methods and methodology - warwick.ac.uk€¦ · web viewtashakkori and teddie (2003) also suggest...

35
Methods and Methodology “Social science is a terminological jungle where many labels compete, and no single label has been able to command the particular domain before us. Often ... researchers simply ‘do it’ without worrying about giving it a name.” Lofland and Lofland (1984) In this chapter, I will consider the different methods that could be applied to a study of ICT use by teachers and the advantages and disadvantages of these methods. I will also consider the methods I have used, their boundaries and limitations. Within educational research different areas can be identified, for example Verma and Mallick (1999) identify seven areas, of which three are related to performance at different levels, i.e. individual, group and school. The other four relate to management of schools with school management, interpersonal relationships, the curriculum and teachers. This study falls into the latter category. Once the problem has been identified there is the question of selecting the appropriate method or methods. There are many approaches to educational research although it is generally recognised that there are two key distinctions. In the quest for the discovery and development or an organised body of knowledge, Cronbach and Suppes (1969) (cited in Verma) suggest that research, or systematic inquiry, is categorised according to purpose and point out that there are different approaches available for research relating to decisions for policy makers (decision orientated) and that which follows the researcher’s interests (conclusion orientated). This study falls into the latter category. 1

Upload: vudien

Post on 04-Jun-2018

214 views

Category:

Documents


0 download

TRANSCRIPT

Methods and Methodology

“Social science is a terminological jungle where many labels compete, and no single label has been able

to command the particular domain before us. Often ... researchers simply ‘do it’ without worrying about

giving it a name.”

Lofland and Lofland (1984)

In this chapter, I will consider the different methods that could be applied to a study of ICT use by

teachers and the advantages and disadvantages of these methods. I will also consider the methods I

have used, their boundaries and limitations. Within educational research different areas can be

identified, for example Verma and Mallick (1999) identify seven areas, of which three are related to

performance at different levels, i.e. individual, group and school. The other four relate to management

of schools with school management, interpersonal relationships, the curriculum and teachers. This study

falls into the latter category.

Once the problem has been identified there is the question of selecting the appropriate method or

methods. There are many approaches to educational research although it is generally recognised that

there are two key distinctions. In the quest for the discovery and development or an organised body of

knowledge, Cronbach and Suppes (1969) (cited in Verma) suggest that research, or systematic inquiry, is

categorised according to purpose and point out that there are different approaches available for

research relating to decisions for policy makers (decision orientated) and that which follows the

researcher’s interests (conclusion orientated). This study falls into the latter category.

Many writers will make distinctions using the terms quantitative and qualitative. Quantitative methods

are concerned with statistics and the data collected is measurable, and countable. Qualitative methods

are not based in statistics, and use different techniques such as coding and content analysis. In the

debate concerning quantitative or qualitative methods a sharp distinction is frequently drawn around

them, not on their appropriateness for the task but concerning their epistemology and ontology (Cohen,

Manion and Morrison, 2007, Guba, 1990). The quantitative and qualitative methods demonstrate a

difference in the language and the way that key ideas are expressed and take on different logical

characters. According to Pring (2004) quantitative methods can be described as those which are

appropriate for the physical world while qualitative methods address:

1

‘that which is distinctive of the personal and social, namely the ‘meanings’ through which

personal and social reality is understood’.

Pring (2004) p.45

However, quantitative and qualitative methods sit at the extremes of a continuous spectrum and in the

past, many researchers aligned themselves to one or other of the two research paradigms (alternatively

named as positivism or objectivity and anti-positivism or subjectivity) and tended to select methods

accordingly. Mixing paradigms is a possible approach, building on their different strengths. In discussing

quantitative and qualitative methods Silverman (1985) wrote,

‘.... that there are distinctive advantages to be gained from the juxtaposition and integration of

these two styles of research ...’.

Tashakkori and Teddie (2003) also suggest that different research methods should be viewed as

complimentary, especially as social science research does not fit into the qualitative-quantitative divide,

whilst Fuller (2002) states that knowledge cannot be about the world unless it is situated in the world

and that its price is any distortion that might occur. Whilst May, (2001) states:

‘Social researchers do not then have to content themselves with one paradigm, as Kuhn would

suggest. Social sciences are dynamic discipline lines within which, depending upon the position of

the researcher within an academic field, other paradigms can be considered. This enables as

understanding and explanation of empirical inquiries, while also adding to the challenging of

assumptions about social life as an important part of research practice.’

May, (2001, p.7)

Other writers will use the terms positivism and interpretivism. Positivism has its basis in the physical

sciences. From these results, generalisations can be made. Pring, (2004), defines these as traditions in

educational research. The first tradition, positivism, is sympathetic with the objective world of physical

things that are measurable and data can be collected using experimental methods. The data sets tend to

be large, allowing generalisations to be constructed. The second, interpretivism, is more concerned with

situations as they exist, inner thoughts and the more private world of people. The first sees educational

research as a subset of social sciences that ‘seek the general laws or conditions which will enable

teachers or policy makers to predict, ‘What will happen if .... ?’, i.e. looking to achieve certain goals

through the use of statistical analysis of data, while interpretivism ‘purports to reveal the

understandings and perceptions of the subjects to research – the phenomenology of the mind.’ As

people (often the subject of study in educational research) are individuals who do not necessarily

conform to a set model, then significant generalisations are not possible. Research tends to look at what

2

is unique and distinctive, adding an interpretation, but modern educational research does not confine

itself exclusively to one or other tradition.

Tradition 1 Tradition 2

Objective world of physical things

Public world of outer reality

Quantitative methods on scientific model

Mind and body publically accessible

Subjective world of meanings

Private world of inner thoughts

Qualitative methods based on phenomenological

exposure

Privately privileged exposure

After Pring (2004) p.33

The two traditions represent the contrasts with the objective world of physical things that conform to a

pattern and a subjective one of meanings.

