A pre-registered multi-lab replication of ego-depletion: implications for the replication “crisis” and health psychology

Abstract

Background. Self-control has been conceptualized as a limited resource with the acute state of self-control failure termed ego-depletion. A substantive body of research, including meta-analyses has supported the ego-depletion effect in multiple contexts, including health behaviour. However, subsequent analyses has raised questions over the strength of the effect, with some estimates suggesting it is no different from zero. In response, I coordinated a pre-registered replication of an ego-depletion experiment will collaborators in multiple labs (N= 23) and meta-analyzed the resulting data set (N=2,141). The analysis indicated an effect that was no different from zero for multiple outcome variables. Methods. I will present a narrative review and interpretation of the implications and impact of the replication project. I will summarize the process of the replication project from inception to dissemination. I will outline the impact of the project including responses from the media and scientific community, and its broader implications for the ‘replication crisis’. Findings. The replication has catalyzed subsequent reanalyses that have further extended its findings including the application of mediation analyses on the role of effort and statistical tests to detect bias. The replication has influenced theory development and further research. However, it also has limitations and the bounds of its contribution should be recognized. Discussion. I will provide recommendations to guide future replications including identification of an effect of interest, consultation with original study authors, pre-registration, coordinating labs, analysing results, and disseminating findings. I outline how replications may be conducted in health contexts and what kind are needed.