Using case study_in_research

Using Case Studies in Research
by Jennifer Rowley
Introduction
Case study as a research strategy often emerges as an obvious option for stu-
dents and other new researchers who are seeking to undertake a modest scale
research project based on their workplace or the comparison of a limited
number of organisations. The most challenging aspect of the application of
case study research in this context is to lift the investigation from a descrip-
tive account of ‘what happens’ to a piece of research that can lay claim to be-
ing a worthwhile, if modest addition to knowledge. This article draws
heavily on established textbooks on case study research and related areas,
such as Yin, 1994, Hamel et al., 1993, Eaton, 1992, Gomm, 2000, Perry,
1998, and Saunders et al., 2000 but seeks to distil key aspects of case study
research in such a way as to encourage new researchers to grapple with and
apply some of the key principles of this research approach. The article ex-
plains when case study research can be used, research design, data collec-
tion, and data analysis, and finally offers suggestions for drawing on the
evidence in writing up a report or dissertation.
When to use Case Studies
Case studies as a research method or strategy have traditionally been viewed
as lacking rigour and objectivity when compared with other social research
methods. This is one of the major reasons for being extra careful to articulate
research design, and implementation. On the other hand, despite this scepti-
cism about case studies, they are widely used because they may offer in-
sights that might not be achieved with other approaches. Case studies have
often been viewed as a useful tool for the preliminary, exploratory stage of a
research project, as a basis for the development of the ‘more structured’
tools that are necessary in surveys and experiments. For example, Eisen-
hardt (1989) says that case studies are:
Particularly well suited to new research areas or research areas for
which existing theory seems inadequate. This type of work is highly
complementary to incremental theory building from normal science
research. The former is useful in early stages of research on a topic or
when a fresh perspective is needed, whilst the latter is useful in later
stages of knowledge (pp.548-549).
This is however a somewhat narrow conception of the application of
case study research. As discussed below case studies are useful in providing
answers to ‘How?’ and ‘Why?’ questions, and in this role can be used for ex-
ploratory, descriptive or explanatory research.
16 Management Research News
Biographical Note
Professor Jennifer
Rowley can be
contacted at the School
of Management and
Social Sciences, Edge
Hill College of Higher
Education, Ormskirk,
Lancashire, England
L39 4QP.

The first stage is to decide whether case studies can be useful for a spe-
cific kind of investigation. There are three factors that determine the best re-
search methodology:
• The types of questions to be answered
• The extent of control over behavioural events, and
• The degree of focus on contemporary as opposed to historical
events.
The issue of types of research question is the most significant in deter-
mining the most appropriate approach. Figure 1 (based on Yin, 1994, p.6)
summarises the different kinds of research questions and methods that are
most appropriate. Who, what and where questions can be investigated
through documents, archival analysis, surveys and interviews. Case studies
are one approach that supports deeper and more detailed investigation of the
type that is normally necessary to answer how and why questions.
Case study research is also good for contemporary events when the rele-
vant behaviour cannot be manipulated. Typically case study research uses a
variety of evidence from different sources, such as documents, artefacts, in-
terviews and observation, and this goes beyond the range of sources of evi-
dence that might be available in historical study.
In summary then, case study research is useful when:
A how or why question is being asked about a contemporary set of
events over which the investigator has little or no control. (Yin,
1994, p.9)
In contrast to surveys, typically the number of units studies in a case
study is many less than in a survey, but the extent of detail available for each
case should be greater. As compared with an experiment, the case study re-
searcher has much less control over the variables, than if an experiment were
used to investigate a situation. In a survey data may be collected from a
number of organisations in order to generalise to all other organisations of
the same type. In contrast in a comparative case study across a number of dif-
ferent organisations, the objective is to compare or replicate the organisa-
tions studied with each other in a systematic way, in the exploration of
different research issues.
Volume 25 Number 1 2002 17
Using
Case Studies
in Research
Figure 1: Choosing a Research Strategy
Strategy Form of research question
Experiment
Survey
Archival analysis
History
Case study
How, why
Who, what, where, how many, how much
Who, what, where, how many, how much
How, why
How, why