Verma and Mallick (1999) state:

“A recent trend in educational research has been to a greater emphasis on applied research rather

than basic research .... Many consumers (particularly classroom teachers) agree that there is a

wide gap between research findings and practical applications”

Verma and Mallick (1999) p.39

Scott and Usher (1999) question how educational research is understood. They suggest that the

positivistic, scientific paradigm, as it relates to education, is not sustainable as the researchers act

independently from the subjects of the research to produce generalisations. But for many people, both

within and outside the educational field, this form of research is seen as superior to practical knowledge

and practice is seen as an efficient application of theory developed by professional researchers and not

practitioners. This develops a tension between advice, for example, about how to conduct schools and

teach, and those responsible for implementing this advice, that is the practitioners.

An alternative classification is through the purpose of the research, whether it is basic information or to

be applied in a practical sense. Like the quantitative and qualitative, positivistic and interpretive

methods, this represents a continuum rather than discrete methods. It can be argued that the outcomes

of the basic research can lead to future practical applications.

3

Basic Research Applied Research

Adds to a body of scientific knowledge

Does not necessarily produce results of immediate

practical value

Learning and transfer of training

Developing systematic knowledge

Studies of interaction

Pushes forward boundaries

Vital link for future research

Adding to scientific knowledge is secondary

Solves an immediate practical problem

Solves a specific problem

Information provided for immediate use

Frequently involves surveys

After Verma and Mallick (1999) and Travers (1978)

Verma and Mallick (1999) suggest that there are four types of social science research, pure (or basic),

applied (or field), action and evaluative, each focussing on different types of outcomes. Without the

basic research, much applied research would lead to superficial changes, as there is a need for a deeper

understanding of educational issues, such as teaching methods, and their effect on learning. Pure

research concerns the development of theories by discovering broad generalisations or principles with

the aim of discovering facts and is usually carried out under laboratory or controlled situations. For the

solution of day-to-day problems, and often used in educational research, fieldwork will normally involve

a large number of cases following the application of new knowledge to assess the effect of this. By its

nature it has less control built in that pure research. Frequently carried out through surveys,

observations or interviews, applied research is used to investigate situations such as the implementation

of changes such as teaching methods or strategies following a developmental (basic) stage. The third

type, action research, is found in social studies such as social psychology, social work and education. It is

concerned with immediate application, rather than the development of a theory, and focuses on a

specific problem within a particular setting. Action research typically involves a small sample, for

example ten, with the aim of implementing change. Evaluative research is often used to evaluate

programmes where change has been implemented and can be formative of summative in nature.

Much educational research arises from issues and problems rather than substantive theory. Frequently

researchers will be investigating the same substantive problem but their approaches will differ according

to their own preferred methods and approaches. Attempts to engage the concept of a paradigm, as

Kuhn defined it, have led to some problems within educational research in so much as there is a

tendency to focus on certain methods, some of which are given more credence than others, within

4

research establishments. This, according to McNamara (1979) leads to constraints being applied, when

focus on the appropriateness of different methods ought to be at the fore. McNamara also suggests

that:

“Educational researchers ought not to seek the protection of frameworks and argue that

legitimate criticism of their work is confined within the limits of ‘their’ paradigm.”

Paradigm, originally a term developed by Kuhn, is a theory about relationships in the natural world and

is found in reference to educational research. A paradigm can be thought of as a “basic system of

beliefs” or “world view”, as fits the facts at a particular moment in time, which guides the investigation

encompassing beliefs, values and methods within which research takes place. (Guba and Lincoln, 1989).

Davis (1994 p.148) suggests that it is a set of rules within any field for solving one or more puzzles,

where a puzzle is defined as a scientific question for which it is possible to find a solution to in the near

future. This distinguishes it from the many important and interesting questions that do not have an

answer at any particular stage of progress. However, McNamara (1979) questions the use of the term in

educational research and argues that the term:

“can lead to muddled and unclear thinking amongst educationalists”.

McNamara bases his thinking on the fact that Kuhn’s work was based in a different field, notably

scientific work where knowledge is acquired, then replaced over time as this knowledge develops or is

superseded. By developing the original idea the researcher’s work is cumulative. McNamara suggests

that for educational purposes these two qualities do not form the basis for evaluative research where

the emphasis is on method rather than theory. In the natural world it is possible to conduct research

that can be repeated whilst educational research does not lend itself to laboratory experimentation and

complete replication cannot happen because of the fact it is dealing with human beings, their time

related responses, and the effect of being a participant in the research. Hammond and Wellington

(2012) suggest that ‘positivism and interpretivism are better thought of as paradigms’ rather than

methods (p.161).

Pring (2004) outlined different philosophical positions including positivism, functionalism,

constructivism, post-modernism, interpretive theory, phenomenology, ethno-methodology, and

ethnography while Verma and Mallick (1999) identified seven perspectives, empiricism,

experimentalism, quantitative, qualitative, psychometric, sociometric and interpretive. In considering

some of Verma and Mallick’s perspectives for the study I needed to acquaint myself with their

5

attributes. The starting point for empiricism is an observation or experiment followed by an explanation,

or identifying influential factors. In this case interpretation is after the data collection, to avoid distortion

by the researchers own knowledge, beliefs and attitudes. Experimentalism tests theories or hypotheses

under closely controlled conditions, which can be re-tested to confirm the results. They key to this

perspective is reliability, validity, probability and random sampling. Much discussion has also taken place

concerning the roles of interpretivist and positivist research around the type of data collection and

mode of analysis, with quantitative researchers questioning the reliability, and validity of qualitative

methods. Whilst quantitative data can be counted or measured, and these discrete units are compared

using statistical methods, qualitative methods reflect experiences, feelings or judgements and reflect

social processes to provide meaningful explanations. Using the second method, how the samples are

selected and the data analysed is key for plausibility. Background biographical evidence, e.g. age, is

commonly relevant to these studies. These methods are not mutually exclusive, many studies including

elements of both methods. This study will involve elements of both, but leans more to the use of

qualitative methods in so far as the initial questionnaires will be used to form a focus for the selection of

the participants and the data to be developed and collected in the second phase. Psychometric and

sociometric studies both involve the measurement of factors that are central to, or contributory to, the

topic under study. Psychometric studies are about mental capacities while socio-metric studies look at

the relationships within groups and this can include nationality and ethnic groupings. Interpretive

perspectives are situated within qualitative research but focus on why something has happened from an

insider’s point of view. As a former head of mathematics and ICT coordinator, this includes the

researcher using their experience of settings to ‘read’ the acquired information. This study will involve

an element of this interpretive perspective.

