The evolution teaching interventions significantly improved student understanding with effect sizes above the 0.40 implementation threshold in both tranches¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q3 · i2 medium
Currency · newest study 2020 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i2 Pre- to post-test scores increased significantly in tranche 1 (Cliff's d = 0.55) and the gain replicated in tranche 2 (Cliff's d = 0.49), both above Hattie's 0.40 hinge point. → Buchan 2020 q3 i2 The paired within-student change in score also increased significantly in both tranches (Cliff's d = 0.55 and 0.48), showing improvement is not an artefact of between-student comparison. → Buchan 2020 (2)

Evidence¶

Buchan 2020¶

Buchan, L., Hejmadi, M., Abrahams, L. & Hurst, L. D. (2020). A RCT for assessment of active human-centred learning finds teacher-centric non-human teaching of evolution optimal. npj Science of Learning 5:19. https://doi.org/10.1038/s41539-020-00078-0

q3 · i2 · causal · r2

Unpaired comparison of all pre- and post-test scores in the 2×2 RCT, analysed by Wilcoxon rank-sum tests. The intervention raised "the mean student test score" by 2.44 marks in tranche 1 and 2.37 marks in tranche 2, with large effect sizes.

"The mean student test score increased significantly in tranche 1 by 2.44 marks (16.27%, P < 2.2 × 10−16) with a large effect size (Cliff's d = 0.55, 95% CI: 0.52 –0.59). This was replicated in Tranche 2 (increase in mean by 2.37 marks, 15.80%, P < 2.2 × 10−16, Cliff's d = 0.49, 95% CI: 0.45 –0.52)."

Buchan 2020 (2)¶

Buchan, L., Hejmadi, M., Abrahams, L. & Hurst, L. D. (2020). A RCT for assessment of active human-centred learning finds teacher-centric non-human teaching of evolution optimal. npj Science of Learning 5:19. https://doi.org/10.1038/s41539-020-00078-0

q3 · i2 · causal · r2

Paired within-student analysis of students taking both pre and post tests (Fig. 2), using Wilcoxon rank-sum tests on change scores. Both tranches showed significant gains with large effect sizes above the 0.40 threshold.

"The mean student change in score was +2.42 marks ( +16.13%, P = < 2.2 × 10−16, Cliff's d = 0.55, 95% CI: 0.51 –0.59) in tranche 1, replicated in tranche 2 (mean increase 2.32 marks, 15.47%, P = < 2.2 × 10−16, Cliff's d = 0.48, 95% CI: 0.44 –0.52)."

Discussion¶

Related Claims¶

• Teacher-judged science ability repeatedly predicts gain in understanding, while age does not and gender effects are weak and non-replicable — related
• Each of the four schemes of work individually produced significant pre-post improvement, with all effect sizes above the 0.4 threshold except SoW 1 in tranche 2 — related
• The four schemes of work showed replicable significant heterogeneity in effectiveness, with the predicted-least-effective teacher-centred/non-human scheme among the most effective — related
• Serially taught lesson combinations produce replicable positive interaction effects, with the two most effective schemes showing positive interaction terms — related
• Comparing lesson 2 (student- vs teacher-centred) and lesson 4 (human- vs non-human-centred) activities in isolation yields weak, inconsistent or negligible effects — a narrower finding that bears on this claim
• Most teacher-level predictors of student performance failed to replicate between tranches, but improved teacher confidence in teaching evolution was a replicable predictor — related