Presentation is loading. Please wait.

Presentation is loading. Please wait.

Case Studies and Case Selection Research Methods Workshops Oisín Tansey.

Similar presentations


Presentation on theme: "Case Studies and Case Selection Research Methods Workshops Oisín Tansey."— Presentation transcript:

1 Case Studies and Case Selection Research Methods Workshops Oisín Tansey

2 Definition of a case study George and Bennett (2005) – “the detailed examination of an aspect of a historical episode to develop or test historical explanations that may be generalizable to other events”. So: Intensive study A case is an instance of ‘a class of events’

3 Common Critiques Arend Lijphart criticisms Findings are not generalisable - Cannot test theories Difficult to replicate

4 Types of Case Study Atheoretical Interpretive Hypothesis generating Theory testing - confirming or infirming Deviant case studies Crucial case studies - Most-Likely and Least-Likely

5 Other uses of Case Studies So, a role for case studies in both generating and testing theories. Other (related) uses: Concept formation Identifying omitted variables Analysis of time and sequence

6 Trade-offs in Case Study Research Theoretical parsimony ‘v’ explanatory richness Causal mechanisms ‘v’ causal effects Theory testing ‘v’ theory generation

7 Why have multiple cases? Comparison allows control: Taking into account alternative possible explanations We can establish if apparent relationships in one case also exist elsewhere We might find they are spurious relationships Raises issues of case selection

8 Most Similar Systems Design (based on Mill’s Method of Difference’) Compare similar cases that differ in the outcome of interest Differences in the key explanatory variables across otherwise similar cases can explain the different outcomes E.g. compare two neighbouring countries

9 Most Different Systems Design (based on Mill’s Method of Similarity) Compare diverse cases with similar outcomes Similarities in key explanatory variables may explain similar outcomes Brownlee: ‘amid diversity in culture, history and economics, political commonalities elucidate general patterns’

10 Selection bias Selection choices can skew findings Classic Barbara Geddes argument: Avoid selection on the dependent variable So, identify the range of variation in your DV, and select across that range

11 Geddes on Asian Tigers

12

13 Van Evera response Selecting on DV is fine in three common circumstances:  If comparing cases to a known average  If there is large within-case variation for congruence procedures  if cases are sufficiently data-rich to allow for process tracing

14 George and Bennett response Selecting on the DV can allow you to identify which variables are not necessary or sufficient conditions Also, can aid theory building: in the early stages of a research program, selection on the dependent variable can serve the heuristic purpose of identifying the potential causal paths and variables

15 Exercise If your argument is that X leads to Y Can you be sure that: 1). Y does not exist in cases without X 2). All cases with X also have Y Solution: look for variation on DV and IV

16 Examples from the literature Page Fortna, Does Peacekeeping Work Compares cases with peacekeeping and without – correcting a flaw in the literature Compares cases with successful and failed peace efforts

17 Examples from the literature Chenoweth and Stephen, Why Civil Resistance Works Iran, Palestine, Burma, Philippines Four key reasons for case selection: Variation on the dependent variable Variation in the independent variable Allows for within-case comparison The cases are intrinsically interesting

18 Examples from the literature Judith Kelley, Ethnic Politics in Europe IO influence on ethnic policies Latvia, Estonia, Slovakia and Romania Reasons for case selection: Variation in IO treatment (IV) Variation in domestic politics (IV) Variation in national policy (DV)

19 Van Evera Two general criteria, eleven specific ones Cases should best serve the purpose of the inquiry Case selection should maximise the strength and number of tests that can be performed Strong tests evaluate predictions that are certain and unique

20 Exercises What is your case study a case of? What is the domain of your study? - where does your theory apply? Why have you selected your particular cases? Can you be sure your explanatory variable is not also present in cases with a different outcome? Or that your outcome is not also present where your explanatory variable is absent? Critically review case selection choices in key books in your literature (and look out for case selection discussions/justifications).

21 References Arend Lijphart, ‘Comparative Politics and the Comparative Method’, in American Political Science Review, Vol. 65, No 3, 1971. Jason Brownlee, Authoritarianism in an Age of Democratization (CUP, 2007)


Download ppt "Case Studies and Case Selection Research Methods Workshops Oisín Tansey."

Similar presentations


Ads by Google