Résumé
NTRODUCTIONTwo studies were conducted among elite young judokas to examine (a) whether those who persistedin national training centers were different from dropouts in their perceptions of coach-, parent-, and peer-induced motivational climates,goal orientations, self perceptions, perceived competence, and intention of dropping out, and (b) whether these variables variedduring the persisting athletes' first two years in these centers. STUDY 1METHODPARTICIPANTS.The participants (Mage = 17.9 ± 1.3) were selected from two cohorts of judokas who had just enteredone of France's national training centers at the beginning of the study. The selected sample included 52 dropouts (30 males, 22females) who voluntarily gave up within their first two years in their training center and 52 persisting judokas (34 males, 18females) who were randomly selected among athletes who did not give up. PROCEDURE.The participants completed the following set of questionnaires six times during a period of two years:– Le Bars et al.'s (2006) Significant Others' Goal-Involving Roles in Sport Questionnaire that measures athletes'perceptions of coaches-, athletes-, and parents-induced task- and ego-involving climates.– Durand et al.'s (1996) Questionnaire de Perception du Succès en Sport that measures task and ego orientations in sport.–Ninot et al.'s (2000) Inventaire du Soi Physique that measures the perceptions of the self in the physical domain.– Two items of perceived competence in judo.– Two items of Intention of Dropping Out of high level judo.For dropouts, the last measures that were recorded before they gave up were considered. For persisting athletes, thetimes of measurement that were retained were yoked to dropouts'.RESULTSThe 2 x 2 multivariate and univariate analyses of variance (Sex ? Participation: Dropout vs. Persisting)revealed a significant main effect of participation on all the significant others-induced task-involving climates, and athletes' taskorientation and intention of dropping out (6.93 < F(1, 100) < 19.01, p < .01). Compared to persisting athletes, dropouts perceivedthe roles of the significant others as less task-involving, were less task-oriented, and intended more to drop out. No significantdifferences were observed regarding ego orientation and perceptions of significant others-induced ego-involving climates. Asignificant main effect of sex was found on self perceptions which were higher for males than for females (4.26 < F(1, 100) <12.91, p < .01).STUDY 2METHODPARTICIPANTS. The participants (n = 82; Mage = 16.7 ± 1.1) came from the initial two cohorts of judokas that servedas a basic pool for the first study. Only persisting athletes who had answered all the questionnaires at the six times of measurementwere retained for this study.PROCEDURE.The conditions and the procedure of investigation of this study were exactly those of the first one and thesame questionnaires were used.RESULTSThe 2 x 6 multivariate and univariate analyses of variance (Sex x Time) with repeated measures for Time revealed asignificant main effect of time on all the significant others-induced task-involving climates, coach-induced ego-involving climate, intentionof dropping out, and self perceptions (2.33 < F(5, 400) < 9.51, p < .05). Athletes'' perceptions of coach-, parent-, and peer-induced task-involving climates decreased (p < .05), while perceptions of a coach-induced ego-involving climate and the intention of dropping outincreased in spite of more positive self perceptions (p < .05). Athletes' task (M = 4.44) and ego (M = 3.49) orientations were high and didnot significantly change over the two years, thus being impervious to the variations of perceived motivational climates. Gender differencesfavoring males were observed for self perceptions only (9.27 < F(1, 80) < 21.99, p < .01).DISCUSSIONThe main distinction between persisting athletes and dropouts was that the latter perceived the roles ofcoaches, parents,and peers as less task-involving and were less task-oriented. The perceptions of ego-involving climates as wellas ego-orientation did not differ between dropouts and persisting athletes. These findings corroborate achievement goal theorists'conclusions about the necessity of promoting self-referenced criteria of success in sport (e.g., Roberts et al., 2007). They also showthat, although highly ego-involving, the context of high level competition is not detrimental for athletes' motivation as long astask-involving climates are concomitantly sustained. However, persisting athletes' perceptions of coach-, parent-, and peer-induced task-involving climates almost continuously decreased during the first two years spent in elite training centers. During thesame period, the perceptions of coach-induced ego-involving climate and the intention of dropping out increased in spite ofathletes' self-concept increases. These last findings bring an additional emphasis on the need for all the agents of athletes' socialenvironment to promote task-involving climates, and all the more in elite competition contexts which seem naturally prone todegrade such climates.