Selecting Units of Analysis in Qualitative Research

• In Event: Qualitative Research: Rethinking Boundaries and Expanding Frontiers

Abstract

This chapter of a book project on the modifiable areal unit problem (MAUP) and the study of spatial aggregate units explores how the MAUP affects core inferential tasks of small-N research. I first explore its impact on research that begins with observed variation in an outcome of interest and seeks to explain its causes. Second, I turn to case selection. Third, I show that the MAUP also affects a prior step in the research process--the description of the overall population. Fourth, I explore whether inferences in within-case analysis are also vulnerable to the MAUP. Finally, I show that matching techniques in large-N research are also vulnerable to the MAUP in ways parallel to those I have identified for small-N research. The chapter uses literature on the causes of religious riots in India as an exemplar to guide the discussion.

Author

• Hillel David Soifer, Temple University