You are on page 1of 26

Masters Level Module – Research Methods

Chapter 6

Overview of Research Methodologies

Learning Outcomes

After studying this chapter, you should be able to:

1. Describe the paradigms which underpin the main research


methodologies.

2. Discuss the range of methodological techniques available to the


researcher.

3. Explain how to match methodological techniques to data


requirements.

4. Justify your selection of a methodological technique.

5. Show how to apply a methodology robustly and rigorously.

6. Explain the issues surrounding the mixing of research methodologies.

6-1
Masters Level Module – Research Methods

6.1 Introduction
The term research methodology is used in two inter-related ways which
are often not very clearly separated. Methodology is frequently used as an
overarching term to refer to the principal paradigms of an approach to
researching a problem, for instance the researcher takes a qualitative
approach, or a quantitative approach. The distinction between these two
approaches is explored in the first part of the chapter. Strictly speaking,
this is defining the methodological strategy according to the data form.
Assuming that a universal research target is the robust treatment of the
collection and the interpretation of the data, this distinction between
qualitative and quantitative approaches relates to differences in the
methodological techniques which are used, and the way in which they are
applied. The term methodology is also used as a definition of an
operational research technique, hence the researcher may use The Delphi
Approach, or undertake a Questionnaire-Based Study, use a Case Study
Technique, or conduct Semi-Structured Interviews. The techniques vary
but tend to accord with the clearly definable methodological paradigms
of qualitative or quantitative research.

It is quite possible to conduct a satisfactory piece of research without


being expert in all methodological techniques, so long as you are expert
in the particular technique that you are using. Of course the task of
selecting the technique which is best suited to the problem requires some
knowledge of all the possible techniques. If you intend to mix
methodologies then somewhat more detailed knowledge of their
principles and operation are required. The remainder of this chapter is
designed to provide an introduction to the range of research methods
used for gathering data.

6.2 Distinctions Between Qualitative And


Quantitative Approaches To Research
Looking first at the underpinning paradigmatic distinctions between the
methodological schools, the two broadly differing approaches to
describing the collection and handling of data are usually referred to as
the qualitative approach and the quantitative approach. They depend
upon qualitative paradigms or quantitative paradigms respectively.

6-2
Masters Level Module – Research Methods

The quantitative approach, also known as the scientific method, is


founded on the assertion that there is a single reality which is objective.
That is, if you are to research an issue, it is possible and necessary to
separate the phenomenon from the surrounding environment and make a
freestanding assessment of this objective reality. The quantitative
approach asserts that it is both possible and necessary to maintain
distance or objectivity from the subject of the research. The approach is
characterised by the image of the disinterested researcher, who observes
without inter-relating with what is observed. This is also known as the
positivistic ideal. Quantitative approaches are therefore assumed to be
repeatable, and capable of isolation from reality without compromising
the cause and effect being researched. The paradigm assumes that the
researcher will be independent from that which is being observed.
Conceptualisation is usually by deduction from the evidence. It is
therefore axiomatic that the phenomena under study can in some way be
measured using a numeric or quantitative approach. As a consequence,
these techniques tend to be better for theory testing than theory building.

The qualitative approach to research is based on the assumption that


there is no singular objective reality, and that the nature of an observed
reality is related in some way to the researcher’s interaction with the
phenomenon. The conventional defence of the qualitative approach
proceeds to assert that the quantitative paradigm of creating a sterile
objectivity is a fallacy. Instead, real-world phenomena need to be
assessed from within the context of that reality. Qualitative research
forces a (realistic) complexity on the study, since the isolating
assumptions usually imposed to create the necessary control for a
quantitative experiment cannot be made. As a consequence, the
qualitative research tends to yield rich, complex data which does not
render itself to isolation of singular realities or generalisable and
quantified causes and effects. The findings illuminate the qualities of
phenomena rather than their numeric measurement. The richness of the
findings will usually provide contextual location for the research
question.

The qualitative approach is distinguished categorically from the scientific


method by its inductive nature. It is more probable that unanticipated
findings will emerge, and that the evidence will illuminate the process

6-3
Masters Level Module – Research Methods

involved in the research phenomenon and the causal explanations of its


dynamic. Hence these techniques tend to be better for theory building
than theory testing and it is common for there to be a proposition
associated with the research question rather than a hypothesis.

