Effects and moderators of dialogic reading on children’s reading literacy: A three-level meta-analysis on studies from 2000 to 2025

Abstract

Dialogic reading (DR) is a well-established reading intervention that fosters children’s reading development; yet, its effects across multiple dimensions of reading literacy and the sources of variability in effect sizes have not been systematically evaluated. Based on a holistic reading literacy framework, this study synthesized post-Whitehurst DR interventions (2000–2025) on children’s reading literacy development from preschool to primary school using a three-level mixed-effects model. Pooling 260 effect sizes from 64 studies involving 10,463 typically developing children, the findings indicated that the model produced a significant overall effect on reading literacy (g = 0.76), with substantial between-study and within-study heterogeneity. The domain-specific models reported large effects on decoding (g = 1.10) and comprehension (g = 0.74), medium effects on response (g = 0.50), and large but inconclusive effects on motivation (g = 2.72). Moderator analyses identified age and measurement type as robust sources of variation, with smaller effects observed for standardized assessments and younger children. The study’s findings support the idea that DR has positive impacts on a broad bandwidth of reading literacy, while highlighting developmental readiness and assessment sensitivity as key considerations for research and practice.