The sport model : commitment and outcome behaviours of age-group triathletes

The purpose of this study was to understand the factors that predict commitment for adults in the sport of triathlon using a conceptual model of commitment. Secondary purposes were to examine whether levels of commitment predict training behaviours (number of hours of practice per week) and to evaluate and compare two variations of the Sport Commitment Model. One hundred and forty-four (males: n-69; females: n=75) age group triathletes were recruited at local triathlon races, expos, and club practices and volunteered to participate in the study. Participants received pre-stamped and pre-addressed envelopes with the questionnaire pack to fill out at their convenience. Two statistical models were tested, a direct effects model and a mediational model (Weiss et al., 2001). The direct effect model predicted that sport commitment would be positively correlated to enjoyment, social support, personal investments, involvement opportunities, and negatively correlated to involvement alternatives and social constraints. The mediational model predicted enjoyment to mediate the effects of the other five variables on sport commitment. Sport commitment was also examined as a positive predictor of actual behaviour. Regression analyses did not provide support for the mediational model with the total sample or in separate models for females and males. Support for the direct effects model was observed. With the total sample, the results showed that sport enjoyment (β =.16), personal investments (β =.27), involvement opportunities (β =.26) and alternative (β =-.38) were significant (p < .05) independent predictors of sport commitment, accounting for 57% of the variance. Gender differences emerged from the analyses. For males, enjoyment (β =.25), opportunities (β =.24), attractive alternatives (β =-.27) and investments (β =.28) predicted 56% of sport commitment variance. For females, only opportunities (P =.31), attractive alternatives (β =-.48,) and investments (β =.23) were significant predictors predicting 62% of sport commitment variance. For males (β =.64) and females (β =.46), investment was the only predictor of training behaviour. Sport commitment was a weak significant predictor of training behaviour in all models except for the non-significance found for the female only model. Important findings include the lack of evidence for a mediational model, evidence for the direct effect model for key predictors, evidence for gender differences in the model, and a weak link between commitment and training behaviour.Education, Faculty ofKinesiology, School ofGraduat