One of the most frequently-voiced criticisms of the qualitative approach


is that the views of the researcher colour the findings produced by
qualitative techniques. As such, the findings are not considered to have
objectivity in the style of the quantitative independence. Modern
qualitative school counter-arguments to this have emerged with the
ascendance of action research. This is a particularly invasive technique,
since the researcher deliberately catalyses a change and learns from the
observed changes and dynamics that this change unleashes. A concept of
‘sinning bravely’ has arisen, where researchers acknowledge
involvement in the real world scenario being observed. Rather than trying
to remain disinterested or independent, researchers acknowledge their
role and try to understand qualitatively the nature of their contribution to
the phenomenon being observed. The tensions this causes for the
advocate of the scientific method are not difficult to see but there are
clear advantages for learning which such cases can illuminate. The key
challenge for the action researcher is to produce findings which can be
generalisable to other circumstances irrespective of their involvement in
the particular case under study. It is here that the defence of the rigour of
the research can be particularly challenged under the scientific spotlight.

PF6.1

Since the fundamental principles of quantitative and qualitative


approaches differ so much, do you think either can be truly appropriate
for researching an issue?

PF6.2

Can ‘fly-on-the-wall’ documentaries be objective?

6.2.1 The Experimental Approach


This is the classical quantitative approach to research, the operation of
which is conventionally based on the principle of systematically isolating
variables in order to manipulate them carefully and assess specified

6-4
Masters Level Module – Research Methods

causes and effects. With this research approach, the research question
and hypothesis usually refer expressly to identifying and testing specific
causes and effects.

6.2.2 Requirements
Experimentation requires the manipulation of independent variables,
followed by the measurement of the effect of this on the dependent
variables. An independent variable is a variable which is assumed to
come first in a relationship between two variables. The dependent
variable is assumed to follow the independent variable, and therefore
depend upon it. Hence changes in the independent variable cause changes
in the related, or dependent, variable. This approach requires a scenario
where variables can be identified and distinguished as dependent or
independent. Also the variables in question must be capable of
manipulation, which requires an understanding (or conceptual theory) of
how the experimental focus relates to external factors. This implies a
tightly-defined boundary to the research, and a relatively high degree of
advanced conceptual theory. Quantitative approaches to experimentation
tend to require a high degree of detailed conceptual underpinning.
Therefore the classical experimental approach tends to be used for theory
testing rather than theory building. Also as a consequence of this,
experimentation does not tend to be well suited to complex systems.

Any problems with the degree of isolation or the control experiment can
produce profound difficulties for the experimental process. It is essential
that the experimenter has complete control over the independent
variables and that there are no uncontrolled or ‘invisible’ forces acting on
the experiment. Accordingly, it is usual for the experiment to be reported
in sufficient detail to allow its replication by another researcher to verify
the phenomenon or results. Hence a requirement of the analysis of
experimentally-derived data (and the scientific method in general) is that
the experiment-as-reported can be repeated. This condition alone requires
an abstraction of the experiment from real world pressures, since any
circumstantial specifics which are unrepeatable or unverifiable will
seriously effect the experimental validity.

6-5
Masters Level Module – Research Methods

6.2.3 Operation
The independent variables selected for experimentation are isolated and
then manipulated in some controlled manner to produce an identifiable
and measurable response. The hypothesis usually defines the expected
identifiable and measurable result, and therefore the experiment tests the
hypothesis. Accordingly, the research question and conceptual model
will detail the assumptions and concepts upon which the experiment is
based, and the conditions which are to be created and manipulated to
create the causal effect on the experimental scenario.

It is normal to establish a control, that is a duplicate scenario to which the


systematic change is deliberately not performed. By comparing the effect
of the deliberate cause in the experiment with the expected and
confirmed lack of change in the undisturbed control case, experimenters
increase their understanding and confidence in the link between created
cause and observed effect. This also provides support for the
phenomenon when it is reported.

6.2.4 Strengths And Weaknesses


The advocates of the experimental approach note that the technique is
ideal for examining demonstrable cause and effect, thereby allowing
extensions of theory to be empirically supported. The opponents of the
experimental technique counter by suggesting that in many cases the
degree of limiting assumptions required to establish a focused experiment
(with isolated independent variables which the researcher can
manipulate) involves making assumptions which render the scenario (and
findings) unrealistic. In such cases, the lack of realism in the
experimental conditions may invalidate the interpretation of the findings
within realistic environments. The corresponding applicability of the
findings may be severely compromised.

The arguments against experimentation generally relate to the lack of


demonstrable realism, in particular the difficulty in establishing that the
limiting assumptions required to make the experiment controllable are
not fatal to its representativeness of any meaningful situation. It is also

6-6
Masters Level Module – Research Methods

essential that a representative data sample is used, since the approach


generally relies on taking a controlled manipulation of a generalisable
phenomenon, and then generalising the results. Without a representative
sample of data, it will not be possible to safely generalise the outputs.
Hence data sampling is particularly sensitive for experimentation. Some
experiments may be extremely expensive or require such a large
collection of data that the resource implications become prohibitive.

