The impact of trial stage, developer involvement and international transferability on universal social and emotional learning programme outcomes: a meta-analysis

89 studies, ages 4-18; 43 cluster-randomised trials, 14 randomised trials, 22 matched comparisons, 2 waitlist, 8 other

School students aged 4-18, international.

A pre-stated three-hypothesis test - efficacy beats effectiveness, developer-led beats independent, and home beats away - conducted by an independent Manchester team with no programme ownership. Splits were efficacy 61 versus effectiveness 25; developer-led 34, developer-involved 25, independent 30; home 71 versus away 18. This is the paper the archive expected to supply a quantified developer-bias gradient, and the honest result is that it does not.

Main effects - social-emotional competence 0.53, attitudes 0.17, prosocial behaviour 0.33, conduct problems 0.28, emotional distress 0.19, ACADEMIC ACHIEVEMENT 0.28 (0.18 to 0.40, k = 15). THE EFFICACY-TO-EFFECTIVENESS DECAY IS CONFIRMED - mean difference 0.13, significant on four of seven outcomes, with prosocial 0.37 versus 0.18, conduct 0.34 versus 0.15, and academic achievement 0.38 (efficacy, k = 11) versus 0.22 (effectiveness, k = 5). THE DEVELOPER-INVOLVEMENT HYPOTHESIS IS NOT SUPPORTED - for social-emotional competence the led/involved/independent series runs 0.21 / 0.48 / 0.69 and for conduct problems 0.20 / 0.25 / 0.37, that is, INDEPENDENT evaluations produced the LARGEST effects; the hypothesis held for only two of seven outcomes and neither was statistically significant. The fidelity explanation was also rejected (chi-square = .633, p = .718). The home-versus-away hypothesis was supported on four of seven outcomes, all significant, with social-emotional competence 0.56 at home versus 0.06 abroad.

Medium - a quarter of the pool is non-randomised matched-comparison work.

Not applicable as a synthesis. Its negative result on developer involvement is important precisely because it cuts against this archive's prior, and it is recorded rather than explained away. Note that the contrary trial-level evidence remains strong - the Positive Action Chicago schools produce opposite conclusions depending on who analyses them.