At the heart of the epistemological debate is what constitutes truth, reality and verification. Pring,

(2004) states that:

‘research is often focussed upon people’s ‘perceptions of reality’ where one lot of perceptions is

as good as another’.

In other words, what is the researcher’s position with regard to the situation? This may be influential as

to how the research is carried out and how the findings are interpreted. This suggests that there is no

one reality or truth as realism is ‘socially constructed’, and in which culture and social situations will

have some effect. Perhaps it is more accurate to say that the findings are ‘true in that situation’ rather

than being absolute truth. Objectivity considers ‘how the world is’ and the steps that are taken to

ensure that this is taken into consideration when drawing conclusions by following recognised

6

procedures. However, whilst events may appear to be correlated, there are difficulties in making causal

explanations in that there may be other factors that have influence, such as social networks and

interactions, background, and motivation making exceptions to the rule of which the researcher is

unaware. Another difficulty for the researcher is that it is possible to make different interpretations

relating to the same actions. This arises through, amongst others, difference in motives, intentions, an

ability to understand situations and meanings on part of the researcher (to what degree are they able to

display empathy) and participants. For example, ‘Are you good at using ICT?’ initiates a response relating

to the participant’s conception of using ‘who is good at using ICT’ rather than a recognised scale. Unless

the participant knows the criteria for ‘good’ and can measure themselves against it a statement such as

‘75% of the respondents are good at using ICT’ has little value. The insertion of ‘think they are’ does

qualify the statement but also gives it a different context, i.e. in their view and not the researcher’s.

The interpretation of ‘truth’ also causes some difficulty. The word ‘truth’ suggests that something is

correct, but whether correct in all situations is a matter of debate. The same is true for ‘verification’ in

that not all ‘truths’ are able to be verified but that does not mean that they are ‘false’. This ties in with

developing a ‘theory’ through theoretical interpretations rather than what appears to be common sense

or practical knowledge and understandings, that might agree or disagree with the theoretical stance.

Again, conditions, as with ‘reality’, are influencing reporting. In defining knowledge, there is a distinction

between ‘that which is believed’ and ‘that which is known’, the latter being concerned with ‘truth’,

‘certainty’ and ‘verification’. In the scientific world, defined research outcomes will be obtained when an

experiment is repeated, confirming, or otherwise the truth and reality of the situation. Quantitative

social science methods fall within the ‘positivist’ paradigm where human behaviour is governed by rules

and can be investigated using methods developed in the natural sciences with behaviour regarded as a

response to external environmental stimuli or internal stimuli. The very nature of the research means

the findings lie in the past, with the researcher distanced from them. In the social world, situations do

not always conform to rules, so following these methods has not always enabled socially based

questions to be answered and other methods have been suggested. The ontological debate surrounds

the particular philosophical positions. In qualitative research the research is seen to ‘get to the world’ as

it is in that place and at that time, recognising that the ‘reality’ can never be independent of the person

researching it, the participants nor their experiences, and so an identical study may not produce

identical results.

7

In addition to categorising research along the epistemological and ontological spectrums (for example

positive/interpretive), research has been classified using different points of view, according to discipline

(such as sociological, philosophical, psychological aspects), methods (including historical, descriptive,

and experimental) and types of data collected (examples being interviews, observations and testing

being two examples). As with other methods, these categories are not discrete, but are on a continuum.

In many educational research programmes a combination of research methods will be used as it is

widely accepted that all research methods have strengths and weaknesses and different purposes are

better served with appropriate research formats. Methods involving measurement and statistical

analysis of the results, i.e. that which are quantifiable, served well those situations that require hard

facts (such as examination results, finances), where large amounts of data can be collected, and the

absence of the need for the researcher to be involved in the situation. Having roots in the natural

sciences it is possible to use control and experimental groups with a systematic design, the process

moving from identifying the problem, developing a hypothesis, identifying variables, deciding how to

control them and selecting the variable to be manipulated to design. The process then follows with a

sample selection, pilot, final experiment followed by data processing and analysis. The sample needs to

be large and carefully selected. By manipulating an independent variable under controlled conditions,

the researcher measures the effect on the situation or item under investigation. Controlling other

relevant variables aims to avoid contaminating the results.

However, not all problems and issues can be addressed in this measurable manner and a different

approach is more appropriate. Within qualitative methods, there is more of a concern for the individual

with behaviour having meaning and being intentional. There is a plethora of terms associated with the

qualitative type of enquiry, many of which have a degree of overlap, because persons who profess to

practice qualitative research tend to take different views of what this implies. Some researchers prefer

the terms ‘interpretive’ or ‘naturalistic’ research rather than qualitative to describe their form of work.

Within ‘qualitative’ methods, there is more of a concern for the individual with behaviour having

meaning and being intentional and research methods use a variety of empirical methods, such as

personal experience, case study, observation, action research to describe meanings in people’s lives.

Pring (2004) states that qualitative methods

‘embrace symbolic interactionism, phenomenology, ethnography and hermeneutics. And within

any one piece of research there is frequently the employment of different approaches as different

questions are asked’.

(Pring 2004 p.48).

8

Within qualitative research there are also a number of different paradigms, including ethnography,

phenomenology, and critical theory. This suggests that there are many facets associated with this form

of social enquiry and there is no one form of investigation to fits all situations. For example, in some

paradigms, such as critical theory, there is a focus on action as a consequence of the behaviours, (i.e.

future orientated), whilst others such as ethnography look more at past situations. The very nature of

social situations will mean that there are differences in groups being studied, through the different

attributes and experiences of group members or environmental conditions so the results from one study

might not echo those of another. This suggests there are multiple realities, and ‘truth’ is defined as a

consensus amongst the informed (Pring, 2000) rather than absolute truth in positivistic, scientific

research. For this study, I will look to working within the interpretivist paradigm.