6.2.5 Uses
Properly applied, experimentation can establish causality links and their
direction. The relationship between independent and dependent variables
can be clarified in terms of their directionality. Where it is possible to
adjust variables and parameters in a controlled and measured manner,
experiment can produce very useful sets of data which allow the
extension of findings beyond the simple demonstration of cause and
effect, onto calibration. Experiment can be a very effective method of
testing the principles of a technique in a controlled and cost effective
manner, and may be used successfully as a piloting or feasibility tool.

The experimental approach can also be used in an analogous or


metaphorical manner for modelling real world phenomena to test or
validate basic abstract theory before costly real applications. For
instance, modelling the movement of people in buildings has traditionally
been supported by the case study analysis of them whilst really moving.
By triangulating (that is comparing) with experimental analogies using
ball bearings or flowing water, it has been possible to demonstrate
queuing applications theory for people movement through physically
controlled spaces. This cannot be extended directly to real life use of
buildings, but it can extend conceptual theory to produce a better
hypothetical framework for further real world case study analysis. In
such cases experimentation can be used as one of several methodologies.
As described here it is used serially, or reiteratively with case study
techniques. In other applications experimentation may be used in parallel,
in a true mixed-methodology application.

6-7
Masters Level Module – Research Methods

6.2.6 Summary Points Of The Experimental Approach


A deliberate intervention to alter independent variables and assess the
outcome(s).

Controlled, selective, and follows the researcher’s model of reality (and


therefore their selection of issues to study).

Must alter single variables − the best variables are exogenous.

Requires a robust/ prescribed model.

Contrived circumstances are possible.

PF6.3

Outline the minimum requirements for a successful experiment, and the


risks to the successful operation of an experimental process.

PF6.4

Identify some common applications of experimentation in construction.


What is it about these applications that renders them open to
experimental research?

6.2.7 The Quasi-Experimental Approach


True experimentation is limited to controllable situations. Real-world
phenomena which occur at unexpected times or in situations where the
cause and effect is complex and/or obscure will not yield to the full
requirements of the experimental approach. In such cases the quasi-
experiment may be a suitable technique.

6.2.8 Requirements
The principal distinctions between true experiments and quasi-
experiments lie in their requirements. The quasi-experiment
approximates to a true experimental approach in the sense that
explanations for cause and effect are sought. However, most quasi-

6-8
Masters Level Module – Research Methods

experiments operate with data that are not random. This is because there
is a lack of complete control over the experimental data design which the
researcher cannot solve. The quasi-experiment is therefore a pragmatic
solution to experimenting without random data. Also, there may be no
intervention of the researcher.

6.2.9 Operation
The quasi-experiment operates as an experimental mode of analysis
rather than within the controlled circumstances of the true experiment. It
may be that the researcher has no control over the independent variables,
or for some reason it is invalid to control them (because the change has to
be a natural phenomenon to be realistic, perhaps). With the quasi-
experiment technique, the researcher uses the opportunity afforded by an
accidental or chance change in the independent variable which stimulates
the phenomenon that is then observed.

It may still be possible to be selective over the independent variables to


observe, even if they cannot be controlled. In such cases, the quasi
experiment may involve measurement of the dependent variable in the
quasi-experimental group and comparison of the results with a control
group if this can be achieved. This approach can suffer from problems
with the prior assignation of subjects to the control and quasi-
experimental data groups. It may be that the splitting of data groups has
to be done after the quasi-experiment has occurred, on the basis of
distinguishing between those subjects who were affected by the incidence
of the independent variable changing, and those who were not.

6.2.10 Strengths And Weaknesses


There is of course a fundamental loss (or release) of control over
extraneous variables with the quasi-experimental technique. The
conventional true experiment pre-occupation with replication may remain
a possibility depending on the circumstances of the research, but because
the research method relies on an unplanned or uncontrollable change in
the status of the independent variable, replication is not typically a
strength of quasi-experimental research. However, the randomness and
6-9
Masters Level Module – Research Methods

lack of artificial control of the quasi-experimental technique can give a


high degree of population validity by avoiding some of the conventional
problems of artificiality associated with true experiments.

6.2.11 Use
The quasi-experiment is extremely useful for illuminating potential
relationships between variables which are involved in complex
phenomenological systems. It is also more potentially applicable to real-
world research problems. In the right circumstances, it can be used to
assist in creating conceptual models and is consequently more useful for
theory building than the true experiment. The researcher uses logic to
analyse the data provided by the quasi-experiment to make deductions
about the phenomenon. Natural, uncontrollable scenarios may yield to
quasi-experimental research where the control requirements for true
experiment are not achievable (or ethical).

