What Can We Know about the Effectiveness of Coaching? A Meta-Analysis Based Only on Randomized Controlled Trials

The study involved a comprehensive meta-analysis of 37 randomized controlled trial (RCT) studies of workplace and executive coaching programs written in the English language between 1994 and 2021, using rigorous inclusion criteria designed to align with the robustness of statistical significance and coaching applications. Analysis of 39 coaching RCT samples, with a total sample size of n = 2,528, yielded a statistically significant effect of workplace coaching across all leadership and personal outcomes. Our best estimate of a standard effect size for coaching of g = .59 fell well within the moderate range, although there were indications of significant publication bias, as expected. Furthermore, we develop a model that is based on the leverage from coregulation between coach and coachee. The model was confirmed in five different predictions: moderation analyses indicated that effects are larger with self-reported outcomes, as distinct from observed outcomes; they are larger with qualified coaches in nonleadership applications; and they are somewhat larger with female coachees, while they do not depend much on the length of the overall coaching assignment. Taken together, these findings provide clear and new evidence of the efficacy of coaching interventions in a variety of applications.