Syntes: Delen et al. (2026) — out-of-school learning raises achievement (ḡ = 0.529)

Källa: Delen, İ.; Özüdoğru, F.; Şen, S.; Biasutti, M. (2026). The Impact of Out-of-School Learning on Academic Achievement in Elementary and Secondary Education: A Meta-Analysis. https://doi.org/10.3390/su18031437

Verified content

• [abstract] 42 peer-reviewed studies, 54 effect sizes.
• [abstract] Overall medium-to-large effect of out-of-school learning on academic achievement: ḡ = 0.529, 95% CI [0.392, 0.665], statistically significant.
• [abstract] Stronger effects generally in elementary education and in science and social sciences; high heterogeneity; effectiveness depends on instructional alignment and implementation quality.

Full-text numbers (read 2026-11-08)

• [full text] Table 2: random effects, k = 54, g = 0.529, SE 0.070, 95% CI [0.392, 0.665], Q = 744.810, df 53, p < 0.001. Studies published 1970–2022.
• [full text] Table 3, random assignment: yes k = 29, g = 0.307 [0.167, 0.447]; no k = 16, g = 1.020 [0.722, 1.318]; Q_B = 17.998, p < 0.001. Discussion: "stronger claims regarding effectiveness and generalizability should rely primarily on evidence from randomized or well-controlled quasi-experimental studies."
• [full text] Table 3, region: North America k = 15, g = 0.149 [−0.010, 0.308]; Scandinavia k = 4, g = 0.180 [−0.093, 0.454]; Central Europe k = 11, 0.414; Middle East k = 12, 0.976; Asia k = 6, 1.267; Africa k = 3, 0.732. Country: USA 0.149, Türkiye 1.023 (k = 11), Germany 0.421 (k = 10).
• [full text] Table 3, discipline: science k = 36, 0.632; maths k = 5, 0.295 [0.053, 0.538]; social science k = 5, 0.757; reading k = 3, 0.166 [0.069, 0.263].
• [full text] Table 3, school level: elementary k = 14, 0.805; middle k = 12, 0.617; secondary k = 27, 0.500; mixed k = 6, 0.209 [−0.035, 0.454]. Table 4: grade slope B = −0.136, SE 0.044, p = 0.002 (k = 35).
• [full text] Table 3, duration: 2–6 h k = 7, 0.034 [−0.163, 0.230]; 7–12 h k = 3, 0.637; 13–24 h k = 6, 0.552; more than a week k = 10, 0.360.
• [full text] Setting: designed 0.252, unstructured 0.524, mixed 0.573 (Q_B p = 0.119).

Limits

• Read level: full text since 2026-11-08 (was abstract only).
• Random-effects pooling of 54 effect sizes treated as independent (CMA v3); one outlier (g = 3.941) excluded.
• Test delay is not coded or reported as a moderator, so durability is not addressed.
• Assessment-item-type figures differ between Table 3 (conceptual 0.343, declarative 0.258) and the text (0.682, 0.336).
• Subgroup differences (level, region, subject) are not adjusted for design.

Why it is in the library

Main OA meta for an out-of-school-learning candidate (Phase B Near seed).

Added 2026-10-27 by the project (Phase B expansional push). [abstract] = published abstract; [full text] = full text; [assessment] = judgement; [secondary] = secondary summary pending paper check.