^aPolitical Science, Carl von Ossietzky University Oldenburg, Oldenburg D-26111, Germany

Correspondence: Hiltrud Nassmacher, E-mail: Karl.H.Nassmacher@uni-oldenburg.de

Abstract

Empirical comparisons of political systems can hardly avoid shortcomings. Many political scientists have turned to quantitative analyses of mass data, because case studies are assumed to have only explorative value. A variable-oriented design for comparisons with a small N would provide a basis for bridging the gap between the two empirical efforts. Until now, scholars who have dealt with this problem have not identified the key variables for the selection of cases, and for this reason the issue has never been discussed previously.