Interpretivism uses everyday practices and ordinary life as its subject matter. It seeks to make sense of

these through the understandings of those being researched. Whilst appropriate for these participants,

their contributions may be objectively faulty or misleading so generalisations are not a viable option.

This will be particularly true when asking the participants to recall events that have taken place

sometime previously and the surrounding situations are not able to be recalled accurately. The

implication is that no interpretation can be undeniably correct as this presupposes that there is an

authentic interpretation in dealing with people who have their own understandings and interpretations

of their, and others, actions. As work progresses the researcher may modify their own interpretations

and be able to compare and contrast those of others so that a degree of consensus is achievable. The

goal of this form of research is to provide an interpretation of the actions and practices within this

context. Linking to phenomenology and hermeneutics the interpretivist paradigm focuses on human

actions, with the assumption that these actions have meaning, can be interpreted and understood.

Interpretivism has also been associated with post modernism, which encourages the researcher to look

at what is meant by knowledge and truth and their verification. It questions the dominance of ‘one-view

world’, suggesting that there is a variety of viewpoints and that those with most power control the idea

of knowledge. Post modernism severs the link between knowledge and certainty, and questions how

they sit with different perspectives and whether the outcomes are consistent. Interpretivism takes

everyday practices and ordinary life as its subject matter, to be worked with and made sense of through

the understandings of the researched. These can be revealing but also objectively faulty or misleading so

making wide generalisations from the findings is not a viable option. According to Scott and Usher

(1999):

9

‘the goal of the research becomes that of providing interpretations of human actions and social

practices within the context of meaningful, culturally specific arrangements.’

Scott and Usher (1999) p. 26

An interpretivist approach has its roots in persons attaching meaning to themselves and others. In

understanding others, the researcher needs to understand their interpretations about what they are

doing, their intentions and their motives. Researchers would define this as subjective meanings of those

being researched (Pring 2004). In putting any interpretation on evidence where the researcher has been

involved (as in the case of case study, action research, observation, interviews) subjects the research to

the researcher’s own unique experiences. Also how participants interpreted the question and their

intentions in doing do (whether the responses were ‘true’ or ‘the expected’), and visual cues, such as

gestures, and how they might be interpreted will affect the data in some form or another. A

misinterpretation (and one that is acted upon) will change the social reality of the situation. The

phenomenological stance is about uncovering the subjective consciousness, the importance of the

interpretation of events or others’ actions in light of the subjective meanings through life experiences,

and their uniqueness, cannot be discounted. For those who follow the positive traditions the locating of

research in ‘traditions’, and hence subjectivity, is problematic as this affects, according to positivists, the

objectivity of the research and the independence of the researcher. Scott and Usher (1999) suggest that

understanding something is always prejudiced as there is an anticipation of meaning and this will have

influence in how the research is carried out, whatever the paradigm. Within any type of research there

is always the danger of removing the findings out of the context in which, and through which, they were

constructed and making generalisations, even when experimental control groups have been used and

results analysed quantitatively. In the medical world this is seen where there have been conflicting

generalisations, and subsequent advice given to people, regarding the use of drugs (such as aspirin), and

alcohol. Distortions of ‘truth’ may also arise where the sponsor of the research wishes to promote their

own agenda and ‘play down’ that which does not support their case. Pring (2004) suggests that:

‘failure to recognise complexity of enquiry and the nature of that which is being enquired into,

which causes the blurring of the distinctions with the so-called paradigms’

Pring (2004)

is a problem in that some writers elaborate on the distinctions between quantitative and qualitative

modes of research, rather than accepting they can be complementary in drawing informed conclusions

from the evidence obtained. Scott and Usher (1999) suggest that the positive tradition of locating

research in ‘traditions’ and hence subjectivity is a problem as this affects, according to them, the

objectivity of the research and the independence of the researcher. Scott and Usher point out:

10

knowledge cannot be without some subjectivism as the initial projection or understanding is

part of the subject’s situatedness – the subject’s location and standpoint’.

Scott and Usher (1999) p.25

Criticisms of qualitative methods (and by inference those in the non-statistical style), that come from

those who espouse a positivistic approach, include that it does not provide wide-ranging generalizations

or objective standards for verifying or rejecting accounts (Carr and Kemmis 1986). When carrying out

research into social situations the researcher should have an awareness of the criteria fundamental

concepts of reality and objectivity being presented, the case independently of the researcher’s own

beliefs and ideas. Krauss (2005) goes one-step further suggesting that many qualitative researchers

believe that the only way to understand a situation is to be part of it, and allow questions to develop

and change as the inquiry progresses. Within the naturalistic enquiry, the two prime tenets are that the

researcher does not affect the process of the enquiry nor its outcome. This form of enquiry looks deeper

into situations than one that uses quantitative methods and takes into account what has happened in

the past as well as in the present. Pring (2000) highlights the need to recognise the links between

knowledge, truth, certainty, and verification. Researchers need to have an awareness that participants’

behaviour can be moderated by previous experiences and this needs to be investigated and considered

in analysis. A definition of truth and fact rather than fiction is pertinent to the recording of observations

and emphasises the importance of ‘what-is’ rather than ‘what-is-thought-to-be’. There also needs to be

a distinction between common sense and practical understandings as opposed to the theoretical

position and knowledge against opinion. When conducting research, researchers need to be aware that

it will not be easy to make generalisations with a small number of people, unlike quantitative research,

where a good representative sample enables a degree of generalisation.