What iscase study reseach? Yin (1994) p.13 defines a case study thus:
A case study is an empirical inquiry that:
• Investigates a contemporary phenomena within its real life
context, especially when
• The boundaries between phenomenon and context are not clearly
evident.
This statement emphasises that an important strength of case studies is
the ability to undertake an investigation into a phenomenon in its context; it
is not necessary to replicate the phenomenon in a laboratory or experimental
setting in order to better understand the phenomena. Thus case studies are a
valuable way of looking at the world around us. On the other hand, it is im-
portant not to confuse case studies with ethnographic and other strictly
qualitative research paradigms. Case study research can be based on any mix
of quantitative and qualitative approaches. Typically, it uses multiple data
sources including two or more of: direct detailed observations, interviews,
and documents. In addition, case studies can involve single or multiple cases
as discussed in the next section on research design.
Research Design
Research design often seems to be something of a mystery to new research-
ers, and the proneness of research philosophers to engage in sophisticated
debates using terminology that is inaccessible to the novice does not help.
On the basis that it is necessary to grasp the basics, and undertake some re-
search before arriving at the position where some of these debates start to
have some meaning, this section takes a very practical approach to research
design. For those that need a health warning, this section takes a positivist
and deductive approach to case study design. It urges the definition of ques-
tions and propositions in advance of data collection. This is in contrast to al-
ternatives such as the grounded theory or inductive approach, in which
questions, insights, propositions, and pictures emerge from the data collec-
tion. The authors are of the opinion that the positivist approach provides a
firmer foundation for understanding and managing issues such as validity
and reliability, and for structuring data collection and analysis, and is there-
fore a more straightforward process for the new researcher.
A research design is the logic that links the data to be collected and the
conclusions to be drawn to the initial questions of a study; it ensures coher-
ence. Another way of viewing a research design is to see it as an action plan
for getting from the questions to conclusions. It should ensure that there is a
clear view of what is to be achieved by the case study. This involves defining
the basic components of the investigation, such as research questions and
propositions, appreciating how validity and reliability can be established,
and selecting a case study design.
Using
Case Studies
in Research

Components of research design
A research design has the following components:
• The study’s questions
• The study’s propositions
• The study’s units of analysis
• The logic linking the data to the propositions
• The criteria for interpreting findings.
The previous section has already explored the nature of research ques-
tions. Starting with clearly formulated questions is useful for all research
projects. Formulating research questions is never easy. Theory as embodied
in the literature of a discipline is important in pointing towards appropriate
research questions. Both practitioners and other researchers can generate
questions that are of general interest, and that therefore might be fully ex-
plored in the context of the proposed case study. Sometimes with exploratory
research the questions may have yet to be formulated; in this case the purpose
of the research still needs to be defined.
Descriptive and explanatory studies need propositions. Research ques-
tions need to be translated into propositions. The researcher has to make a
speculation, on the basis of the literature and any other earlier evidence as to
what they expect the findings of the research to be. The data collection and
analysis can then be structured in order to support or refute the research
propositions.
The unit of analysis is the basis for the case. It may be an individual per-
son (such as a business leader, or someone who has had an experience of in-
terest), or an event, (such as a decision, a programme, an implementation
process or organisational change), or an organisation or team or department
within the organisation. It can sometimes be difficult to identify the bounda-
ries of the unit of analysis. A key issue is that the case study should only ask
questions about the unit of analysis, and any sub-units; sources of evidence
and the evidence gathered are determined by the boundaries that define the
unit of analysis.
Selecting the unit of analysis, or the case is crucial. Case selection must
be determined by the research purpose, questions, propositions and theoreti-
cal context, but there will also be other constraints that impact on case selec-
tion. These include accessibility (whether the data needed can be collected
from the case individual or organisation), resources (whether resources are
available to support travel and other data collection and analysis costs), and
time available (if time is limited, it may be easier to analyse a small business
rather than a large business, or to identify a unit of analysis within a large or-
ganisation, rather than seek to study the organisation in its entirety.
Using
Case Studies
in Research

