Confounds and Objectives in Economic Experiments

Abstract

The key general problem in design, evaluating and defending an experimental design is to determine whether there are factors affecting experimental behavior that are not being duly taken into account. We present a relevance, distinctiveness and plausibility (RDP) framework for systematically evaluating potential experimental confounds, and present examples. Relevance relates to the functional relationship between potential confounds and objectives. Distinctiveness is about whether potential confounds and objectives relate essentially to the same thing. Plausibility can be based on direct or indirect evidence or, failing those, to a more global judgment based on all available knowledge at a point in time.

Keywords: confounds, methodology of experimental economics, experimental objectives, relevance, distinctiveness, plausibility

JEL Classification: B41, C90