6.2.12 Summary Points For The Quasi-Experimental


Approach
It may be invalid or impossible to control the independent variables, but
some accidental manipulation or other change may occur which the
researcher can use to their advantage. A pragmatic solution to
experimenting without random data.

Selective, but may not be controlled or deliberate. Researcher remains


external to the problem. The research may be observational or facilitating
rather than controlling.

Uses a less robust/ prescribed model than true experiments. Does not
meet the full criteria for rigorous true experiment research.

May be used to create conceptual models. Use logic and reasoning to


analyse the data rather than by extra control of variables (which may also
be possible). Tends to be limited to causal explanations, but can be
extremely useful in uncovering potential relationships between variables.

Good for social science approaches to researching problems.

6-10
Masters Level Module – Research Methods

PF6.5

Compare and contrast the conceptual assumptions and operation of true


experiments and quasi experiments.

6.3 Questionnaires And Interviews


Surveys are a popular approach to collecting data for theory testing. For
such applications, the researcher is frequently using a prescribed
conceptual frame of reference, and the collection of the data is designed
to populate that conceptual framework. There is a range of possible
approaches which includes questionnaires, interviews, the examination of
documentary evidence, observational studies, the review of diaries, and
schedules. Questionnaires and Interviews are very common and will be
discussed here. The other techniques are less widespread, and suggested
further reading is included at the end of this chapter.

6.3.1 Requirements
One of the key requirements for a successful interview or questionnaire-
based data collection exercise is the sample size and sample
representativeness.

To deal with the natural and methodologically-induced variability in


responses (scatter), it is important to select sufficient questionnaire
respondents or interviewees to allow the researcher to analyse the results
confident in the assumed representiveness and confidence in the results.
Targeting the sample population can usually only be done to a certain
extent prior to circulating a questionnaire or blind interview. In such
cases, it is very worthwhile to give some thought to including filtering
and categorisation questions in the early sections of the questionnaire/
interview to allow the data analysis to be sub-divided or pruned
according to the priorities of the research.

Questions have to be screened carefully to eliminate leading or


ambiguous content or style. Use a pilot study with de-briefing of
respondents and/ or a negative brainstorming exercise. Coupled to this is
6-11
Masters Level Module – Research Methods

the need to restrict the issues encompassed in the questions, so that


compounding or offsetting opinions are not hidden within the answers.

6.3.2 Operation
Some of the most commonplace approaches to collecting data for
construction research involve questionnaires and interviews. Many of the
principles of their operation are similar, the principal distinctions
between a questionnaire and an interview lie in the administering of the
questions.

With the questionnaire, the questions are self-administered, which can


introduce a particular layer of uncertainty into the data validity. With an
interview, researchers or their agent administer the questions. Both
methodological techniques are designed with the explicit criterion of
repeatability and categorisation of data in mind.

Interviews may be further distinguished from questionnaires as being


structured (also known as standardised) or semi-structured (also referred
to as unstandardised, or unstructured).

The structured interview operates along the same parameters as the


questionnaire. The focus and detail of the data collection is prescribed
before it is collected, and one of the major strengths of the methodology
lies in its apparent repeatability and consistency. This is the more-
common approach to data gathering using interviews, but requires a
firmer conceptual hypothesis than that needed for the sound operation of
the semi-structured interview. There is therefore a distinction in their
operation based on whether they are to be used for theory-testing or
theory-building.

The presentation of questionnaires is very important, since they are


designed to be filled in remotely and without supervision or control by
the researcher. It can be very difficult to establish whether the person
who filled in the questionnaire was taking it seriously, had the necessary
knowledge, or was indeed of the sample category that the researcher
anticipated or assumed. There is therefore a potentially large degree of
variability in the results which needs to be taken into account when

6-12
Masters Level Module – Research Methods

drawing conclusions. A common response to this is to select large


samples and examine the responses carefully for deviation from the
standard band of responses.

It is also important to avoid including extra questions which do not


actually add to the data required for testing the hypothesis but tend to
clog up the processing system. Questionnaires and interviews should be
concise and targeted on specific data collection purposes. Careful
consideration of the coding purposes prior to finalisation of the
questionnaire or structured interview can save a great deal of time and
effort in the analysis stage.

Always pilot a questionnaire or interview (structured or semi/


unstructured). Errors across a full set of questionnaires or interviews can
easily invalidate the entire findings, and ambiguity in the questions is
surprisingly easy to achieve but difficult to detect. The other advantage
of piloting is that other issues which could be usefully included (or
excluded) in the questionnaire/ interview can come to light following the
de-briefing of the pilot respondents.