Finally, it is necessary to decide what data is necessary in order to sup-
port or demolish the propositions, and to reflect on the criteria for interpret-
ing the findings. These issues are explored in more detail later in the article.
Generalisation, Validity and Reliability
These three concepts establish the basis on which other researchers should
regard a piece of research as knowledge that can be assimilated into the
knowledge base of a field of study. It is therefore important to demonstrate
that these issues have been fully considered.
Generalisation of the case study so that it contributes to theory is im-
portant. Generalisation can only be performed if the case study design has
been appropriately informed by theory, and can therefore be seen to add to
the established theory. The method of generalisation for case studies is not
statistical generalisation, but analytical generalisation in which a previously
developed theory is used as a template with which to compare the empirical
results of the case study. If two or more cases are shown to support the same
theory, replication can be claimed. In analytic generalisation, each case is
viewed as an experiment, and not a case within an experiment. The greater
the number of case studies that show replication the greater the rigour with
which a theory has been established.
Four tests have been widely used to establish the quality of empirical
social research:
1. Construct validity - establishing correct operational measures for
the concepts being studied. This is concerned with exposing and
reducing subjectivity, by linking data collection questions and
measures to research questions and propositions.
2. Internal validity (for explanatory or causal studies only, and not
for descriptive or exploratory studies) establishing a causal
relationship whereby certain conditions are shown to lead to other
conditions, as distinguished from spurious relationships.
3. External validity: establishing the domain to which a study’s
findings can be generalised. Generalisation is based on replication
logic as discussed above.
4. Reliability: demonstrating that the operations of a study - such as
the data collection produced can be repeated with the same results.
This is achieved through thorough documentation of procedures
and appropriate recording keeping.
Many of the approaches for ensuring validity and reliability are dis-
cussed further below in the sections on data collection and analysis.
Using
Case Studies
in Research

Case Study Designs
As summarised in Figure 3 case study designs can be categorised along
two dimensions, reflecting whether the number of case studies contributing
to the design, and the number of units in each case study respectively.
The differentiation between single case and multiple case designs needs
to be clearly made. A single case design is akin to a single experiment. Single
case studies are appropriate when the case is special (in relation to estab-
lished theory) for some reason. This might arise when the case provides a
critical test to a well-established theory, or where the case is extreme, unique,
or has something special to reveal. Single case studies are also used as a pre-
liminary or pilot in multiple case studies.
Multiple case designs are preferred. On the basis of the replication logic
discussed above, multiple cases can be regarded as equivalent to multiple ex-
periments. The more cases that can be marshalled to establish or refute a the-
ory, the more robust are the research outcomes. A frequent question is how
many cases should be included in a multiple case study. There is no simple
answer to this question. Cases need to be carefully selected so that they either
produce similar results (literal replication), or produce contrasting results but
for predictable reasons (theoretical replication). Typically within say six to
Using
Case Studies
in Research
Figure 2: Checking Case Study Design
Tests Case Study tactic Phase of re-
search in which
tactic occurs
Construct validity
Internal validity
External validity
Reliability
Use multiple sources of evidence
Establish chain of evidence
Have key informants review draft case study
report
Do pattern matching
Do explanation building
Do time series analysis
Use replication logic in multiple case studies
Use case study protocol
Develop case study database
Data collection
Data collection
Composition
Data analysis
Data analysis
Data analysis
Research design
Data collection
Data collection
Figure 3: Case Study Designs
Single Case Designs Multiple Case Designs
Holistic (single unit of analysis)
Embedded (multiple units of analysis)
Type 1
Type 2
Type 3
Type 4

ten cases, a few cases might be used to achieve literal replication, whereas
others might be designed to peruse other patterns of theoretical replications.
If all of the cases turn out as predicted then there is strong evidence for the
initial set of propositions. If, however, the cases reveal a variety of different
outcomes, it may be necessary to revisit the propositions, and consider con-
ducting further research. The number of cases depends on the nature of the
propositions to be substantiated.
Case studies can also be divided into holistic or embedded studies. Ho-
listic case studies examine the case as one unit. They might, for example, fo-
cus on broad issues of organisational culture or strategy. This approach
ensures a helicopter view of the case, but can be superficial, and may miss
changes in the unit of analysis that could impact on the appropriateness of
the original research design. Embedded designs identify a number of sub
units (such as meetings, roles or locations) each of which is explored indi-
vidually; results from these units are drawn together to yield an overall pic-
ture. The biggest challenge with embedded designs lies in achieving a
holistic perspective from the analysis of the sub-units.
Data Collection
Data collection, and in general the execution of a good case study, depend
crucially upon the competence of the researcher. Unlike, with say, a ques-
tionnaire, the evidence to be gathered is defined as it is collected, and the re-
searcher is an active agent in the process. This means that the researcher
undertaking data collection needs to be able to ask good questions, to listen
and to interpret the answers. This involves having a sound grasp of the ques-
tions and propositions of the case study, and being able to approach the study
in an unbiased, and flexible manner.
Data collection should be guided by a case study protocol. This proto-
col needs to include the following sections:
1. An overview of the case study project.
2. Field procedures, such as use of different sources of information,
and access arrangements to these sources.
3. Case study questions, or the questions that the case study
researcher needs to keep in mind when collecting data. These
questions are posed to the researcher, and not to any specific
respondents, although they may be used to guide the formulation
of questions to interviewees, and members of focus groups. In
complex cases studies it is important to differentiate between the
questions asked of specific interviewees and used to interrogate
documents, questions asked of the individual case, and questions
to be asked across multiple cases.
Using
Case Studies
in Research