Another of the criticisms of qualitative research by those engaged in quantitative research in its purest

context is that analysis is subjective. Subjectivists reject the notion that reality is ‘out there’ and believe

that it is what the person or group makes it out to be. This is based on the premise that the standards of

attitudes and beliefs are at the level of the individual, or their community, and knowledge and truth are

created rather than being discovered. Pring (2004) discusses ‘common sense language’, contrasting this

with the technical language (i.e. precise and scientific). He suggests that the common sense account

determines what kind of explanation is relevant. This raises the query as to whether these attitudes and

beliefs are rational. One of the problems with positivism is that human behaviour cannot be defined as

can the natural world. Pring, (2004) uses the term ‘common sense’ as a range of unquestioned beliefs

11

that provide the basis for societal living and decision-making. These are not static, but change over time,

from one person to another and in light of experiences. This ‘common sense’ relates to attitudes and

beliefs and only by questioning and critical reflection can the move to a more theoretical stance be

made. In many respects, acknowledging that there is no one fixed answer demonstrates rationality, i.e.

recognition that there are differences and problems might need different solutions. An example of the

distinction between subjectivists and those who think there might be socially agreed ways of describing

and interpreting reality is frequently found in education. In recognising that not everyone learns in the

same way, for instance people with dyscalculia or dyslexia benefit from exposure to practical tasks

within mathematics lessons rather than learning from texts, and are hindered by written tests which

require short term memory recall, their ability to demonstrate their true aptitude is thus not recognised.

The ‘one-size-fits-all’ testing and subsequent data for these people is not a true reflection of their ability

and they are disadvantaged. Allowing for a more subjective form of data collection to be allowable

alongside the formal rigid test would provide a more rational picture and reduce the falsity of the

analysis of the test data with respect to how many pupils are truly functioning at any particular

academic level. Within this study, consideration of the participant’s situation will be born in mind,

alongside their aspirations, allowing them the opportunity to report what they would do if existing

constraints on their practice were not present.

An issue raised with qualitative research is one of validity and reliability. According to Winter (2000)

there is a problem in actually defining both ‘validity’ and ‘reliability, there being no one definition for the

terms. Hammersley (1987) suggests that if the account accurately represents the features of the

phenomena then it is valid or true, however it is relative to the person and the belief systems involved.

Winter, in summarising the definitions of ‘validity’ and ‘reliability’, identifies accuracy of measurement

and whether the intended measurements are being made for ‘validity’, and whether it is possible to

replicate the process with respect to ‘reliability’. Cohen and Manion (2007) suggest that reality is

‘multilayered and complex’. As qualitative research involves non-static concepts of people’s lives and

experiences there will be differences between individuals and groups. In data collection, an effort can be

made to reduce contradictions between information collected and description by presenting the

information to the participants for checking and reporting ‘as seen or heard’. In adding an interpretation

the researcher, intentionally or unintentionally, introduces discrepancies through their selection and

choice on what and how to report. The researcher needs to be aware of introducing subjectivity into

their report that adds to these discrepancies. Maxwell (1992) suggests that participants may not be able

to interpret their actions with any more validity than other people can. At the interpretation stage,

12

there may be ethical problems requiring careful handling whether from the participant’s viewpoint or

because of agencies and stakeholders involved in the research. In using purposeful sampling (as

opposed to random) interpretations of the findings are not widely generalisable and a similar study

might produce different results. According to Winter (2000) the terms ‘validity’ and ‘reliability’ are more

appropriate to research orientated to quantifiable data and where the researcher wishes to justify their

results as the only possible one, i.e. generalisations can be made giving an external validity. Winter

suggests that if the research is measuring what it is intended to measure, then it is to be considered

reliable.

The development of a theory as a framework for research enables concepts to be drawn together to

help identify missing ideas, links and data for exploration. From this, two groups of researchers emerge

the rationalists who are concerned with theory development (Travers, (1978), cited in Verma) and

empiricists who focus on data or fact collection, although most social science research does have an

element of both. Educational research has to deal with a multiplicity of variables, only some of which

can be controlled, including the effect of the interaction with subjects and the effect of them being

involved in the research. This will include subjects changing their beliefs and behaviours from thinking

about the questions posed by the research rather than any direct influence by the researcher. This effect

will vary from one participant to another. I will be conducting this study from a predominantly rationalist

standpoint and suspect that in posing some of the questions relating to their personal practice the

participants will question what they are currently doing and might/could change what they do in the

future. Motivation (in the case of this study teachers using ICT in their lessons) is not measurable, it is a

theoretical construct and hence the study will lean towards qualitative analysis methods rather than

statistical analysis. At the current time there is a lack of tools for measuring these constructs, and as

some of these are dealing with quite complex issues, e.g. attitudes, it is difficult to make wide

generalisations as attitudes and beliefs will differ from one individual to another, and these will not be

consistent as their situations change over time. As there are constraints placed on teachers’ practice

from both internal and external situations, I will also enquire what they would do in an ideal situation to

enable them to express ‘their ideal world’.

This study follows an interpretive route where the researcher needs to be more in contact with the

participants of the investigation, as it seeks to find out what it is that encourages some teachers of

mathematics to use ICT in spite of internal and external constraints that might exist in their particular

situation. By being more involved, this raises an issue in that the researcher’s own values, attitudes and

13

perceptions are present to a greater of lesser extent. When undertaking interpretive research the

researcher seeks to be more interconnected, trustworthy and less detached from the situation. This

requires steps to be taken to reduce the impact of the researcher’s involvement to avoid the research

being subjective rather than objective. In the early stages of the research the issues surrounding the

hypothesis (in this case the use or non-use) might appear to be quite apparent, but reading around the

subject and/or conducting a pilot survey help to make the hypothesis and research question more

focussed and specific. Gadamer (1975) suggests that the ‘subject’ and the ‘object’ of the research are

situated in known worlds hence it is impossible for a researcher to disassociate themselves completely

from the situation. These pre-understandings will formulate the framework for the investigation. The

implication is that no interpretation can be undeniably correct as this presupposes that there is one

authentic interpretation. In dealing with people who bring their own understandings and interpretations

the researcher is already dealing with interpretation so the process of interpretation is more cyclical or a

spiral than the linear model of positivist traditions where interpretation can be objective. Through

interactions with the subjects, the researcher may modify their own understandings as the work

proceeds, leading to re-interpretation and different meanings being place on them. This also enables

different and conflicting interpretations to be compared and contrasted leading to some consensus

between the parties involved and with similar research. The initial proposal for this study was to

establish the constraints teachers experienced and why some mathematics teachers use ICT while

others either do not, or use very little. From the pilot and subsequent questionnaires, it became

apparent that some teachers were overcoming difficulties and believed that using ICT was beneficial to

teaching and learning.