6.3.3 Strengths And Weaknesses


Questionnaires and interviews both perform well in circumstances where
there is a firm conceptual model which needs populating with data.
Interviews have an additional advantage over questionnaires for
formative research since they can be used in a semi-structured manner to
gather conceptual understanding as well as measurable responses to more
prescribed (theory-testing) issues.

The strength of questionnaires lies usually in their generalisability, and


therefore a random but representative sample is a central requirement. It
can be difficult to establish why the data collected does not match that
anticipated, since the richness and quality of data is sometimes lost in the
search for high-volume simple quantitative (or semi-quantitative) data.

Structured interviews require an absolutely consistent questioning format


by the researcher, which can be problematic where there is more than one
field worker collecting the data. There can also be a hidden and

6-13
Masters Level Module – Research Methods

unintentional leading by a fieldworker using a semi-structured or


unstructured interview format where they have a particular personal bias.
The phrasing of questions can easily place pre-conceptions into a
respondent’s mind, or constrain them from giving a truly neutral
response.

The widespread application of questionnaires means that there is


frequently a very strong possibility that a similar form of questionnaire to
that suitable for the data collection has been used before. If it can be
found and proves to have been validated, it may be feasible to modify it
for use and increase the certainty of the data collection process.

Be aware that when collecting opinion-based findings there is always the


risk that an unforeseeable event will create a profound change in the
focus of opinion being sought. This is an external validity issue which
could create externally-originated bias in the entire data sample. If part
way through the data collection, such an event could cause a differential
skew in the data which would require careful analysis to deal with. A
chronological record of the collection of data using questionnaires or
interviews is a useful precaution, and also allows the customer of the
research to observe patterns in the data which may be due to factors
invisible at the time of the initial research analysis.

6.3.4 Uses
The careful selection of measurement scales can allow semi-quantitative
or quantitative data to be gathered using questionnaires or interviews. An
alternative is the use of fixed alternative questions, which operates rather
like a multiple-choice exercise. By devising scales or set responses the
researcher pre-determines the categorisation of the data and makes
processing and analysis much more straightforward. Where this data
would be too constrained for the intended use, a semi-structured
interview could be used to allow open-ended questions to be asked. The
resulting data would be less focused, although the start point (and
therefore the initial attachment of data to the conceptual frame of
reference which the interview was designed to gather data for) would be
preserved. Whilst being less focused and therefore more difficult to

6-14
Masters Level Module – Research Methods

codify, the data would be richer and could support theory-building


research.

Cyclical questionnaires or structured interviews can be used to support


the Delphi Approach, a methodological tool which allows the researcher
to gather expert opinion in a structured manner. A series of rounds of
questioning provides qualitative data on an issue of contention where
expertise opinion is either diverse or not conclusive. Collecting the data
together to form a consensus allows the researcher to develop theory in a
structured manner − it can also be used to capture quantitative opinion
data.

6.3.5 Summary Points For Questionnaires and Interviews


Critically dependent on sample size, response rate and sample
representativeness.

Randomness issue for generalisable results.

Can be used for theory building, but tend to be applied to theory testing.

According to format can deliver qualitative and/ or quantitative data.

Risks of bias. Piloting essential.

Strength lies in consistency across large samples.

Both are recognised data collection techniques.

PF6.6

What are the operational requirements for the use of questionnaires or


interviews?

6-15
Masters Level Module – Research Methods

6.4 The Case Study Approach


This technique has been used for many years as a powerful tool in social
science research. Case studies usually serve to illuminate the rich
circumstances of reality within a particular data boundary.

6.4.1 Requirements
Case studies have few generalisable requirements other than access to the
level of data required for the study. Consistency of depth of access is
important to get a coherent picture. It is also essential to pay attention to
the frame of reference or explanatory template (more on this in Chapter
7) used for the analysis of the data with this sort of research.

6.4.2 Operation
Case study data is, by definition, particular to the case from which it was
collected. Therefore the generalisability of the results is limited to those
principles which emerge and could be of a general nature. Establishing
the generalisability of results is usually done using the criteria of
replication. Cross-case analyses may be employed for this purpose to
compare the results from several broadly similar cases to assess if there
are any consistent and generic trends which can be concluded.

It is quite common for case studies to embed other methodological


techniques in a mixed methodological approach. For instance, case study
data may be collected using ethnographic studies, questionnaires,
structured and/or unstructured interviews, or the inspection of documents
and organisational protocols. All of these could be applied. It is also
increasingly common to find that an action research methodology
originated from a case study, therefore case study may itself be
embedded in other methodological strategies.

