Only perceived accuracy differed across educational contexts, rated lower in CTeSP than in secondary and undergraduate education¶
Claim · All claims
Evidence · 2 studies · 2 associationalr1·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? At the response level, only accuracy differed across contexts, with accuracy rated higher in secondary education (M = 4.63) and undergraduate education (M = 4.42) than in CTeSP (M = 4.20). → Mendonça PC 2026
q3 i? At the report level, no dimension differed across contexts after Holm correction. → Mendonça PC 2026 (2)
Evidence¶
Mendonça PC 2026¶
Mendonça PC, Quintal F, Figueiredo M and Mendonça F (2026) Perceived usefulness and intention to use large language model-generated feedback across three educational levels: a user-centred study in programming. Front. Educ. 11:1934069. https://doi.org/10.3389/feduc.2026.1934069
q3 · i? · associational · r1
Kruskal–Wallis comparison of the three educational contexts on response-level dimensions; only accuracy was significant [H(2) = 10.90, p = 0.022], with Dunn comparisons showing "accuracy was rated higher in secondary education (M = 4.63) than in CTeSP". A gender-adjusted sensitivity analysis confirmed the CTeSP difference (b = −0.40).
"accuracy was rated higher in secondary education (M = 4.63) than in CTeSP (M = 4.20; z = 3.07, p = 0.007, rank-biserial = 0.49) and higher in undergraduate education (M = 4.42) than in CTeSP (z = 2.63, p = 0.017, rank-biserial = 0.28)"
Mendonça PC 2026 (2)¶
Mendonça PC, Quintal F, Figueiredo M and Mendonça F (2026) Perceived usefulness and intention to use large language model-generated feedback across three educational levels: a user-centred study in programming. Front. Educ. 11:1934069. https://doi.org/10.3389/feduc.2026.1934069
q3 · i? · associational · r1
Kruskal–Wallis comparisons of the six report-level dimensions across the three contexts found no significant differences after Holm correction; the smallest adjusted p-value was for cognitive load [H(2) = 8.32, p = 0.094].
"At the report level, no dimension differed across contexts after the Holm correction, as shown in Table 6, and the Welch analysis of variance agreed with this result."
Discussion¶
Related Claims¶
- Students rate LLM-generated consolidated reports favourably, with cognitive load rated lowest — related
- A design-based study with 42 undergraduates reported a 15% improvement in rubric alignment and faster correction cycles from proactive agentic feedback loops — related
- Students rate LLM-generated response-level feedback favourably, with clarity and perceived accuracy highest and actionability and usefulness lowest — related
- Students perceive AI scoring as accurate for grammar, mechanics, and organization but skeptical of its ability to evaluate diverse styles, complex writing, tone, and emotion — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
perceived-accuracy-lower-ctesp-context-difference |
title |
Only perceived accuracy differed across educational contexts, rated lower in CTeSP than in secondary and undergraduate education |
description |
Only perceived accuracy differed across educational contexts, rated lower in CTeSP than in secondary and undergraduate education |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
generated |
by process:wiki-ingest on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
sources |
2 — mendonça-pc-2026, mendonça-pc-2026-2 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.