Cases, case populations and generalization

In a comparative case study method like QCA, cases must be comparable. One cannot learn from comparing apples and oranges. Consequently, researchers must homogenize their case population on key scope conditions. Given populations are too heterogeneous, and samples reproduce this heterogeneity. Therefore, QCA researchers must select a homogeneous case population on the basis of a careful definition of what is a case (i.e., the process of casing). This makes generalization analytical rather than empirical. QCA generalizes the substantive explanation (to similar cases), rather than the empirical pattern from which it is interpreted. If one has carefully defined one’s cases and selected them accordingly, the solution one finds is always the correct solution for that case population. Changing the case population, one expects to find a different solution, which makes QCA’s case sensitivity a strength rather than a weakness.