Case studies of social interventions need not exchange relevance for rigor, provided analysts keep three methodological issues in mind. First, certain critical tests can and should be performed to help establish causal relations even when empirical data are limited. Second, "causell is an...