ABSTRACT
Optimally, a variety of nontrivial cases not in a Very-Small-N nonrandom sample will share important characteristics with one or more cases in the study. (Chapter 9 will return to this point.) Sensitivity to these possibilities during case selection can increase the "leverage" (King, Keohane and Verba 1994:29-31) we get from our research. In other words, the explanatory power of our inferences increases. In contradistinction, a very high degree of unit homogeneity-desirable in Large-N studies-may severely limit the applicability of inferences made from a Very-Small-N study focused on continuity and change in foreign policy decision making activity in one or two countries.
