Motivation gains need a local learning test
Does generative AI make students more motivated to learn? A new meta-analysis combines 42 controlled studies involving 6,059 participants and reports an average positive effect of 0.764 standard deviations. That average conceals substantial variation: the prediction interval stretches from a negative effect to a large positive one. The included studies span primary, secondary and university education; exploratory estimates were smaller in school settings than in universities. The finding supports testing thoughtfully designed uses, but does not establish that a particular district product will help. Motivation was the outcome under review. Better grades, durable understanding and independent performance cannot be inferred from the pooled motivation result.
Why it matters for leaders
Editorial recommendation: Have curriculum leaders define success before expanding a pilot. Pair a brief motivation survey with an unaided task that uses new material and a delayed assessment. Compare with normal instruction and examine students needing additional support separately. A student who enjoys the tool but cannot explain the work needs a different instructional response from one who gains both confidence and understanding. Budget teacher time for reviewing those differences.
The evidence
The synthesis covers 56 effects from experimental and quasi-experimental studies published in 2023-2025. Heterogeneity was high: I-squared 93.7%.
Who was studied
Primary, secondary and university learners; mixed educational contexts.
What the evidence cannot tell us
Study-level subgroup differences are exploratory. A positive average does not guarantee local benefit; instruments, interventions and populations vary substantially.
Read the original sources
Summary and leadership analysis from the Week ending 2026-09-25 brief. The edition’s selection notes explain source access and evidence limitations.