Conceptual Change Texts and Students’ Learning Outcomes: A Meta-Analysis of Experimental Studies

Abstract

Conceptual change texts (CCTs) are widely used in science education to address students’ misconceptions and promote meaningful learning. However, findings across individual studies vary, and the magnitude and consistency of their effects remain unclear. This meta-analysis synthesized 28 studies to examine the effectiveness of CCT-based interventions in improving learning outcomes compared with conventional or alternative instructional approaches. A random-effects model was used, with Hedges’ g as the standardized effect-size measure. Heterogeneity, prediction intervals, moderator effects, and funnel-plot asymmetry were also examined. The results showed a statistically significant positive overall effect of CCTs on learning outcomes, g = 0.98, SE = 0.24, 95% CI [0.49, 1.47], t(27) = 4.08, p < .001. Substantial between-study heterogeneity was observed, Qₑ(27) = 275.66, p < .001, I² = 95.63%, and τ² = 1.35. The 95% prediction interval ranged from −1.45 to 3.41, indicating considerable variation in effects across future contexts. Meta-regression showed that study year and research design were not significant moderators, whereas country was significant, F(4, 6) = 4.41, p = .05. Funnel-plot and asymmetry tests indicated potential small-study effects. The findings indicate that CCTs can improve learning outcomes, although their effects vary substantially across contexts. Science teachers should use CCTs as part of learner-centred conceptual change instruction, beginning with elicitation of students’ prior conceptions and explicitly addressing discrepancies between alternative and scientific explanations.