Identifying context and cause in small-N settings: A Comparative Multilevel Analysis

Abstract

Qualitative small-N comparisons face the challenge to detect context-bound causality under conditions of limited empirical diversity. Rather than treating context as a causal factor, we test the usefulness of the novel method of comparative multilevel analysis (CMA) to identify and understand the role of context as a contingent necessary condition that enables a causal relationship to unfold. Combining CMA with pairwise comparisons, we assess how organ donation policies in Switzerland and Spain affect relatives’ refusal rates in a small-N setting exhibiting multiple contextual levels. To tackle limited diversity systematically, we suggest to refine the CMA methodology by accounting for several contexts and referring to higher-order constructs. Applying CMA with these refinements, we find voluntary information measures only affect refusal rates in contexts of a credible state explicitly supporting organ donation. The fact that CMA can easily be combined with other analytical and conceptual approaches makes it an effective technique to identify contextual effects in small-N research.