The relationship between self-efficacy and performance exclusively within the sports environment is yet to be quantified. Hence, we meta-analysed this relationship by following the PRISMA guidelines. Two previous meta-analyses, five relevant databases, and Google Scholar were searched. Forty-four articles published between 1983 and 2021 met the inclusion criteria, with 55 independent samples. Comprehensive meta-analysis software version 4 was used for all meta-analytic calculations using a random-effects model to calculate the mean effect size, and a mixed-effects model was used for moderation analyses. The mean pre-event self-efficacy and performance effect size was r = 0.31 (95% CI 0.22, 0.40). For moderation analyses, notable mean differences (p values ≥ 0.08) resulted for concordance [concordant (r = 0.37), nonconcordant (r = 0.22)], sports skill [closed (r = 0.37), open (r = 0.23)], and athlete level [elite (r = 0.40), sub-elite (r = 0.28)]. The true effect prediction interval ranged from negative (i.e., self-efficacy impairing performance) to positive (self-efficacy improving performance) for all moderator variables except self-referenced vs. other-referenced performance. In conclusion, the relationship between pre-event self-efficacy and performance is positive and moderate in magnitude, although with prediction intervals ranging from debilitating to facilitating performance. Researchers and practitioners should note that high athlete-rated self-efficacy might not always improve impending competitive sports performance.
Keywords: athletes; competitive sports; confidence; quantitative review.