Early mathematics interventions in pre-primary education: A three-level meta-analysis of intentional teaching and exploratory play approaches

Early mathematics interventions in pre-primary education are generally effective, but prior syntheses rarely distinguish the dominant pedagogical approach used and provide limited evidence on which study or implementation features are associated with effects within each approach. This three-level meta-analysis synthesized 74 studies (2000–2025) of early mathematics interventions in preschool and kindergarten, focusing on two broad pedagogical approaches: intentional teaching and exploratory play. Studies were classified by dominant pedagogical emphasis as intentional teaching ( k = 56) or exploratory play ( k = 18), yielding 566 effect sizes. Both approaches were associated with positive effects on children’s learning (intentional teaching: g = 0.32; exploratory play: g = 0.21), and the difference between approaches was not statistically significant. Moderator analyses examined heterogeneity by program design, intervention conditions and outcome evaluation. For intentional teaching, effects were larger for advanced numeracy outcomes (quantitative comparison: g = 0.42; calculation: g = 0.45) than for foundational number and counting skills ( g = 0.21). Effects also attenuated over time, with smaller follow-up than immediate post-test effects (post-test: g = 0.34; follow-up: g = 0.20), suggesting partial fadeout. For exploratory play, interventions incorporating contextual scenarios showed larger effects ( g = 0.47) than those without ( g = 0.13), and effects were stronger for math-specific outcomes ( g = 0.26) than for non-math outcomes ( g = -0.01). Overall, the findings support the use of both pedagogical approaches in early mathematics programs and highlight the need to align instructional design with targeted numeracy goals and to sustain learning gains over time.