The Potential of Statistical Methods in “Small N” and “Medium N” Situations

Abstract

Mainstream statistical tools encounter some problems in small N and medium N situations. Can some of these mainstream tools (such as log-linear analysis, classical correlation techniques, etc.) or other more specific “robust statistics” tools, be useful in medium N social science research ? Can these tools be usefully combined with specific small N techniques such as QCA ? Each paper treats some of these issues on the basis of concrete applications.

Title	Details
QCA and correlational methods	View Paper Details
Two macro-level research designs applied to the same set : invalid ballots in Brazil and in Belgium studied through ecological approach and QCA	View Paper Details
Differential registration and electoral participation between French people of foreign origin in France and “French born people” : methodological issue in using longitudinal data and log-linear models	View Paper Details
Is pooling data a reliable method of empirical argumentation relating to African development ?	View Paper Details