The positive GenAI writing effect is robust across dependence-aware, sensitivity, and risk-of-bias specifications, but the wide prediction interval spans near-zero to large effects¶
Claim · All claims
Evidence · 1 study · quant-synthesisr3·q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q3 i? The 95% prediction interval was wide (−0.59 to 2.19), indicating true effects in new settings could plausibly range from near-zero to large positive values. → Mark Feng Teng 2026
Evidence¶
Mark Feng Teng 2026¶
Mark Feng Teng. (2026). Effectiveness of GenAI in enhancing writing performance: a meta-analysis. Asian-Pacific Journal of Second and Foreign Language Education. https://doi.org/10.1186/s40862-026-00444-2
q3 · i? · quant-synthesis · r3
Prediction-interval diagnostic reported alongside the primary synthesis. The article states the interval "was wide (−0.59 to 2.19)", underscoring that AI-supported writing interventions may produce near-zero or even negative effects in some settings.
"The 95% prediction interval was wide (−0.59 to 2.19), indicating that although the average effect is positive, true effects in new settings could plausibly range from near-zero to large positive values."
Discussion¶
Related Claims¶
- Risk-of-bias classification is the only robust moderator of GenAI writing effects, with unrated studies showing significantly larger effects than high-risk studies — related
- Effects of AI literacy interventions vary considerably across contexts, with most variance between studies (71.361%) and a wide prediction interval — related
- GenAI effects on externally assessed learning are highly heterogeneous and conditional on prior knowledge, scaffolding, verification, and task design, with pooled effects unstable under publication-bias correction — related
- Effects of game-based learning in STEM education vary substantially across studies, with prediction intervals spanning trivial-to-negative through very strong effects — a broader claim this one bears on
- Effect sizes vary widely between studies and especially between comparisons within studies — related
- Meta-analytic evidence on GenAI and critical thinking shows a positive mean association with a prediction interval spanning harm to benefit — related
- Effect sizes across studies show substantial heterogeneity, with wide prediction intervals spanning large negative to large positive true effects — related
- Effect sizes for gender differences in L2 achievement are highly heterogeneous across studies (I² = 89.82%) — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
genai-writing-effect-robustness-checks |
title |
The positive GenAI writing effect is robust across dependence-aware, sensitivity, and risk-of-bias specifications, but the wide prediction interval spans near-zero to large effects |
description |
The positive GenAI writing effect is robust across dependence-aware, sensitivity, and risk-of-bias specifications, but the wide prediction interval spans near-zero to large effects |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — mark-feng-teng-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.