. 1988. Organizational learning. Annual Review of Sociology
14: 319-340. Levitt, R. E., and M. E. Nissen. 2002. Toward simulation models of knowledge-intensive
work processes. CIFE Working Paper #77, Stanford University. Levitt, R. E., G. P. Cohen, J. C. Kunz, C. I. Nass, T. R. Christiansen, and Y. Jin. 1994.
The Virtual Design Team: Simulating how organization structure and information processing tools affect team performance. Edited by Carley, K. M. and M. J. Prietula. Computational and Mathematical Organization Theory. Hillsdale, NJ: Lawrence Erlbaum.
Lewin, A. Y., and H. W. Volberda. 1999. Prolegomena on coevolution: A framework for
research on strategy and new organizational forms. Organization Science 10 (5, Focused Issue: Coevolution of Strategy and New Organizational Forms): 519-534.
March, J. G., and H. A. Simon. 1958. Organizations. New York: John Wiley & Sons,
Inc. Marwell, G., and P. Oliver. 1993. The critical mass in collective action: A micro-social
theory. Cambridge, UK: Cambridge University Press. McPhee, R. D., and M. S. Poole. 2001. Órganizational structures and configurations.
Edited by F. M. Jablin and L. L. Putnam. The new handbook of organizational communication: Advances in theory, research, and methods (503-543). Thousand Oaks, CA: Sage.
Miller, C. C., W. H. Glick, Y. D. Wang, and G. P. Huber. 1991. Understanding
technology-structure relationships: Theory development and meta-analytic theory testing. Academy of Management Journal 34: 370-399.
285
Mintzberg, H., and F. Westley. 1992. Cycles of organizational change. Strategic Management Journal 13 (Special Issue: Fundamental Themes in Strategy Process Research): 39-59.
Monge, P. E., and N. S. Contractor. 2003. Theories of communication networks. Oxford,
UK: Oxford University Press.
Moreland, R. 1999. Transactive memory: Learning who knows what in work groups and organizations. In J. Levine (Ed.) Shared Cognition in Organizations: The Management of Knowledge. Hillsdale, NJ: Lawrence Erlbaum Associates.
Nadler, J., L. Thompson, and L. van Boven. 2003. Learning negotiation skills: Four
models of knowledge creation and transfer. Management Science 49 (4): 529-540.
Neighborhood Reinvestment Corporation. 1994. How to be your own developer: Making the development decision. Washington, D.C.: Neighborhood Reinvestment Corporation.
Nelson, R. R., and S. G. Winter. 1982 (in Scott 1998). An evolutionary theory of economic change. Cambridge, MA: Belknap Press of Harvard University Press.
Nissen, M. E. (in review). Understanding knowledge-flow patterns for design. Nissen, M. E. 2002. An extended model of knowledge-flow dynamics. Communications
of the Association for Information Systems 8: 251-266. Nissen, M. E. 2005a. Harnessing knowledge dynamics: Principled organizational
knowing and learning. Hershey, PA: Idea Group Publishing (forthcoming).
Nissen, M. E. 2005b. Toward designing organizations around knowledge flows. In: K. Desouza (Ed.), New frontiers in knowledge management (forthcoming).
Nissen, M. E. 2005c. A computational approach to diagnosing misfits, inducing
requirements, and delineating transformations for Edge Organizations. Proceedings International Command and Control Research and Technology Symposium, McLean, VA (June).
Nissen, M. E., and R. E. Levitt. 2002. Dynamic models of knowledge-flow dynamics.
CIFE Working Paper #76 (November), Stanford University. Nissen, M. E., and R. E. Levitt. 2004. Agent-based modeling of knowledge dynamics.
Knowledge Management Research and Practice 2 (3).
Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization Science 5 (1): 14-37.
286
Nonaka, I., and H. Takeuchi. 1995. The knowledge-creating company- How Japanese companies create the dynamics of innovation? New York: Oxford University Press.
Palazzolo, E. T., D. Serb, Y. She, C. Su, and N. S. Contractor (in review). Co-Evolution
of communications and knowledge networks as transactive memory systems: Using computational models for theoretical integration and extensions.
