The case selection is an important component in comparative politics, because selection bias associated with small N problem in qualitative research (argued by King, Keohane and Verba), as well as with selecting cases based on the dependent variable both in quantitative and qualitative researches leads to the difficulty of justifiable generalization and representativeness, overestimation or underestimation of conclusions or even inability to test and develop new theories in an appropriate way.
First, as we can conclude from reading written by Collier’s and Geddes’, selection bias can emerge 1) because of subjective selection of cases by researcher or 2) because of non-random selection of cases. For example, O’Donnell’s single-country study of Argentina, Skocpol’s analysis “States and Social Revolutions: France, Russia and China” or Asian "tigers" study proves the problem of representativeness of cases which were observed and the problem of selecting cases based on the dependent variable.…
