Testing reciprocal causal relations among expectancy, value and academic achievement of early adolescents: A longitudinal study.

This dissertation examines the relative contribution of students' expectancy or self-concept of ability (SCA) and value for a specific task to the prediction of their actual academic achievement (Eccles et al., 1983). Given that few studies have systematically tested the reciprocal causal relations among expectancy, value, and academic achievement, this study addresses this gap in the literature with a longitudinal design. Based on the structural equation modeling of 4-semester, longitudinal data involving 826 sixth graders (mean age of 11.6) who experienced the junior high school transition, this study tested the reciprocal causal relations among SCA, intrinsic value (i.e., interest and liking), and mathematics performance of early adolescents. Generally, the results showed some significant, reciprocal causal relations between academic beliefs and achievement. Furthermore, these reciprocal causal relations were somewhat moderated by both the gender of adolescents and the school/classroom context. For example, contrary to common beliefs, positive SCA alone did not always have a positive influence on subsequent academic performance and value when other factors were controlled for. While positive SCA relative to actual achievement level had a slightly enhancing effect on subsequent performance for males, such a causal effect did not exist for females. If there were any SCA effect at all, females' self-concept beliefs seemed to be negatively related to their performance. Unlike before or after the transition to junior high school, SCA had a significantly negative impact on subsequent performance during the transition for both females and males. In addition, the causal effect of performance feedback on students' SCA and value increased in the context of the junior high school. Finally, in contrast to the mixed relations between SCA and performance, intrinsic value and performance tended to influence each other in a consistently positive manner. Results are discussed in the light of cultural and contextual perspectives. Implications for education are also discussed.PhDEducationEducational psychologyPsychologyQuantitative psychologySocial psychologyUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/129913/2/9635640.pd