Paulson, B. C., Jr. 1976. Designing to reduce construction cost. Journal of the
Construction Division 102 (CO4), Proc. Paper 12600, December: 587-592.
Peiser, R. B., and D. Schwanke. 1992. Professional real estate development- The ULI guide to the business. Washington, D.C.: Dearborn Financial Publishing, Inc. and ULI-The Urban Land Institute.
Polanyi, M. (1967). The tacit dimension. London, Routledge and Keoan Paul. Krackhardt, D., and J. R. Hanson. 1993. Informal networks: The company behind the
chart. Harvard Business Review 71: 104. Ren, Y. 2004. Why and when does transactive memory matter: Opening the black box.
Proceedings of the NAACSOS Conference 2003 in Pittsburgh, Pennsylvania, June 24-26 by the North American Association for Computational Social and Organization Science.
Roberts, K. H., and C. A. O’Reilly. 1979. Some correlates of communication roles in organizations. Academy of Management Journal 22: 42-57.
Roethlisberger, F., W. Dickson (1939). Management and the worker. New York: Wiley. Schindler, M., and M. J. Eppler. 2003. Harvesting project knowledge: a review of project
learning methods and success factors. International Journal of Project Management 21 (3): 219-228.
Schmitz, A. 2000. Multifamily housing development handbook. Washington, D.C.: ULI-
the Urban Land Institute.
Schreiber, C., and K. Carley. 2003. “The impact of databases on knowledge transfer: Simulation providing theory,” Proceedings of the NAACSOS Conference 2003 in Pittsburgh, Pensylvannia, June 26-28 by the North American Association for Computational Social and Organization Science.
Scott, J. 2000. Social network analysis. Sage Publications, London. Scott, W. R. 2003. Organizations: rational, natural, and open systems. New Jersey:
Prentice Hall.
287
Thompson, J. D. 1967. Organization in action: Social science bases in administrative theory. New York, NY: McGraw-Hill.
U.S. Department of Housing and Urban Development. Available at
Spradley, J. P. 1980. Participant observation. Fort Worth: Harcourt College Publishers. Szulanski, G. 2000. The process of knowledge transfer: A diachronic analysis of
stickiness. Organizational Behavior and Human Decision Processes 82 (1): 9-27.
Thomas-Hunt, M. C., T. Y. Ogden, and M. A. Neale. 2003. Who's really sharing? Effects of social and expert status on knowledge exchange within groups. Management Science 49 (4): 464-477.
Thomsen, J. 1998. Virtual Team Alliance (VDA): Modeling the effects of goal
incongruence in semi-routine,fFast-paced project Ooganizations. Ph. D. Diss., Department of Civil and Environmental Engineering, Stanford University.
Thomsen, J., R. E. Levitt, J. C. Kunz, C. I. Nass, and D. B. Fridsma. 1999. A trajectory for validating computational emulation models of organizations. Journal of Computational & Mathematical Organization Theory 5 (4) Dec: 385-401.
http://www.hud.gov/. Von Hippel, E. 1994. "Sticky information" and the locus of problem solving: Implications
for innovation. Management Science 40 (4): 429-439. Wasserman, S., and K. Faust. 1999. Social network analysis: Methods and applications.
Cambridge: Cambridge University Press. Weber, M. 1947. The theory of social and economic organization (A. H. Henderson and
T. Parsons, Eds. and trans.). Glencoe, IL: Free Press. Wegner, D. M. 1987. Transactive memory: A contemporary analysis of the group mind.
Edited by B. Mullen, and G. R. Goethals. Theories of Group Behavior. New York: Springer-Verlag, 185-208.
Yuan, Y., J. Fulk, M. Shumate, P. Monge, J. A. Bryant, and M. Matsaganis (in review). Individual participation in organizational information commons: the impact of team level social influence and technology-specific competence.
Wegner, D. M. 1999. A computer network model of human transactive memory. Social
Cognition, 13(3): 319-339. Yin, R. K. 2003. Case study research: Design and methods. Thousand Oaks: Sage
Publications, Inc.
288
289
Zellmer-Bruhn, M. E. 2003. Interruptive events and team knowledge acquisition. Management Science 49 (4): 514-528.