David Lopez, PhD Candidate and National Science Foundation Graduate Fellow in political science at UW, shared with the QUAL Community two mixed method case selection strategies from his dissertation work. Lopez studies the political foundations of national education systems in developing democracies. He grounds his work in the idea of infrastructural power as described by Jackson School Professor Joel Migdal.
When it came to selecting the cases for his mixed method research, Lopez relied on John Gerring’s case study definition. Lopez recommended three considerations when planning which cases (in his dissertation work, which countries) to study in a given case study research project:
Scope conditions and purpose of the research
Lopez argued for purposeful case selection rather than a randomized draw of countries from a region of study. For his dissertation prospectus, the scope conditions he considered were – the presence of a national education system, permanent state institutions, and the level of economic and political development in a given country. He then created a table with countries within and outside of this scope. The purpose of your study, the variation you are trying to explain, will shape your case selection but also your research design overall, Lopez said.
Data availability and feasibility
Data availability and access to it are next on the considerations list. Time and funding for a project such as a dissertation are very good disciplinary measures to help you progress your research, he acknowledged.
Multiple approaches to case selection
Lopez described two case selection strategies he explored and discussed a third one he examined after his cases were chosen.
Nested Analysis is a systematic way of selecting cases, especially well suited when there is a lot of quantitative data available. Lopez shared a diagram of choices (see slide 21) a researcher could follow after examining quantitative data for a given phenomenon. The limitations of this approach to Lopez’s own research were that he had a medium-N – several dozen countries to choose from – and his outcome of interest is a causal process which doesn’t necessarily show up in quantitative data, he explained.
Lopez then explored the Millian Methods for Cross-Case Research (from John Gerring’s “Case Study Research” book) approach to case selection, suitable for both hypothesis testing and for generating hypotheses. Using this approach, Lopez decided on studying post-independence Chile and Argentina. He used Comparative Historical Analysis as his method and dove in to get some preliminary findings from the two cases.
“Using qualitative evidence I was able to control for international factors of influence,” he said. He controlled for the influence of France, the United States and Germany in the two countries of interest. Lopez is thus far finding, in his early study of Chile and Argentina, that they conform with his theory. But he readily admits there doesn’t appear to be a perfect case selection strategy, and he is still considering adding or subtracting cases.
Looking back on his case selection process, Lopez realized he ended up doing something similar but inverse to Nested Analysis. And thus, after the fact, he discovered the work of Ingo Rohlfing (see slide 29).
For summaries of each of the three case selection methods Lopez discussed, see the slides from his presentation.