In designing an inquiry and working within the qualitative paradigm certain issues need to be kept to the

fore. A key issue in any research is the validity of the research, especially when the researcher ‘becomes

immersed’ in the situation as their personal or socially constructed ideas may have influence on both

the situation and the reporting of the investigation. The recording of ‘what is’ rather than fiction or an

interpretation of what has been recorded has to be firmly established from the outset with the

researcher being aware of the responses angled to what the ‘researcher wants to hear’ as well as

recording situations without bias or interpretation. An attempt at verification of the credibility or

replicability needs to be made whilst conducting the research, this may be carried out by negotiation

with the parties involved to ensure that they reported facts are, indeed, true. As time passes additional

work might add to rather than conflict with the original research, resulting in a more informed and

sophisticated piece of work. In the case of this study I moved from the initial data gathering by

14

questionnaires into what I call the ‘second phase’, to find out more about some teachers who regularly

use ICT in their lessons and whether there are any similarities in experiences or themselves as teachers

that contribute towards this.

Methodologies

Having elected to take an ontological position within the perspective of constructivism, i.e. one that:

‘asserts that social phenomena and their meanings are continually being accomplished by social

actors, It implies that social phenomena are not only produced through social interaction but they

are in a constant state of revision’

(Bryman, 2001, pp16-18)

and an ‘interpretivist’ epistemological position that:

‘is predicated upon the view that a strategy is required that respects the differences between

people and the objects of natural sciences and therefore requires the social scientist to grasp the

subjective meaning of social action’

(Bryman, 2001, pp12-13)

The methodology chosen is one of narrative enquiry, seeking to understand why some teachers of

mathematics will use technology in their lessons whilst others are more reluctant to do so and whether

this is something that relates to their personality or their past experiences of training or working in

schools. In using a narrative approach several ethical issues are raised, namely confidentiality, access,

informed consent and negotiation. Within these, there are disadvantages of using this type of research

and certain safeguards need to be put in place. In dealing with data analysis, processes and transcribing

materials into a more formal academic form there are issues of relationships and accountability. Doucet

and Mauthner, (2003) point out that maintaining relationships is important, and there is a need to

include those who do not appear to fit into the researchers’ frameworks and analytical concepts. On the

periphery there are those who are associated with the group and research, but not actually participating

who should also be included. I have undertaken a group interview at one school to allow for the fact

that not all teachers in a department will make the same use of given and available resources. The

second point they make is one of accountability. They suggest that reflexivity is concerned with holding

the ontology, epistemology and methodology together and that the researcher must decide how much

they should let the participants know. When data is analysed there needs to be an awareness of

possible different agendas regarding the outcome from the stakeholders as it could be exposing power

and privilege. Through negotiation with participants there is a reduction in the risk that the

interpretation is of ‘what is’ rather than ‘what is thought to be’ and hence a check is made on the

15

validity of the research. In the case of this research, there are no stakeholders and the participants

selected for interview are, apart from the group interview, from different establishments and are

unlikely to know each other’s work.

The study includes elements of phenomenology in that deals with everyday life and peopl’s experiences.

Its basis is narrative, and deals with perceptions, attitudes, beliefs and feelings (Denscombe, 2003, p.96).

Denscombe also illustrates phenomenological approach as dealing with issues in depth so providing a

detailed description of the ‘experience’ with the interpretation resting with the participant and not the

researcher, so minimising the reliance on the researchers own beliefs and attitudes. The analysis is of

the ‘story’ of the participants rather than an attempt to generalise form a small sample to the general

population or to form policy. The study will also incorporate elements of grounded theory as originated

from work of Glaser and Strauss (1967). Grounded theory has, however, been adapted by researchers to

meet their own purposes. The original concept was that fieldwork should be carried out before the

research of literature to minimise the risk of influencing the outcomes by the researcher and their

agenda. The theory then developed from this point. The initial pilot survey on teacher’s usage was

carried out before reviewing literature and the results analysed. This lead to the development of the

second pilot questionnaire which investigated certain aspects in more detail to investigate questions

that might be worth pursuing in more detail. In turn, these led to the development of the semi-

structured interview questions. This fits with the idea of grounded theory as there were no fixed ideas

about who or what would be included before the data was analysed and paths were chosen based on

these outcomes. Grounded theory emphasises practical needs and is exploratory research on human

interaction rather than proof and policy making. Strauss and Corbin describe grounded theory as:

‘The initial questions or area for observation are based on concepts derived from literature or

experience. Since these concepts .... do not yet have proven theoretical relevance to the

evolving theory, they must be considered provisional. Nevertheless they provide a beginning

focus, a place for the researcher to start.’

Strauss and Corbin, (1990) p.180

It might be said that data collection in grounded theory is cumulative in that it builds on that previously

collected, while becoming more focussed with fewer categories being explored.

16

1st pilot questionnaire for teachers

2nd pilot questionnaire for teachers, questionnaire for students

Training provider interviews

Teacher interviews

Using different methods will result in triangulation and support validity.

Denscombe (2003) suggests methods for data collection include questionnaires, interviews, documents

and observation to provide a breadth of study and a focus on a snapshot in time. In deciding methods to

use, there was the issue of accessibility to participants. He suggests that surveys are a research strategy

rather than a method and that they are wide and inclusive and as such, capturing a moment in time,

suitable for empirical research with a view to generalising, the sample being carefully selected to be

representative. Verma and Mallick, (1999) describe surveys as ‘planned data collection for explaining or

answering questions as a guide to action’. Data collection methods do include questionnaires, interviews

and standardised tests with some researchers choosing to verify and collect additional data through

using the option of a follow-up interview. Documentary evidence is seen an additional source. In

surveys, the researcher stays at a distance from the participants and is more concerned with wider

issues than with individual characteristics. The sample size needs to be large to enable generalisations to

be made, and broad or narrow in scope e.g. a census. Apart from the issue of gaining access to a large

number of mathematics teachers, this would not have given the more personal approach and the

opportunity to gain extra, more personal, information about the person and their beliefs. The data only

gives information on the facts being investigated, not on the causes or reasons why and is situated at

one point in time, the information will also include recollections on ‘what has been’ rather than ‘what is

happening now’ so introducing a degree of uncertainty as it relies on participants’ memory. Surveys are

17

able to reach more people than observation alone, however the participants are able to put their own

interpretation on questions, also there are problems of ambiguity in the chosen language, especially by

non-native speakers. Many are longitudinal studies; thus, they have the problem that they are

potentially expensive to carry out. In addition, there is the problem of remaining in contact with

participants over time.

