Common outcomes: conceptual and analytic issues

The challenges of studying rare events such as diseases affecting just a minority of people have received some attention in the literature. By contrast, the opposite of rare events, i.e. common events or experiences, have received less attention, perhaps because it is harder to develop a theoretical framework for such events. Nevertheless, some research questions of interest do concern common events and therefore researchers ought to pay attention to analytic challenges arising from skewed outcome distributions and asymmetry. This paper draws on Qualitative Comparative Analysis (QCA), a method well suited to analysing asymmetry, new goodness-of-fit indicators developed by Luna De Souter, and classification indices. The conventional consistency index used in QCA to assess how close a condition is to being sufficient for some outcome is of limited use for causal analysis if the outcome is very common because if an outcome is common, then it will tend also to be common for cases characterised by a wide range of conditions, and consistency for these conditions will be high accordingly, whether or not a causal link exists between them and the outcome. Using numerical examples and an empirical illustration drawing on the German ALLBUS data, the paper demonstrates that especially in the case of skewed outcome distributions additional indicators such as those developed by De Souter in conjunction with theoretical considerations will be helpful for the interpretation of findings. The paper focuses on high prevalence, but De Souter’s indicators would also be instructive for the other form of skewness, rare outcomes.