Larger case studies, with multiple cases and embedded designs with
several sub units in each case study are likely to need a team of researchers.
When more than one researcher is engaged in gathering evidence, the case
study protocol is a central communication document for the team. However,
it is not sufficient to draft such a document, and leave it for researchers to
read. The success of the project depends upon the quality of the researcher’s
engagement with the problem and the agreed process. Training and partici-
pation in research design are important in ensuring that a team has an appro-
priate level of familiarity with the case study investigation.
Gathering evidence
Typically case studies draw on multiple sources of evidence. These include
documents, archival records, interviews, direct observation, participant ob-
servation, and physical artefacts. Each of these different sources require dif-
ferent approaches to their interrogation, and are likely to yield different kinds
of insights. Each source has its strengths and its weaknesses, and the richness
of the case study evidence base derives largely from this multi-facetted per-
spective yielded by using different sources of evidence.
Whichever sources of evidence are used, there are three key principles
of data collection that need to be observed:
1. Triangulation - one of the great strengths of case studies as
compared with other methods is that evidence can be collected
from multiple sources. Triangulation uses evidence from different
sources to corroborate the same fact or finding.
2. Case Study Database - A case study database of the evidence
gathered needs to be collected. Whilst a report or dissertation may
be the primary distillation of the case study, a further outcome
which strengthens the repeatability of the research, and increases
the transparency of the findings is a well organised collection of the
evidence base. This base may include case notes made by the
investigators, case study documents that are collected during a case
study, interview notes or transcripts, and analysis of the evidence.
When preparing a dissertation it will be useful to agree with a
supervisor whether some elements of this evidence base should be
presented as appendices to the dissertation.
3. Chain of Evidence - The researcher needs to maintain a chain of
evidence. The report should make clear the sections on the case
study databases that it draws upon, by appropriate citation of
documents and interviews. Also, the actual evidence needs to be
accessible in the databases. Within the database, it should be clear
that the data collection followed the protocol, and the link between
the protocol questions and the propositions should be transparent.
Using
Case Studies
in Research

Analysis of Case Study Evidence
Analysing case study evidence is not easy. Typically a case study database
will include a multitude of different evidence from different sources. Data
analysis of this rich resource is based on examining, categorising and tabu-
lating evidence to assess whether the evidence supports or otherwise the ini-
tial propositions of the study. The preferred strategy for analysis is to use the
propositions that encapsulate the objectives of the study, and which have
shaped the data collection. The researcher trawls through the evidence seek-
ing corroboration or otherwise of the initial propositions, and then records
relevant evidence and makes a judgement on whether the positions have
been substantiated. This is where you discover whether the propositions
were well formulated in the first place!!
In exploratory case studies that typically do not start with propositions,
an alternative approach needs to be adopted. Here an alternative analytic
strategy is to develop a descriptive framework for organising the case study.
Thus a framework of sections reflecting the themes in the case study are de-
veloped and evidence is gathered within relevant themes, and analysed and
compared in these categories, in order to achieve a description of the case
study that can be corroborated from multiple sources of evidence.
In general, there are no cookbook procedures that have been agreed for
the analysis of case study results, but good case study analysis adheres to the
following principles:
1. The analysis makes use of all of the relevant evidence
2. The analysis considers all of the major rival interpretations, and
explores each of them in turn
3. The analysis should address the most significant aspect of the case
study
4. The analysis should draw on the researchers prior expert
knowledge in the area of the case study, but in an unbiased and
objective manner.
Writing the Case Study Report
Writing the case study report can be a daunting task, because at this point the
researcher needs to discriminate between what is to be included and the
wealth of evidence that will not appear in the report, but stays in the case
study database. Effective analysis of the results will assist in providing a
structure. The task of writing a report or dissertation will appear less over-
whelming if the researcher has observed the advice to all researchers which
is to write up as the research proceeds. Drafts of literature review and meth-
odology sections can be written in parallel with data collection. A key factor
in determining the coverage and presentation of the case study report is the
intended audience. Case studies have a range of potential audiences, includ-
Using
Case Studies
in Research

