36.0 Case study research
- 36.1 Ethical issues of case studies
- 36.2 Documents and other records
- 36.3 Risks of case studies
- 36.4 Data review by participants
- 36.5 Managing staged consent
Case studies are an important and popular research strategy, typically involving multiple sources of information, and a focus on particular people and their context. A case study can involve:
- a range of different research methods and approaches (interviews, focus groups, questionnaires, action research),
- use of existing information (policy documents, minutes, correspondence), and/or
- prospective data collection.
The National Statement includes case studies as one of the methods of qualitative research.
When designing your study, think through the various aspects of the research, and how they relate to each other. If you are planning a survey, followed by in-depth interviews, followed by focus group discussion of the findings, how will each part of the research relate to each other part? You will need to carefully schedule each element - including ethical review.
Case study work can include the analysis of policy documents, correspondence or other documentary sources or records. Such analysis, by itself, is not human research and does not require ethics approval. If the documents are not in the public domain, you will need the approval of the organisation or other parties involved.
Human ethics approval is required as soon as the project involves human participants or access to identifiable records that are not in the public domain. (See Section 2 for a full discussion of what research needs ethics approval.)
If your project involves both documentary analysis and human research, include a brief description of the documentary analysis in your ethics application. This provides context and demonstrates research merit and integrity.
Section 7 discusses the different kinds of risks that can be associated with human research, and strategies to address them.
Some common risks arise from the nature of the 'case':
- it will usually involve people with an existing relationship to each other whether work, personal, leisure activity or other, and
- the specific information detailed in the case study may make the people and organisation/s involved recognisable either to people who already know them or to the public at large.
Potential harms tend to be in the areas of:
- privacy and confidentiality,
- relationships, or
- reputation: personal, professional, or the standing of the organisation/s involved in the study
Risk amelioration is often a matter of ensuring that participants are fully informed of the possibility of identification by inference. This depends on the nature, identity and location of the people and organisations being studied. When designing your project, make a realistic assessment of whether
- the identity of the group or organisation can be kept confidential (eg where the groups/organisations are a small number among many with similar characteristics, then their identity can probably be concealed); or
- if the group or organisation may be identifiable, is it sufficiently large or structured in a way that allows the identity of contributors can be concealed, (eg,is it large enough and homogeneous enough for pseudonyms or other measures to disguise identities); or
- possible identification by inference is unavoidable.
Where both the organisation and the individuals involved are likely to be identifiable by inference, you must clearly explain this to potential participants, so if they consent to the research they understand that their identity may not remain fully confidential.
Case studies often involve existing relationships, and have the potential to impact on them in both positive and negative ways, for example:
- If the research raises contentious issues this has potential to harm relationships. The material may be identifiable between participants, even if it is anonymous to people outside the case study group.
- If the subject of the research is sensitive, it may add emotional overtones to the situation, and tensions between participants.
Be alert for these possibilities, and plan how you will manage them if they do occur.
In some case studies you, as the researcher, may already have a relationship with the participants, or may form one over the course of the project.
Keep in mind the difference between your role as a researcher and any other roles that you may have. Where you have another relationship with participants (work colleagues, friends, members of a church or club) you will need to find a way to be clear to everyone (including yourself) when you are a researcher gathering data.
If the project involves an organisation or institution which may be identifiable, one possible harm is damage to its reputation. Discuss this with the organisation ahead of time. Take what steps you can to maintain confidentiality, for example:
- do not refer directly to the organisation: 'a school in Melbourne's eastern suburbs';
- identify the collective eg 'universities in Melbourne' without saying which organisation obtained which results 'universities 1, 2, 3 etc'.
If the organisation is still likely to be identifiable, make sure they understand the ramifications before agreeing to be involved. Section 12.4 and 12.5 discuss obtaining approval from other organisations.
A key ethical issue in the conduct of case study-based research is the degree to which participants may be identifiable. Participants' "stories", in their own words, can potentially identify them or any third parties mentioned, especially to people who know them.
You must make sure participants are fully informed and forewarned of the possibility of identification by inference. Think through the potential consequences of identification, both to the person/s, the organisation, and third parties who might be identifiable by context.
A funny story about a mother in law is less funny if the mother in law hears about it and is upset or outraged. If the person telling the story is identifiable, so is the mother in law.
Where "stories" are likely to be identifiable, it is ethically advisable for participants to review and "sign off" on them before publication of the research outcomes, both for accuracy and "identifiability".
If you plan to have participants review their "stories", you must include details in the informed consent materials: how a review will happen, the reasons for a review, and limitations (if any) on what changes a participant may make.
See also 32.6 for a discussion of reviewing interview or focus group transcripts. The same process should apply to other media such as sound bites or edited video recording.
If your case study involves phases and different methodologies, you may obtain approval 'in principle' for the whole program of research, and final approval for the first phase, and any others for which you have 'final' methodological information. Final approval for the remaining sections of the research will be subject to provision of final research documentation.
In the example in 36.1 above, the focus group themes rely on the findings of the in depth interviews, which in their turn depend on the outcomes of the initial survey. The interview and focus group format and themes can only be finalised once the previous component of the project has been completed. This would normally be dealt with by gaining staged approval for the project.
- The initial application covers approval of the project overall and final approval for the survey.
- When the survey results are in, the themes and format for the interviews are finalised and a modification is submitted for final approval of the interviews.
- When the interview results are in, the themes and format for the focus groups are finalised and a modification is submitted for final approval of the interviews.
If your research involves multiple stages, the consent process may be complex. You may need to manage or "scaffold" consent separately for the different components of the research. This is particularly relevant if participants will be involved in one or more parts of the research, but not in all. For example:Your case study requires key stakeholder interviews, focus groups, worker interviews, and one or more surveys. Few, if any, participants will be involved in every component activity of the project, but most will be involved in more than one.
See also 9.12 for further discussion of informed consent in multiple stage research.
These guidelines were produced by Deakin Research Integrity in consultation with the Deakin University Human Research Ethics Committee and Human Ethics Advisory Groups
© Deakin University 2010. This material incorporates or is based upon part or all of Griffith University's research ethics arrangements.