A meta-analysis of teacher and student-centered practices and processes in undergraduate science education

Abstract

This meta-analysis investigates the effects of four instructional dimensions rated on a scale from more Teacher-centered (T-C) to more Student-centered (S-C) plus several coded moderator variables on the achievement of undergraduate students in science education courses. More student-centered conditions served as the ‘treatment’ while more teacher-centered conditions were considered the ‘control.’ Hedges’ g, operationalized as the adjusted standardized differences between treatment and control means, served as the outcome measure. The weighted average difference between groups was ḡ = 0.34, k = 140 (random effects analysis), indicating an overall difference in favor of student-centered instruction. Out of four rated dimensions (Pacing, Teacher’s Role, Flexibility, and Adaptation) only Flexibility was significant in meta-regression as a negative predictor of effect size. Two demographic variables (i.e., class size & subject matter), and one instructional moderator variables (i.e., technology use) were also significant when added to Flexibility, producing a model that accounted for 36% of total variation in effect size.