Why Can’t a Student Be More Like an Average Person?: Sampling and Attrition Effects in Social Science Field and Laboratory Experiments

Abstract

In the social sciences, the use of experimental research has expanded greatly in recent years. For various reasons, most experiments rely on convenience samples of undergraduate university students. This practice, however, might endanger the validity of experimental findings, as we can assume that students will react differently to experimental conditions than the general population. We therefore urge experimental researchers to broaden their pool of participants, despite the obvious practical difficulties this might entail with regard to recruitment and motivation of the participants. We report on an experiment comparing the reactions of student and non-student participants, showing clear and significant differences. A related problem is that differential attrition rates might endanger the effects found in long-term research. We argue that experimental researchers should pay more attention to the characteristics of participants in their experimental design.