Abstract
The purpose of this study was to test alternative models of sport commitment with a youth sport sample. Competitive female gymnasts (N = 792; Levels 4–10), ages 8–19 years (M = 11.68, SD = 2.33), completed measures designed to assess the determinants of commitment. Data were analysed via path analysis models in the lavaan package in R. Three separate models were fitted: (a) direct effects model, (b) a mediational model testing the indirect effects of the predictors via enjoyment, and (c) direct and indirect effects of the predictors on sport commitment. Of the models tested, the direct and indirect effects model yielded the best fit of the data (RMSE = .03 90% CI [.000, .057], TLI = .989, CFI = .997, NFI = .989; 67% of the variance in commitment explained). The model revealed that enjoyment, involvement opportunities, investments, and attractive alternatives had a direct effect on commitment. With regard to the indirect effects of the determinants on commitment via enjoyment, attractive alternatives, involvement opportunities, investments, costs, parent and best friend social constraints, and coach and teammate social support had direct effects on enjoyment, which predicted the level of gymnastics commitment. Additionally, some differences emerged in the salient predictors of commitment for high versus low level competitors. The role of enjoyment within commitment is complex depending on competitive level, and future research should consider alternative models when examining the role of social influence on sport commitment.
KEYWORDS: