Does Self-Control Training Improve Self-Control? A Meta-Analysis.
Level 1 - systematic review of randomized trials
Meta-analysis of published and unpublished intervention studies
PubMed 28846503 · doi:10.1177/1745691617697076
What was done
The authors conducted a random-effects meta-analysis using robust variance estimation on published and unpublished studies assessing the effects of self-control training (repeatedly overriding dominant responses). Moderator analyses and bias-correction techniques were applied.
What was found
Across 33 studies and 158 effect sizes, self-control training showed a small-to-medium effect of g = 0.30, 95% CI [0.17, 0.42]. Training effects were larger for stamina than strength, for inactive compared to active controls, for males than females, and when proponents of the strength model were (co)authors. Bias-correction techniques for small-study effects and/or publication bias yielded lower estimated effects ranging from g = 0.13 to 0.24.
Why it matters
This meta-analysis demonstrates that apparent benefits of self-control training are modest and likely inflated by study artifacts, author allegiance, and publication bias.
Limits
The underlying evidence base is affected by small-study effects, publication bias, and moderator effects such as inactive control comparisons. Mechanisms remain poorly understood, with insufficient evidence that overriding dominant responses is the critical active element.