From the ‘Frying-Pan’ into the ‘Fire’? Antidotes to Confirmatory Bias in Case-Study Research

Abstract

This paper sets out three antidotes against confirmatory bias in case-study research. We take a step back, as it were, from the generalizable-or-not debate and consider a prior issue: the local accuracy of the findings. The antidotes are method free and “substance free” (Schum, 2001), that is to say, they transcend the differences in types of research methods and evidentiary data. The properties of these antidotes/tests are described and located in both accounting and a wider literature. The characteristics of each antidote are further considered, and the adverse consequences of the failure to apply them are illustrated through analysis of a number of case-studies by leading accounting and management writers. The unpacking of these cases also shows how engagement with case-study situations through the three processes opens up the possibility of richer and more valid research using this methodology.