Claims

Differentiated-AI materials exceeded unified-AI materials on reading comprehension within each proficiency stratum, most strongly for high-proficiency learners (d = 1.40)

Claim · All claims
Evidence · 3 studies · 3 causal r2 · q3 · i0–i3
Currency · newest study 2026 · checked against its sources 2026-10-19 (automated, 3 of 3 entries) · 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 i2 Within-proficiency planned contrasts showed differentiated-AI exceeded unified-AI on reading comprehension for low-proficiency learners by d = 0.60. → Gao X 2026 q3 i0 For intermediate learners the differentiated-AI advantage over unified-AI was small (d = 0.16), with differentiated discourse cues adding little. → Gao X 2026 (2) q3 i3 For high-proficiency learners differentiated-AI exceeded unified-AI by d = 1.40, the strongest case for added value beyond the unified AI-adapted material. → Gao X 2026 (3)

Evidence

Gao X 2026

Gao X (2026) From unified to differentiated materials: generative AI–supported adaptation of EAP reading materials. Front. Psychol. 17:1887565. https://doi.org/10.3389/fpsyg.2026.1887565

q3 · i2 · causal · r2

Planned within-stratum contrast on reading comprehension (Table 6) holding proficiency constant, comparing differentiated-AI with unified-AI. The low-proficiency difference was "1.56 points (13.23 vs. 14.79), d = 0.60", a moderate advantage attributed to local access support such as glosses and sentence unpacking.

"For low-proficiency learners, differentiated-AI exceeded unified-AI by 1.56 points (13.23 vs. 14.79), d = 0.60."

Gao X 2026 (2)

Gao X (2026) From unified to differentiated materials: generative AI–supported adaptation of EAP reading materials. Front. Psychol. 17:1887565. https://doi.org/10.3389/fpsyg.2026.1887565

q3 · i0 · causal · r2

Same planned-contrast analysis (Table 6): the intermediate-group difference was 0.38 points (14.55 vs. 14.92), d = 0.16. The article reports that "differentiateddiscoursecuesaddedlittleovertheunified-AI version", suggesting general readability support already met part of this group's need.

"In the intermediate group,differentiateddiscoursecuesaddedlittleovertheunified-AI version. General readability support may already have met part of this group's reading need, and paragraph-function labels or claim–evidence outlines did not translate into a large comprehension gain in the current task."

Gao X 2026 (3)

Gao X (2026) From unified to differentiated materials: generative AI–supported adaptation of EAP reading materials. Front. Psychol. 17:1887565. https://doi.org/10.3389/fpsyg.2026.1887565

q3 · i3 · causal · r2

Planned within-stratum contrast on reading comprehension (Table 6) for the high-proficiency stratum: a "2.47 points (16.26 vs. 18.73),d = 1.40" advantage for differentiated-AI over unified-AI, the largest learner-outcome result in the study.

"For high-proficiency learners, the difference was 2.47 points (16.26 vs. 18.73),d = 1.40. This within-stratum comparison identifies the high-proficiency group as the strongest case for added value beyond the unified AI-adapted material"

Discussion

Page metadata
Field Value
type claim
id within-stratum-differentiated-advantage-reading-comprehension
title Differentiated-AI materials exceeded unified-AI materials on reading comprehension within each proficiency stratum, most strongly for high-proficiency learners (d = 1.40)
description Differentiated-AI materials exceeded unified-AI materials on reading comprehension within each proficiency stratum, most strongly for high-proficiency learners (d = 1.40)
status review
evidence_strength moderate
newest_evidence 2026
source_checked 2026-10-19
generated by process:wiki-ingest on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 3 — gao-x-2026, gao-x-2026-2, gao-x-2026-3

Frontmatter is the page's machine-readable record — see the Schema & Guide.