AI-generated revisions employ significantly more appraisal resources than peer-feedback revisions, which exceed first drafts¶
Claim · All claims
Evidence · 1 study · associationalr2·q3
Currency · newest study 2026 · checked against its sources 2026-10-19 (automated, 1 of 1 entry) · 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? Across 168 essays, overall appraisal resource frequency was highest in AI-generated revisions, intermediate in peer-feedback revisions, and lowest in first drafts (all ps < .001). → Zhao H and Ji J 2026
Evidence¶
Zhao H and Ji J 2026¶
Zhao H and Ji J (2026) Human-written and AI-generated revisions: an appraisal analysis of dialogic positioning in argumentative writing. Front. Educ. 11:1904852. https://doi.org/10.3389/feduc.2026.1904852
q3 · i? · associational · r2
Linear mixed model on a corpus of 168 essays (14 participants × 4 tasks × 3 draft types), frequencies normalized per 100 words, found AI revisions (M = 10.811) "employed significantly more appraisal resources than peer-feedback revisions (M = 9.533)"; first drafts averaged 8.151. No effect size printed.
"Bonferroni-adjusted post-hoc comparisons indicated that AI- generated revisions (M = 10.811) employed significantly more appraisal resources than peer-feedback revisions (M = 9.533), which in turn exceeded first drafts (M = 8.151) (all ps < .001)."
Discussion¶
Related Claims¶
- Revision-driven increases in appraisal resources are driven by Attitude and Graduation, while Engagement frequency remains stable across draft types — related
- Entertain is the most frequently used heteroglossic subcategory across all draft types, decreasing in peer-feedback revisions but remaining relatively high in AI revisions — related
- Heteroglossic resources outnumber Monoglossic resources across all draft types, with no significant difference in Engagement distribution across revision conditions — related
- Peer-feedback revisions shift toward a more contractive dialogic stance while AI-generated revisions keep a relatively balanced Contract–Expand distribution — related
- Counter rises in both revision conditions, Endorse rises in peer-feedback revisions but declines in AI revisions, and Pronounce declines in both — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
ai-revisions-most-appraisal-resources |
title |
AI-generated revisions employ significantly more appraisal resources than peer-feedback revisions, which exceed first drafts |
description |
AI-generated revisions employ significantly more appraisal resources than peer-feedback revisions, which exceed first drafts |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
source_checked |
2026-10-19 |
generated |
by process:wiki-ingest on 2026-10-12 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — zhao-h-and-ji-j-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.