In addition to embedding other methodological techniques, case studies


can follow up and elaborate on initial data findings. Hence a widespread
questionnaire or interview may be used to create a general theory of

6-16
Masters Level Module – Research Methods

processes or variables affecting a particular type of situation. This can


then be complemented by a follow-up case study to explore the specific
case of the findings and discover further richer detail about the
underlying process or variables.

6.4.3 Strengths And Weaknesses


The approach can be quite supple, that is the complexity of the situation
and the desired data can inform the structure (or lack of it) which is
designed into the case study. The principal strength of a case study
derives from this, that the researcher can focus on the system dynamics
and processes of a situation, or case, in great detail. The flexibility
extends across into the duration of the research, letting the researcher ‘go
with the flow’ and redefine the focus of the case study as the data
emerges. This sort of application clearly supports theory building as a
series of propositions can be derived and elaborated upon using the case
under focus.

One of the most characteristic problems with case study research lies in
the volume of data which is generated and (paradoxically) its richness. It
is easy to get lost in the detail of the data and lose sight of the bigger
picture and the direction of the research. It is also possible to become
completely swamped by data and find it impracticable to make
conclusions due to data overload.

6.4.4 Uses
The case study approach is useful for allowing a particular issue to be
studied in detail and in the context of its relationship with the real world.
It is also beneficial for some studies that the case study can operate over
a long period of time, which can bring with it other positive certainties
over the consistency and representativeness of data. An extension of the
case study is the Case-Based Reasoning technique, which uses pattern
analysis in unstructured data generated from case study to derive
relational theories; also Ethnographic Studies, which involve a written
representation of a culture based on evidence derived from case study
circumstances. Ethnography often involves a high degree of immersion
6-17
Masters Level Module – Research Methods

in the study environment, similar to that required for action research, but
without the deliberate manipulation of change that characterises action
research. Grounded Theory, an approach which is used to generate theory
from the bottom-up analysis of evidence, is also best supported by a case
study methodological approach.

6.4.5 Summary Points For The Case Study Approach


Usually serve to illuminate the rich circumstances of reality within a
particular data boundary.

Have few generalisable requirements other than access to the level of


data required for the study.

Quite common for case studies to embed other methodological


techniques in a mixed methodological approach (case study as an
umbrella technique)

Case studies can follow up and elaborate on initial data findings


produced from other data collection techniques.

A principal strength of case study derives from its supple nature.

One of the most characteristic problems with case study research lies in
the volume of data which is generated and (paradoxically) its richness

Useful for allowing a particular issue to be studied in detail and in the


context of its relationship with the real world.

It is also beneficial for some studies that the case study can operate over
a long period of time.

PF6.7

What differentiates the case study, a case report, an instrumental case


study, a collective case study, and intrinsic case studies? What is similar
about them?

6-18
Masters Level Module – Research Methods

6.5 Action Research


This methodological technique provides the most problem-oriented
approach to research.

6.5. 1 Requirements
Action research operates on the basis of participation. In contrast to the
scientific method ideal of objectivity and disinterest, the action
researcher adopts a collaborative research strategy with those being
studied. Action research requires a pragmatic researcher for its successful
use. Action research requires access to the case being studied to allow the
necessary participation if it is to be successful. The other principal
requirement for successful action research is that the researcher or their
proxy has the authority and capability to make a change. Full control
over changes is usually needed. This allows them to deliberately
manipulate variables and case environments and observe the effects.

6.5.2 Operation
The researcher becomes part of the research problem and instigates
changes in the independent variable(s). It is common for action research
to emerge from a case study situation, and for the outcomes of the action
research to be jointly analysed by the client and the research team. As the
access required for the case study observations introduces the researcher
to the organisational representatives, a level of trust may build up which
allows the researcher the deeper level of access and role required for
action research.

One of the concerns about action research is the difficulty of finding


representative control cases with which to compare the manipulated
change. Where a case study evolves into a action research case, it may be
possible to use the initial non-participative case study as a contextual
setting for the analysis of the action research phase of the research.

6-19
Masters Level Module – Research Methods

6.5.3 Strengths And Weaknesses


One of the main strengths of action research is that direct intervention is
possible. This avoids the uncertainty or chance element of having to wait
for a change to occur. A high degree of involvement in the case being
researched provides a compensating rich picture of the system dynamics
of the change process and allows a richer analysis of the phenomenon
than a remote, objective observation. Action research is particularly
appropriate for real-world research, since businesses prefer to change
with their instinct and review the change as they proceed, adjusting
where necessary. Action research can be highly appropriate for the
situation, and can allow a rapid application of theories. It is applicable
recurrently.