In my study it has not been possible to gain access to a large, random sample of teachers and, because

of the nature of the enquiry, surveys are not an appropriate method as more is learnt from personal

contact than the participants filling in questionnaires, whether these be postal or via the internet.

Initially I had hoped to select participants from one county, but conditions and staff changed and this

was no longer a possibility. The decision was taken to use purposeful sampling of mathematics teachers

who I knew, or had recommended to me as those who were using ICT rather than probability sampling.

Non-probability sampling is appropriate when it is not feasible to include a sufficiently large number of

examples, it is difficult to contact a probability-selected sample (Denscombe, 2003). According to

Denscombe, purposeful sampling implies that the sample is ‘hand-picked’ as they will provide the most

valuable data. Snowballing, where references are passed from participants to the researcher, enables

the sample to grow purposefully. Some of my participants have been found in this way. This method

also gives a contact point with prospective participants who have been nominated against certain

criteria, giving a purposeful sample. The participants of my survey have all shown an interest in

developing their skills in mathematics education through voluntary attendance on courses run by local

education authorities or national organisations such as Association of Teachers of Mathematics and

National Centre for Excellence in Teaching Mathematics. Fitting in with the interpretative style of

methodology the sample sizes are small put in final figures here and so will not lead to generalisations to

be applied to larger groups.

Use has been made of questionnaires as they provide data economically and are able to be phrased to

produce the information sought with as little ambiguity as to their meaning as possible. Good

questionnaires should be devised so that they cover all of the study, but depending on where and how

they are to be administered be time-efficient for the subjects; too long and the participants lose interest

and questions are not carefully completing leading to a possible bias in the data. Questionnaire

questions may be ‘closed’ or ‘open’ in nature, the former being easier to analyse as the answers are

more controlled. The 3, 5, or 7 point Likart Scale is frequently used, allowing answers to be marked from

strongly agree to strongly disagree (or similar scale), gauging the participant’s view on matters. In some

18

questionnaires, an even number of scale items is used to ‘force’ the response to be one end of the

spectrum. Open–ended questions can be more probing and illuminating, but responses take longer to

code. Responses can be limited by requesting for a set number of factors or reasons. It is useful to have

a small pre-test of between twelve and twenty people to check how they respond to the questions and

allow them space to contribute feedback. I chose to do a pilot questionnaire of teachers who attending

a mathematics conference and who volunteered to fill in the questionnaire. As Denscombe, (2003)

points out the initial research does not need to be representative as relevance is its prime consideration.

Following the pre-questionnaire (or pilot) the questionnaire for the main sample may need revision. My

pilot questionnaire did show that a few questions needed re-working and feedback was that some

participants would have liked the opportunity to expand their answers. This was made possible on the

second version but also gave the appearance of being lengthier as they thought that all boxes needed

completing. Some of the questions can be analysed statistically but as the samples are small,

generalisations cannot be made, except for those samples. Further research would be needed on those

specific criteria to validate the results with regards to a wider population. Questionnaires may be

distributed for independent completion or the participants being assisted by the researcher, particularly

if some questions need more explanation. Internet (or postal) questionnaires are an option, but open

response rate is poor (Denscombe, (2003) quotes less than 20%). This form of response does not appeal

to everyone, presenting a problem of bias. This is frequently done by asking for a participants profile and

adjusting the results according to the profile that is built up. A group of mathematics undergraduates at

Warwick University completed an online questionnaire for this study, enabling me to obtain an insight

into their perspective on the use of ICT in mathematics lessons when they were at school. A face-to-face

request to complete a questionnaire can be richer and the interviewer can get a sense of validity of the

answers by deeper questioning. It is also possible to get a balance of participants by selecting criteria

such as age and gender (Denscombe, 2003). Another alternative is to use telephone methods for

completing questionnaires and interviews. Thomas and Purdon (1995:4) cited in Denscombe (2003 p.9)

suggest that regarding reliability there is no suggestion that these methods are less valid than other

methods. The teacher’s questionnaire was available on line, some participants did take a paper version

one to complete later, but the original pilot and many of the secondary questionnaires were completed

in my presence.

Interviews are used to supplement and complement the data collected through questionnaires. This will

involve the same or different participants. Powney and Watts (1984) describe interviews as:

19

‘a conversation between two or more people where one or more of the participants takes the

responsibility for the reporting of the substance of what is said.’

Powney and Watts (1984)

Interviews, used alongside questionnaires, are particularly useful when the survey provides quantitative

data enabling the development of responses or the addition of further information. Unlike

questionnaires, the interviewer needs to meet the participants, usually at a venue and time of their

choosing; this makes them more expensive in time, at the outset, and in transcription of a recorded

interview. This has the effect of selecting a smaller sample than perhaps might be used in questionnaires

and, as in the case of this study, they might well be purposefully sampled to expand upon data already

collected.

The role of the interviewer is to ask questions, record answers and control the interview. Interviews

range from closely structured (more akin to a questionnaire) to open-ended where the participant is

encouraged to explore issues in depth. Between the two extremes are semi-structured interviews where

researcher has a plan, possibly from an already administered questionnaire, to explore questions in

greater depth and detail. Unlike open-ended interviews these, and the closed interview questions, can

be administered to different participants by different researchers as there are key facts that are being

sought. In conducting interviews, I have used semi-structured interviews in that the questions have been

prepared so that the participant is able to expand outside the question if they wish. In some cases, this

gives new avenues to explore. Having the set questions gives a structure to the interview, so that it

allowing it to be conducted in an orderly way to avoid missing ideas that I would like to explore. In this

type of interview, the researcher has to be strict in how the data is interpreted so as to avoid their own

values, perceptions and prejudices corrupting the analysis. The participants are sent a copy for checking;

correcting and adding comments should they wish in case I have misinterpreted their intention through

my own beliefs and perceptions.

