Quantifying the Extent to Which Junior Performance Predicts Senior Performance in Olympic Sports: A Systematic Review and Meta-analysis

Abstract

Abstract Background To what extent does junior athletic performance predict senior athletic performance (i.e., in the highest, open-age category)? This question is the subject of a lively debate in the literature. Following traditional theories of giftedness and expertise, some researchers and practitioners have proposed that a high level of junior performance is a prerequisite for the development of a high level of later senior performance. Sceptics of this view hold that junior performance has limited predictive value for later senior performance, pointing to empirical evidence indicating that predictors (e.g., participation patterns) of junior performance and of senior performance differ. The straightforward way to resolve this controversy empirically is to test the correlation between junior and senior performance. Objective To provide robust and generalizable evidence on this issue, we performed a systematic review and meta-analysis of relevant studies. The aim was to quantify the overall correlation between junior and senior performance and then test whether correlations vary across junior age categories and subsamples (e.g., types of sports). Methods A systematic literature search was conducted in SPORTDiscus, Eric, ProQuest, PsychInfo, PubMed, Scopus, WorldCat, and Google Scholar from 27 January through 30 April 2022. We searched for original studies that recorded athletes’ junior and senior performance longitudinally and included measures of association between junior and senior performance. Quality of evidence was evaluated using the Mixed Methods Appraisal Tool version for nonrandomized studies. Results The search yielded k = 129 effect sizes from N = 13,392 athletes from a wide range of Olympic sports, 62% male and 38% female, from 2006 to 2021. Four central findings emerged: (1) Overall, the meta-analytic pooled correlation between junior and senior performance was $$\overline{r }$$ r ¯ = 0.148. That is, junior performance explained only 2.2% of the reliable variance in senior performance. (2) The finding was robust across types of sports, sexes, wider or narrower performance ranges, national or international samples, and binary or continuous performance measures. (3) Effects varied across junior age categories: the younger the junior age category, the lower the correlation between junior and senior performance, with $$\overline{{r} }$$ r ¯ ranging from $$\overline{{r} }$$ r ¯ = − 0.052 to $$\overline{{r} }$$ r ¯ = 0.215. That is, across junior age categories, junior performance explained 0–4.6% of the reliable variance in senior performance. (4) The quality of primary studies was high. Discussion The results suggest that junior performance has very little, if any, predictive value for senior performance. The findings run counter to claims from traditional theories of both giftedness and expertise. From an applied perspective, talent selection typically begins around puberty or younger—age ranges where youth performance is uncorrelated or negatively correlated with later senior performance. The evidence presented here raises serious questions about the use of junior performance for talent selection purposes. A PRISMA-P protocol was registered at https://osf.io/gck4a/.