ing academic colleagues, policymakers, practitioner professionals, the gen-
eral public, research supervisors and examiners, and funders of research.
These different audiences have different needs. For example, for non-
specialist audiences, the story that the case study tells may be most engaging,
and they may seek in the case study a basis for action. For a dissertation as-
sessor, mastery of methodology, and an understanding of the way that the re-
search makes a contribution to existing knowledge will be important.
Alternative Perspectives on Case Studies
The guidance given in this article derives largely from the approach sug-
gested by Yin, 1994. This approach is primarily positivistic in perspective,
and can be characterised by the following positions:
1. The analytic approach to generalisation.
2. Theory should inform propositions, and propositions inform data
collection and analysis.
3. The researcher acts as commentator, in representing and
interpreting the case in a way that relates to previous theory.
There is an ongoing debate around each of these positions, which re-
flects the different perspectives taken by the positivist and phenomenologi-
cal schools of research philosophy and strategy (Gomm, 2000). In order to
encourage continuing reflection on these matters it is useful to summarise
some of the questions that can be posed:
1. Generalisation - Is generalisation necessary? The need for
generalisation derives from a positivist approach in which
generalisation on the basis of samples is the norm. Can case studies
just be accepted as insights as they stand, with readers making their
own interpretation, and taking the ideas from the case study into
their own experience (sometimes called Naturalistic
generalisation). Alternatively, can case studies be used as the basis
for the formulation of working hypotheses? If attempts to
generalise are necessary, what does generalisation in this context
mean?
2. Role of theory - Is it necessary to use theory to inform
propositions, or can case studies be used as a basis for grounded
theory development, in which the theory emerges through data
collection and analysis?
3. Authenticity and authority - Whose voice is recorded in the case
study report? Some researchers use case study as a method of
‘allowing the voices of participants to be heard’. This position is
based on a rejection of any authority on the part of the study
researcher. On the other hand, some would argue that the very act
of conducting research undermines the authenticity of the voices to
Using
Case Studies
in Research

which researchers seek to listen, and that the very act of interacting
with the participants changes the case study situation.
The three points above set the agenda for further development of un-
derstanding and sophistication in approaches to case study research. The
readings listed below offer numerous insights into the debate around case
study research and the kind of knowledge that it claims to generate.
Using
Case Studies
in Research

Readings
Easton, G. (1992) Learning from case studies. 2nd edition, Hemel Hemp-
stead: Prentice Hall.
Eisenhardt, K.M. (1989) “Building theories from case study research.”
Academy of Management Review, 14(4), pp.532-550.
Hamel, J. (1993) Case study methods. California: Sage.
Ghauri, P., Gronhaug, K. and Kristianslund, I. (1995) Research methods in
business studies: a practical guide. New York: Prentice Hall.
Gomm, R., Hammersley, M., Foster, P. (Eds.) (2000) Case study method:
key issues, key texts. London: Sage Publications.
Perry, C. (1998) “Processes of a case study methodology for postgraduate re-
search in marketing.” European Journal of Marketing, 32 (9/10), pp.785-
802.
Saunders, M., Lewis, P. and Thornhill, A. (2000) Research methods for busi-
ness students. 2nd edition. Harlow: Pearson Education.
Yin, R.K. (1994) Case study research: design and methods. 2nd edition.
Thousand Oaks, CA: Sage.
Using
Case Studies
in Research

Using case study_in_research

More Related Content

What's hot (20)

Similar to Using case study_in_research (20)

Recently uploaded (20)

Using case study_in_research