Abstract

This meta-analysis assesses the overall impact on undergraduate level science post-secondary student achievement outcomes of instructional environments that are more student-centered versus less student-centered (more teacher-centered). It also considers in which of four instructional events (dimensions) – Pacing, Teacher’s Role, Flexibility and Adaptation – the application of more student-centered pedagogy is more optimal for increasing student achievement outcomes, as well as considers the strength of student-centered pedagogy in each of these four instructional dimensions. Additionally, this meta-analysis considers the impact of a set of instructional and demographic moderator variables – technology use, subject matter, and treatment group class size – on student achievement. Out of an initial pool of 9759 abstracts, 96 full-text sources were chosen for analysis, yielding 141 independent effect sizes. The random effects model weighted average effect size was = 0.34, k = 141, SE = 0.04, z = 8.58, p < .001, suggesting that on average more student-centered classroom studies produce better results on achievement outcomes than do less student-centered classroom studies. However, the non-significant meta-regression result (p = 0.40) compromises the strength of this conclusion. Of the four instructional dimensions, based on simple meta-regression, only Flexibility produced a significant (negative) relationship ( = -0.09, p ≤ .05). Mixed moderator variable analysis yielded the subject matter of chemistry ( = 0.23, p = 0.03) as the best predictor of effect size; studies in which both groups used technology had a significantly lower average effect size ( = -0.18, p = 0.04) than the reference group of studies in which both groups did not use technology; and in studies in which the treatment group used technology and the control group did not, the result was not significantly different from studies in which both groups did not use technology ( = 0.07, p = 0.53). Recommendations include attending to more nuanced moderator variables when introducing student-centered strategies.