One of the counterarguments to this is that action research involves the


application of (premature) formative theories rather than developed
theories. The main criticism from the positivists is that direct intervention
produces a non-objective outcome and view − not only do researchers
distort that which they are intending to observe (and the distortions may
be several or produce incrementing effects), but they also place
themselves so close to the cause, process, and effect that they cannot
distinguish objectivity from subjectivity.

Procedurally, action research control groups are unlikely and usually


impracticable; there is an associated loss of control over extraneous
variables; and the population validity is limited to the sample used.
Action research is difficult to plan, and vulnerable to coincidental
changes in a system clouding the cause-process-effect being assessed.

6.5.4 Uses
Action research has not been accepted in all disciplines. It clearly suits
situations where change is planned or imminent, and allows the
researcher in collaboration with the host organisation to analyse a rapid
and significant change. It is common for action research to start with the
client or the researcher presenting a problem and agreeing mutual goals
and research participation, in addition to mutual agreement over changes
that will be made. Action research will probably produce results in the

6-20
Masters Level Module – Research Methods

form which business can most readily accept and re-apply for useful
effect. At its extreme, it is a hybridisation of research consultancy.

6.5.5 Summary Points For Action Research


Operates on the basis of participation.

Requires access to the case being studied to allow the necessary


participation if it is to be successful.

The researcher or their proxy requires the authority and capability to


make a change.

Involves deliberately manipulating variables and case environments and


observing the effects.

Common for action research to emerge from a case study situation, and
for the outcomes of the action research to be jointly analysed by the
client and the research team.

Difficulty of finding representative control cases to compare the


manipulated change with.

Particularly appropriate for real-world research,

Main criticism from the positivists is that direct intervention produces a


non-objective outcome and view.

Action research is difficult to plan, and vulnerable to coincidental


changes in a system clouding the cause-process-effect being assessed.

Not been accepted in all disciplines.

PF6.8

How can the researcher protect the rigour of the research process and
outcomes when using action research?

6-21
Masters Level Module – Research Methods

6.6 Choosing A Methodological Technique


Select a methodology which answers the research question(s) and
addresses the overall research problem. If the research question is multi-
faceted, there may a requirement for mixed data to answer it. In such
cases a mixture of methodological approaches may be appropriate (or a
simplification of the research question). It is possible to use multiple
methodological techniques and combine the methodological paradigms,
producing a triangulated view on an issue. This can be very powerful,
since it can offer cross-validation of the data and the research techniques,
thereby producing a much richer research picture. It requires skill in
operation and the combined analysis of data however, and is probably not
to be recommended as a first venture into research. The need for mixed
methodological techniques may be indicative of a particularly complex
phenomenon which could be broken down into sub-issues for researching
separately.

Consider using a series of preliminary or pilot studies to generate a


coarse, low rigour data set to establish the viability and value of a more
extensive or more detailed data collection.

Select a methodology which satisfies the highest appropriate level of


rigour for the data requirements, no more. There is no point in using a
high rigour methodology if the variables are trivial. If you are developing
ideas or theories it may be appropriate to use a methodology with a low
or medium level of rigour.

Low Rigour includes qualitative, descriptive and/or narrative data, which


are characterised by a low scope for the classification or the variables.
Anecdotal opinion or single case study, semi-structured interviews may
provide the appropriate level of data for this sort of research. These forms
of data may be very useful for identifying the most appropriate
hypotheses for further testing.

Medium Rigour methodologies include uncontrolled experiments or the


collection of survey data. These are more suitable where there is a
formative conceptual model in place which is not sufficiently detailed or
robust to warrant precision data collection, but has sufficient structure
and support to benefit from some degree of measurement and

6-22
Masters Level Module – Research Methods

manipulation of variables. In such cases it may be possible to explain the


dynamics of some relationships.

High Rigour methodologies are best suited to highly prescribed or


detailed theory testing. The quantitative methods fall into this category,
conducted using the scientific method . In such cases a detailed
hypothesis is probable, built upon identified key relationships between
variables. The data collection may be associated with calibration of a
model rather than with its generation or formation.

Activity

Using your chosen research topic, select a methodological technique


which appears to be suitable for your data requirements and appraise the
strengths and weaknesses of the technique against the requirements of
your research project.

Identify those weaknesses which coincide between the research project


and the methodology. What are you going to do to control the effect of
these on the research?

