Practical Significance of Contextual Sensitivity of Goal Orientations in Youth Sport

Abstract

In goal perspective research, questions have been raised about the stability of goal orientations over time and across different situations since the introduction of the social cognitive approach to achievement motivation. PURPOSE: To examine the sensitivity of goal orientations to different structured sport contexts. METHODS: A sample of 206 male and female high school athletes completed the Task and Ego Orientation in Sport Questionnaire (TEOSQ), with the stem modified to reflect different sport contexts – Physical Education (PE), structured training and practice (TP), and competition (COMP). Validity (and factorial invariance by gender) of the modified instrument was assessed using confirmatory factor analysis for each of the three sport context versions. A combination rule was used to evaluate the adequacy of model fit (< 0.11 for the Standardized Root Mean Squared Residual combined with a value of <0.06 for Root Mean Squared Error of Approximation). As statistically significant differences may be trivial in magnitude, we calculated the probability (percent chances) that the true, population between-context differences in achievement orientations exceeded a pre-defined 'practically significant' effect size (0.2 between-subject standard deviations). These probabilities were calculated using the t distribution, with a modified numerator for the conventional t statistic (observed difference minus the practically significant effect size). All probability values were interpreted using a scale of qualitative descriptors, ranging from almost certainly not to almost certainly practically significant. RESULTS: The combination rule criterion was met for all three versions of the modified instrument, signifying adequate validity. Coefficient alpha for the TEOSQ sub-scales ranged from 0.8 to 0.9 for the different context versions, indicating acceptable internal consistency reliability. For ego orientation, the PE vs. COMP comparison was very likely to be (boys; 97% chance) and likely to be (girls; 91% chance) practically significant. Statistically significant differences for task orientation were observed for PE vs. TP and PE vs. COMP for boys (possibly not and unlikely to be practically significant, respectively), and for PE vs. TP for girls (possibly NOT practically significant). Similarly, for ego-orientation in girls, the statistically significant TP vs. COMP comparison may not be practically significant. CONCLUSIONS: Clearly, practical and statistical significance must be differentiated in interpretation of results. We conclude that task orientation appears relatively stable, whereas ego orientation is contextually sensitive.