Differences between human and AI scoring: A meta-analysis of english language assessments.

Sci Rep · 2026 · PMC13237208 · PMID 41981192

Europe PMC DOI

hedged sentences

0.0542

closest p · 1.1× alpha

0.2171

boldest claim

The sentences

marginally non-significantp =.0542so close (0.05 < p ≤ 0.1)

DL/Hybrid systems showed a marginally non-significant AI advantage ( d = 0.1062, p =.0542).

also in 1,493 other papers

did not reach statistical significancep =.2171not close (p > 0.1)

The results of Egger’s test (see Table 4 ) showed that publication bias did not reach statistical significance ( z = −1.2315, p =.2171).

also in 111,027 other papers

Quoted from the open-access full text in Europe PMC under the licence the publisher applied. The sentence is reproduced exactly as published; the emphasis is ours.