Observations, like interviews can be structured or un-structured and the observer/researcher may

either act as if on the outside of the situation or become immersed within it. However, both methods

present difficulties, not least because the very presence of the observer/researcher might influence the

behaviour of the participants. There are also ethical issues, particularly if the observer/researcher is

reporting back to a more senior member of staff. There is also the question of whether the observed

knows the full details of the observation, presenting a dilemma to the researcher and those permitting

20

the research to take place. I may well include a couple if the teachers are willing to go with the

interviews, they will know the purpose.

One of the other tools available to researchers is documentary and content analysis. This is frequently

used with historical methods, used in documentary enquiry and comparative studies, employing a

systematic collection of, and an objective evaluation of, data relating to past occurrences to look at the

causes, effects or trends presented by the current problem so seeking an insight to the present situation

and predict what might happen in the future. In historical methods, the data is collected, organised,

verified, validated and analysed in accordance with a set of standards, enabling the forming of some

generalisations and predictions. Historical methods data collection includes both primary and secondary

sources. This is not confined to historical methods as this type of analysis is found, most commonly, in

descriptive studies and consists of counting concepts, words or ideas in a document. The purpose of the

exercise is in describing the relative frequency and importance of certain themes, an evaluation of bias,

propaganda or prejudice, as well as assessing the difficulty of reading materials or analysing error types

in students’ work. Sample sizes can be large or small. The collected data is subjected to statistical

analysis before interpretation (Verma and Mallick, 1999). Although I am including a timeline within my

study, it is not relevant to conduct a statistical analysis as it is providing a background to the

development of ICT in schools, particularly in relation to mathematics teachers.

According to Baikie, (1973), research strategies can be assigned to four groups, induction, deduction,

retroduction and abduction. Induction was at the heart of grounded theory (Scott and Usher, 1999) and

supports theory based on data rather than the testing of hypotheses. As grounded theory developed it

became more in line with the realities of quantitative research, (Strauss and Corbin 1990), and social

phenomena. As Hammond and Wellington (2012) point out, an inductive approach offers more

flexibility for the researcher. In reality, a researcher’s own background precludes them from coming into

a situation with no beliefs or prejudices and this will have some effect on how the data is collected and

what questions are asked. My own position is that I was head of mathematics and at the same time in

charge of ICT within two schools where I taught. Believing in leading by example my own position was

that I used ICT where appropriate and expected my colleagues to do likewise, although at a possibly

reduced level. This precludes me from entering the situation with no beliefs or prejudices on the use of

ICT, but being aware of the different attitudes and commitment of colleagues suggested that there are

certain people who embrace the use of technology in the classroom and others who do not when all

other factors such as resource provision are minimised.

21

Deduction is more associated with positivistic research, looking at identifying discrete variables within a

hypotheses and using empirical data to confirm or reject the hypotheses but, as suggested by Schmutter

and Schmitt (2001) it also has a place alongside induction as all research will begin with a theory, albeit

possibly undefined at the commencement of the research. When beginning my study I had the idea that

the training of teachers in using technology in the classroom, at initial training stage or as part of in-

service training, would have some impact and included questions to this effect in the questionnaires.

Using and if .. then statement such as ‘if teachers are given training then they will be able to use ICT in

their lessons’, which may, or may not, be true at the individual level. The problem with deduction is that

it is too theory-led and rejects theories that cannot be, potentially, falsified and results can be

misleading. This is not a method that is appropriate to my research.

Retroduction and abduction have their roots in the early twentieth century and were put forward by

Pierce who used them interchangeably as the method by which hypotheses are engendered. Chiasson (2001)

(http://www.digitalpeirce.fee.unicamp.br/p-abachi.htm states that abduction is still really not understood.

Schmutter and Schmitt (2001) argue that retroduction and abduction are ‘artefacts constructed by

authors who refuse to accept the existence of a deductive- inductive dialectic’ and are a means to

minimise the weaknesses of the induction – deduction strategies. Retroduction applies to research,

according to Scott and Usher, (1999), concerning regularities in observable phenomena, which are not

immediately available to consciousness, and developing a possible model to explain the phenomena

causally. Chiasson (2001) suggests that it is about going back to investigate some surprising fact before

making a hypothesis worthy of testing. Following successful testing of the model the researcher would

have grounds for believing in the existence of these structures and mechanisms. If, in my research, there

was the ability to be able to follow some of the constraints identified by participants through after giving

remedial action, e.g. software training and support, this would be an appropriate consideration but this

has been rejected in line with my focus. Interpretive researchers will use abduction to draw out

meanings used by social actors in their daily lives and unlike the recursiveness of retroduction, it moves

forward, looking for explanations of other possibilities identified in the study.

To summarise, this study will follow an interpretative approach, the theory emerging and arising from

situations as they develop, the data including meanings and purposes of those people who are their

source (Cohen and Manion, 2007). The study is small scale and looks at human actions in a non-

statistical manner but with some ‘subjectivity’ as I, as the researcher, am involved in the situation. The

22

data is to be collected using a mixed methods, i.e. questionnaires and interviews. The participants are all

volunteers chosen using purposeful sampling. The teachers taking part in the questionnaire have been

selected in that they are currently teaching, chosen to attend mathematics specific courses. Those

chosen for interview are known to use ICT in their lessons and are mathematics specialists in secondary

schools. Other groups have been chosen for their interest in mathematics education (the student

groups) or involvement in teacher training (advisors or teacher educators). The data derived from the

interviews will be fed back to the participants for any comments to check for validity. The

questionnaires have allowed the research to become more focussed and they could be regarded as

being pilots and ‘setting the general scene’, allowing me to decide which fields should be further

developed.

23