Introduction
Narrative ability—the capacity to sequence events, maintain thematic coherence, and express complex ideas—is a fundamental pillar of early childhood development. Beyond its immediate role in communication, narrative proficiency serves as a robust predictor of later academic success, particularly in literacy and reading comprehension. One of the most widely advocated pedagogical strategies for fostering these skills is "interactive reading," an umbrella term encompassing techniques such as dialogic reading and shared book reading. While the benefits of these interventions are generally accepted, the specific factors that optimize their effectiveness—such as the child's age, the role of the reading partner, and the duration of the intervention—have remained subject to debate.
Research Objectives
This meta-analysis seeks to synthesize fragmented literature to provide a quantitative assessment of interactive reading's impact on children aged 3 to 6. Specifically, the study addresses three critical questions:
- To what extent does interactive reading improve narrative abilities overall?
- How do moderator variables, such as the type of interactive partner (parent, teacher, or peer) and the child's age, influence outcomes?
- What is the relationship between intervention "dosage" (duration) and the magnitude of the effect?
Methodology
The researchers conducted a systematic review following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. A comprehensive search was performed across both Chinese (CNKI) and English-language databases (including Web of Science, ScienceDirect, and ERIC). From an initial pool of 2,326 records, 25 high-quality studies were selected based on strict inclusion criteria: the use of experimental or quasi-experimental designs, subjects aged 3–6 without developmental disabilities, and the reporting of sufficient data to calculate effect sizes.
The meta-analysis utilized Hedges’ $g$ as the primary measure of effect size, which is particularly suited for studies with varying sample sizes. A random-effects model was employed via Stata 17.0 to account for potential heterogeneity between studies. To ensure the robustness of the findings, the authors conducted sensitivity analyses and tested for publication bias using Egger’s linear regression and funnel plot inspections.
Key Insights and Results
Overall Impact
The synthesis of 123 distinct effect sizes revealed that interactive reading has a moderate and significant overall effect ($g = 0.425$) on the narrative development of young children. This confirms that structured interaction during reading is substantially more effective than passive listening.
The "Peer Power" Effect
One of the most significant findings of this study concerns the identity of the interactive partner. Contrary to traditional pedagogical assumptions that prioritize adult guidance, peer-led interaction yielded the highest effect size ($g = 0.675$). In comparison, parental engagement showed a strong effect ($g = 0.597$), while researcher-led interventions were moderate ($g = 0.459$). Notably, teacher-guided interventions showed the lowest impact ($g = 0.164$).
This suggests that horizontal, less-structured dialogues between children may encourage more narrative experimentation and risk-taking than the vertical, often evaluative, nature of adult-child interactions. In peer settings, children may feel more empowered to negotiate meaning and co-construct stories without the pressure of "correctness" often associated with a teacher's presence.
The Developmental "Sweet Spot"
The analysis identified a critical window for maximum impact between the ages of 4 and 5 years ($g = 0.635$). While children across the 3–6 age range benefited, this specific cohort appears to possess a sufficient linguistic foundation to engage deeply with interactive prompts while maintaining the high levels of neuroplasticity necessary for rapid skill acquisition.
Dosage and Sustainability
The data highlights a clear "more is better" trend regarding the duration of interventions. Short-term interventions (8 weeks or less) produced modest gains ($g = 0.267$). However, as the duration increased to 9–16 weeks ($g = 0.484$) and eventually to 17 weeks or more, the effect size became substantial ($g = 0.644$). This finding underscores the necessity of sustained, long-term implementation in educational settings rather than brief, episodic interventions.
Discussion and Limitations
The stability of these results is supported by the lack of significant publication bias (Egger test $p = 0.0803$). However, the authors acknowledge certain limitations. While the majority of included studies were of high methodological quality, a small subset (Grade C) lacked rigorous randomization. Furthermore, the geographic concentration of the source literature (primarily China and Western contexts) may influence the generalizability of the "agent" rankings, as teacher-student dynamics vary significantly across cultures.
Conclusion
This meta-analysis provides clear evidence that interactive reading is a potent tool for narrative development, with its effectiveness peaking during the 4–5 year old developmental window. The surprising efficacy of peer-led interaction suggests that educators should reconsider the design of classroom reading activities, perhaps incorporating more "child-to-child" dialogic opportunities. Finally, the significant impact of long-term interventions suggests that narrative support should be a consistent, integrated component of early childhood curricula rather than a short-term remedial measure.
Future research should delve into the specific mechanisms of peer interaction—such as the role of play and social negotiation—to better understand why it outperforms traditional adult-led models.
Summary updated