Research Design Paper Guideline for QUALITATIVE METHODS: CASE STUDIES AND COMPARATIVE ANALYSIS
RESEARCH QUESTION: The research question takes the form of: “What causes Y?” or “Why does Y occur in some cases but not others?” or “What is the impact of X on Y?” or “Under what conditions does Y occur?” In all these questions, “Y” represents the outcome that you wish to explain or the “Dependent Variable.” In the paper, you should also briefly address why your research question is interesting and worth studying. (Note: Please make sure you have watched the short Loom video on Moodle under Week 2 in which I give additional guidance on how to pick a research question for this assessment)
CONCEPT & MEASUREMENT: Using the literature and your own ideas, provide an abstract definition for the phenomenon that you are seeking to explain. Then discuss what kinds of criteria you would use to identify the different values of the concept in the cases. Different values of a concept might refer to the existence versus non-existence of a phenomenon (binary concept) or the degree to which a phenomenon exists (continuum).
THEORY (based on existing literature and/or your argument): What have other people said about your research question? What are the existing arguments and their empirical predictions? Do you have a separate argument or are you going test the arguments that already exist in the literature? If you have a separate argument, how is it different from the existing arguments in its empirical/observable implications?
Specify whether you will only rely on comparative case analysis or whether you would use case study analysis combined with statistical methods. Then defend your selection of cases. Remember that “cases” can be different types of units such as countries, subnational units within the same country (i.e. cities, villages, and subnational level states in federal systems), as well as different time-periods within the history of one country. You should defend your case selection based on:
- Your main outcome of interest (dependent variable): How does the dependent variable vary in the cases you compare?
- Controlling for alternative explanations: What are alternative arguments you are controlling for by choosing the cases? When you control for certain conditions, these conditions are similar or same in the cases.
- If your case study is designed to come after large-N analysis, provide a short description of the data you would use and the main hypotheses you would test in the large-N analysis. Then tell us which cases you will select to evaluate the causal logic of different arguments and why these are good cases to focus on.
METHODOLOGICAL TOOLS AND TYPES OF DATA
Specify the methodological tools and types of data that you plan to rely on. Make sure you refer back to the theory section and articulate how your findings would help you support or contradict the empirical implications of different arguments.
Secondary sources: Give examples of published works (for example, specific books, articles, biographies) you would use as well as how you would use them. What kinds of information will you be looking for in these works and why? How would you use secondary sources for process tracing or counterfactual analysis?
Fieldwork: What type of fieldwork tools will you use, e.g. surveys, qualitative interviews, focus groups? What types of findings from these surveys, interviews, or focus groups would support or contradict the theories that you mention in the theoretical section? How would you have access to the site and build the connections that you need to conduct surveys, interviews, or focus groups?
Archival: Which archives will you use? Why? What type of information do the archives entail? What kinds of findings from the archives would support or contradict the theories that you are testing? How will the archives help you with process tracing and/or counterfactual analysis?
POTENTIAL SOURCES OF BIAS:
Regardless of which methods you use, you need to address the issues below (there might of course be additional issues that you need to address depending on your research design paper topic). Briefly discuss the extent to which each constitutes a problem and how you would tackle these problems.
- Bias (selection bias & bias in sources/archives/reporting)