Search

This paper describes the novel use of parallel student teams from a research methods course to perform a replication study, and suggests that this approach offers pedagogical benefits for both students and teachers, as well as potentially contributing to a resolution of the replication crisis in psychology today. Four teams, of five undergraduates each, independently attempted exact replications of Study 8 by Gailliot et al. (2007), which reported that participants' self-control is enhanced by consuming a glucose drink. In a 2 × 2 independent groups design, participants (N = 306) first consumed a glucose drink or a placebo, and then wrote about death, intended to deplete their self-control, or dental pain as a control condition. Absolute levels of self-control were lower here than in the target article (shown by more items left unsolved in a word puzzle), but its main result was replicated, since self-control overall was raised by the glucose drink. Also, the teams reliably reported similar effects for the experimental treatments (ICC = 0.928). Two differences from the target study results were noted: the glucose effect occurred only with female participants, and no effect was found from the writing scenario used.

It is suggested that replication projects may be valuable in teaching research methods, and also address the current need in psychology for more independent verification of published studies. Their use in an undergraduate methods course is described, involving student teams who performed direct replications of four well-known experiments, yielding results which were subsequently published online. Illustrative data are given for the one successful replication and three failures obtained, and practical suggestions are given for incorporating replication projects into a methods course as an alternative to the usual term project. It is also noted that the published success rates of replication attempts appear to be higher for those studies that were performed as class projects.