CONFIRMING A PERFORMANCE-BASED INSTRUCTIONAL-MODEL

This study sought to test alternative performance-based instructional models of growth in students' academic achievement in the elementary s chool. Based upon constructs derived from a synthesis of prior experim ental and descriptive studies, the models tested viewed growth in acad emic achievement as a function of exposure to academic subject matter, task quality, and student engagement. Indices for these constructs we re derived from repeated measures of classroom instruction and standar dized achievement tests in Grades 2 and 3. Results indicated that the best fit was achieved by a structural model that included student enga gement as a construct mediating the effects of instruction on academic achievement as compared to that obtained under alternative direct inf luence and nonmediated models. the implications of the current finding s are discussed.