Record your thought here;

6-23
Masters Level Module – Research Methods

6.7 Summary
The selection of the methodological approach and data collection
technique represents the major commitment in the active research phase.
After working through this chapter you should now be sufficiently aware
of the various mainstream methodological techniques to allow you to do
this. Further reading has been provided for more detailed aspects which
will arise in their use. The next stage in the research process is to collect
the data, and then to analyse it. Chapter 7 deals with the analysis of data,
which should be undertaken on a regular basis as the research proceeds.

6.7.1 Further Personal Feedback Questions


These questions are designed to test your overall understanding of the
concepts presented in this chapter. Try to answer them all before
proceeding to Chapter 7.

PF 6.9

What are independent variables and dependent variables, and how do


they relate?

PF 6.10

What is positivistic research?

PF 6.11

What are the principal conceptual and operational distinctions between


true experiments and quasi-experiments?

PF 6.12

What are the strengths and weaknesses of experiments?

PF 6.13

What are the strengths and weaknesses of quasi-experiments?

PF 6.14

What are the strengths and weaknesses of questionnaires and interviews?


6-24
Masters Level Module – Research Methods

PF 6.15

What are the strengths and weaknesses of the case study approach?

PF 6.16

What are strengths and weaknesses of action research?

PF 6.9

What are the principal differences between the Qualitative Approach and
The Quantitative Approach to research?

6.7.2 Recommended Reading


Bell, J. 1993 Doing Your Research Project: A Guide For First Time
Researchers in Education and Social Science Open University Press (2nd
Edition) ISBN 0-335-19094-4. See Introduction to Section 2 (pp 61 - 66)
and Chapter 6 The Analysis of Documentary Evidence (pp 67 - 74);
Chapter 7 Designing and Administering Questionnaires (pp 75 - 90);
Chapter 8 Planning and Conducting Interviews (pp 91 - 101); and
Chapter 10 Observation Methods (pp 109 - 122).

Mark, R., 1996 Research Made Simple: A Handbook for Social Workers.
Sage Publications. ISBN 0-8039-7427-2. See Chapter 10 Qualitative
Research (pp 206 - 229); and Chapter 12 Sources of Data:
Questionnaires, Interviews, Schedules, and Available Materials (pp 241 -
268).

Mason, J., 1996 Qualitative Researching Sage Publications ISBN 0-


8039-8986-5. See Chapter 3 Generating Qualitative Data: Interviewing
(pp 35 - 59); and Chapter 4: Generating Qualitative Data: Documents and
Visual Data (pp 60 - 82).

Maxwell, J., 1996 Qualitative Research Design: An Interactive


Approach. Sage Publications Applied Social Research Methods Series
Vol. 41. ISBN 0-8039-7329-2. See Chapter 5: Methods: What Will You
Actually Do? (pp 63 - 85)

6-25
Masters Level Module – Research Methods

Yin, R.K., 1984 Case Study Research: Design and Methods Beverly
Hills, Calif. Sage Publications. A classic on case study.

Further Reading on Methodologies

Altheide, D.L., 1996 Qualitative Media Analysis Sage University Paper,


Qualitative Research Methods Series 38. ISBN 0-7619-0199-X. For
document analysis.

Bryan, A., 1993 Quantity and Quality in Social Research Routledge


Contemporary Social Research Series No 18. ISBN 0-415-07898-9. See
Chapter 2 The Nature of Quantitative Research (pp 11 - 44) and Chapter
3 The Nature of Qualitative Research (pp 45 - 71).

Hague, P., 1993 Questionnaire Design Kogan Page The Market Research
Series ISBN 0-7494-0917-7. A practical guide to the questionnaire
methodology.

Howard, K., Sharp, J.A., 1983 The Management of a Student Research


Project Gower ISBN 0-566-00613-8. See Chapter 6 Gathering The Data
(pp 121 - 150).

Kvale, S., 1996 Interviews: An Introduction to Qualitative Research


Interviewing Sage Publications ISBN 0-8039-5820-X. A complete book
on interviewing. Authoritative and analytical.

Moroney, M.J., 1980 Facts from Figures Pelican ISBN 0-14-02-0236-8.

Oppenheim, A.N., 1996 Questionnaire Design, Interviewing and Attitude


Measurement Pinter. ISBN 1 85567 044 5. Thorough.

Sapsford, R., Jupp, V., 1996 Data Collection and Analysis Sage
Publications ISBN 0-7619-5046-X. Covers a wide range of research
techniques and approaches, including quantitative analysis.

Schuman, H., Presser, S., 1996 Questions and Answers in Attitude


Surveys: Experiments on Question Form, Wording, and Content Sage
Publications. ISBN 0-7619-0359-3. Detailed coverage of the principles
and sensitivity of questions in research.

6-26

You might also like