Social Mechanisms and Thin Rationality. Towards Solving the Generalization Problem in Single Case Study Research
Bo Bengtsson and Nils Hertting
Uppsala University,
Department of Government and Institute for Housing and Urban Research
Abstract:
Drawing more general inferences on the basis of observations from single case and small n-studies is often seen as problematic in process tracing and qualitative social research more generally. How can we make more generalized claims based on findings in one single case? Several approaches have been suggested, e.g. related to concepts like Sartori’s ’ladder of abstraction’ or Yin’s ‘analytical generalization’. However, the more precise rationale behind claims of generalization of this type is not always clear. In consequence, we claim, ambitious qualitative studies of complex social and political processes are far too often modestly framed as ‘pilot studies’ generating hypotheses for subsequent and allegedly more conclusive large-n studies.
The paper suggests a logic of generalization based on thinly rational social mechanisms. Assuming that actors in most cases do things for a reason (‘thin rationality’), ideal-type social mechanisms can be derived from empirical observations of social action and interaction in one social context, which can be expected to have relevance in similar settings with similar actor constellations in another context. In this perspective generalization from single case studies is based neither on determinism nor on probability but on expectations about rationality in similar contexts.
In the paper the general logic behind ‘rationalistic generalization’ is presented and related to other ideas about generalization from single or small-n studies. In particular the concepts of ‘thin rationality’ and ‘social mechanism’ are explicated for the purpose of generalization and theoretical inference. The theoretical and methodological argument is illustrated with examples from empirical research by the authors and other social scientists.
