Writing a case study
Case studies are not without controversy, and came to prominence via the University of Chicago’s Dept. of Sociology in the early c20th. It’s important to make a distinction between case studies as a research method, and sociology as a research discipline.
Choosing a case
- A unique event
- New data that hasn’t previously been available
- Applying a theory that isn’t dominant
Types of case
Following Stake (1995) we can consider three types of case:
- Intrinsic – where we learn about a unique phenomenon, and understand why it is different to other potential cases
- Instrumental – where we use the specified case to learn about a broader issue or phenomena
- Collective – where we study multiple cases (either in sequence of simultaneously) in order to learn about an even broader issue or phenomena
Collecting the evidence
A brief overview of the key examples of what constitutes valid evidence, plus ideas on how to get hold of them:
(i) Documentation
- Letters & memoranda
- Minutes of meetings and written reports of events (e.g. from official website)
- Administrative documents (direct from the company)
- Formal studies and evaluations (Google Scholar, Library database)
- Newspaper clippings & mass media (Lexis-Nexus, The Economist, the FT)
Documents will provide the bulk of your analysis (they’re the easiest types of information to find). Rather then base your case study on what the documentation tells you, however, try to use it to corroborate evidence from other sources.
(ii) Archival Records
- Service records – Market share, profitability (company website and annual reports)
- Organizational records – charts and budgets (ditto)
- Maps and charts
- Lists of names
- Survey data – census records etc
- Personal records – diaries, calendars (from interview contacts)
Archival records are crucial to establish the factual content of the evidence you use. Not all of the above information would be accessible/useful, but the final case study will depend on the accuracy of the archival records you uncover.
(iii) Interviews
- Open-ended – ask for the respondents opinions and insights
- Focused – following pre-arranged questions
- Survey – tightly defined questions used for multiple respondents
All methods are suitable, but be clear before conducting an interview which approach to take. Open-ended are very good if you have plenty of time, and the respondent is an expert on the subject. However this must be corroborated with archival records. Be careful not to ask leading questions, and use more focused interviews to double-check.
All interviews suffer from problems about bias and inaccurate information, so do not depend too heavily on interviews – use them to get advice, and to corroborate other evidence.
(iv) Direct Observation
A field trip is useful if the situation is still occurring – i.e. if the company is still trying to solve the same problem that you’re investigating. Observation can help you corroborate what an interviewee has told you about how the organisation is run, but due to the subjective nature it’s better if more than one of you is present.
(v) Participant-Observation
This is when you take an active role within the environment you’re observing, but is usually used if the other sources of evidence are unavailable.
(vi) Physical Artefacts
Obviously you should collect as much evidence as possible, physical artefacts might include technology, tools, instruments etc
Principles for collecting the evidence
Once you’ve got your evidence, there are a few things that ensure a rigorous case study:
- Use multiple sources. Using the classifications above, try to find information from as many different sources as possible
- Create a database. The perfect case study should have the data/evidence completely separate from the report. Whilst you’re creating and writing up your case study bear in mind that the facts should be kept separate from your analysis (as much as possible). This can be done using two stages. Firstly, write up your interview notes, and secondly, write up your analysis. In theory the first part should be made available to other researchers who have completely different research questions, but would still find your data useful.
- Maintain a chain of evidence. To make your analysis reliable the reader of your case study should be able to follow the derivation of your evidence from the research question to your conclusion (or vice versa). This is done by accurate citations (whether it’s documents, interviews or observations), demonstration of the actual evidence (and the manner in which it was collected), and that all this is consistent with the original project plan.
Writing it up
Here is a good resource for the structure of a case study:
- Case study structure, Western Sydney University
A word of advice – if your case studies are chronological you may wish to start off by writing them backwards, i.e. have a section on the current conditions and then go back to the phenomena under investigation, and finish with the firm’s strategic decision-making.
Recommended books:
- Stake R.E,. 1995, The art of case study research. London: Sage Publications Ltd.
- Yin, R., 1994, Case study research: Design and methods (2nd ed.). Beverly Hills, CA: Sage Publishing.