The purpose of this research was to develop a Chinese language scale to measure children and adolescents' expectancy and value beliefs in sport. According to expectancy-value theory (Eccles et al., 1983; Eccles & Wigfield, 1995; Fredricks & Eccles, 2002), individuals' expectancy and value beliefs for an achievement-related activity influence their performance and task choice in the activity. Therefore, examining expectancy and value beliefs of an individual should help to have a valid prediction of his/her achievement-related behaviors, as well as an instructive description of related motivational processes. This research consisted of two studies. In the first study, items from Eccles and Wigfield's measurement were translated into Chinese. New items developed based on the concepts proposed in expectancy-value theory were also added to develop a 32-item scale. Seven hundred and three participants, aged from 9-18 years, completed the scale. A series of item analyses and exploratory factor analysis were conducted. Seven items were eliminated in the analyses and 25 items remained. Factor analysis extracted five reliable factors, including "ability expectancy" (8 items, Cronbach's alpha= .90), "required effort" (4 items, alpha= .82), "psychological cost of failure" (4 items, a= .81), "expectancy cost from significant others" (3 items, alpha= .63), and "task value" (6 items, alpha= .81). In the second study, a new sample with 563 participants, aged from 9-19 years, completed the 25-item scale. Cross-validation procedures were conducted first by a factor analysis with a specified five-factor solution. A result showed that eigenvalues of all five factors were greater than 1 and the allocation of items to the five factors was identical to that of the first study. A confirmatory factor analysis performed by structural equation modeling additionally confirmed the structure of the scale, chi(2)=834.35, chi(2)/df=3.14, GFI=.89, NFI=.95, NNFI=.96, CFI=.97, RFI=.95, RMSEA=.06, CN=217. Predictive validity tests with the amount of physical activity as a criterion, and a known-group-difference-method with gender difference as a criterion were also conducted to examine the construct validity of the scale. Results of a standard multiple regression showed that scores of the scale could predict the amount of physical activity, R=.38, F=18.39, p<. 05, R-2 =.14. "Ability expectancy," "required effort," and "task value" subscales contributed significantly to the prediction (all p < .05), but not "psychological cost of failure" and "expectancy cost from significant others." In addition, results of MANOVA indicated significant gender differences on the scores of the scale, F(5,550)= 9.18, Wilks' Lambda= .92, p<.01. Follow-up univariate tests showed that male and female participants were significantly different in all five subscales (all p < .05). Males had significantly higher scores than females in "ability expectancy" and "task value" sub scales than females, while females scored significantly higher than males in "required effort," "psychological cost of failure," and "expectancy cost from significant others" subscales. Finally, a 25-item Chinese version of "Expectancy and Value Beliefs Scale in Sport for Children and Adolescents" was developed and further examination of its validity was certainly warranted.
