When is a bologna sandwich better than sex? A defense of small-n case study evaluations

This paper discusses the value of quasi-experiments with small numbers of cases - small-n case studies. It argues that these evaluations of crime prevention have value. The first part of the paper shows that small-n evaluations are practical methods for producing information and that discarding this information reduces our ability to learn what works in crime prevention. The solution is to examine small-n evaluations in systematic reviews. In the second part this paper shows that more rigorous evaluations can sacrifice generalizability to gain internal validity and that this, too, reduces our ability to learn. The solution is to create a mixed portfolio of intrusive, less-intrusive, and non-intrusive evaluations. In the third part this paper shows why inference processes that make use of theory should replace naïve inductionism. One implication of this is that ignoring theory also discards evidence. Another implication is that taking theory into account when we review program effectiveness allows us to better use small-n evaluations when we generalize. In the final part, this paper gives three recommendations for making use of small-n evaluations: adopt case study procedures developed in political science, improve the value of process evaluations, and improve crime theories.