CoChat

Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness

Open access
Henderson, Michael, Bearman, Margaret, Chung, Jennifer, Fawns, Tim, Buckingham Shum, Simon, Matthews, Kelly E., de Mello Heredia, Jimena · 2026 · Assessment & Evaluation in Higher Education

The DOI matches its Crossref record; 37 of the 37 references it cites were found in a registry or an index; 20 of the 43 sources its claims cite were checked for support.

Abstract

The rapid integration of Generative Artificial Intelligence (GenAI) into educational contexts has presented both opportunities and challenges for students seeking and using feedback. While AI-generated feedback can offer increased access, timely responses and personalised insights, concerns about the quality of AI-generated feedback still persist, including issues of bias, factual inaccuracies, and homogenisation. This study investigates how students use, value and trust AI-generated feedback compared to feedback from educators. This paper draws on a large-scale cross-sectional survey administered across four major Australian universities. A quantitative analysis of 6960 respondents revealed that half of the students sought feedback from GenAI. Overall, students reported teacher feedback to be more helpful, and especially trustworthy. This paper also reports on the thematic analysis of 8642 open-ended responses, providing deeper insights into students’ experiences. This includes findings that students valued the feedback from GenAI because of its ease of access, timeliness, volume, understandability and that it was perceived to be less risky than seeking feedback from teachers. At the same time, students were concerned about GenAI’s reliability as well as contextual and disciplinary expertise. We argue that GenAI and teacher feedback appear to serve different needs, and therefore are complementary but not interchangeable.

Overview

CoChat Summary

TL;DR

In a cross-institutional Australian survey, about half of 6,960 students used GenAI for feedback, and students rated teacher feedback as more helpful and especially more trustworthy than GenAI feedback. Qualitative responses suggest GenAI and teacher feedback serve different needs and are complementary rather than interchangeable.

Key findings

  • Of 6,960 respondents, 49.7% (n = 3,461) sought feedback from GenAI and 50.3% (n = 3,499) did not. Among GenAI feedback users, 83.9% rated AI feedback somewhat/very helpful versus 82.2% for teacher feedback; 60.1% rated AI feedback somewhat/very trustworthy versus 90.5% for teacher feedback.
  • Paired-samples t-tests found teacher feedback was rated more helpful than GenAI feedback (mean difference 0.14, SD = 1.15, t(3327) = -7.17, p < .001, Cohen’s d = -0.13, small effect) and more trustworthy (mean difference 0.93, SD = 1.09, t(3327) = -49.30, p < .001, Cohen’s d = -0.89, large effect). Means: helpfulness GenAI 4.02 vs teacher 4.16; trustworthiness GenAI 3.53 vs teacher 4.46.
  • Qualitative analysis of 8,642 open-ended responses (11,903 coded instances) found most comparisons concerned the nature of feedback information (54.8%), followed by feedback processes (22.7%). GenAI was seen as less reliable, relevant, and contextualised but more objective and understandable; it was valued for access (ease n = 875, speed n = 614, volume n = 304).
  • Among non-users, reasons included being unaware of the option or how to use it (28.1%, n = 957), trustworthiness concerns (28.7%), quality concerns (37.5%), privacy concerns (2.7%, n = 93), and integrity concerns (9.1%, n = 310). The authors conclude GenAI and teacher feedback are complementary but not interchangeable.

Methods

  • Cross-sectional online survey across four Australian universities; approximately 192,000 students were invited, 5% (n = 10,132) volunteered, and 79% (n = 8,028) completed the survey. This paper draws on 6,960 students who responded to the GenAI-for-feedback item. Sample: 57% women, 72% aged 18–24, 58% spoke English as the main language at home, 89% full time, 58% domestic, 61% undergraduate, 92% on campus.
  • Nine feedback-specific survey items; four Likert items on helpfulness and trustworthiness of GenAI versus teacher feedback were treated as continuous data, with two paired-samples t-tests conducted in SPSS version 29.
  • Qualitative thematic analysis of open-ended responses following Braun and Clarke (2006), using inductive semantic coding; initial 18 codes were refined and expanded to 48 codes organised into themes and subthemes, yielding 8,642 usable text responses and 11,903 coded instances.

Limitations

  • Stated: the response rate of about 8,000 from over 200,000 enrolled students cannot claim to be representative; the optional survey may have attracted students with stronger views about GenAI; Australian higher education may differ from other contexts; and the survey was completed when few students would have experienced GenAI feedback explicitly grounded in their curriculum/assignment resources, which could change perceptions.
  • Stated/acknowledged: helpfulness findings are self-reported perceptions, and the paper notes this for helpfulness; the study does not measure learning impact or feedback accuracy.
  • Evident from the text: qualitative inter-coder agreement was assessed subjectively rather than reported as a statistic, and the paper cites concerns that short open-ended survey responses may not support rigorous latent qualitative insights; the cross-sectional design also cannot show how perceptions change over time.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1080/02602938.2025.2502582

Cites

37 of 37 references verified0 citation errors · 0 not checkable

Statistics

Figures

Image 1
CC-BY-NC-ND · From the PDF
Image 2
CC-BY-NC-ND · From the PDF
Image 3
CC-BY-NC-ND · From the PDF
Figure 1 Visualising the response frequencies in Table 2, concerning the helpfulness and trustworthiness of feedback from GenAI and teachers.
Figure 1 Visualising the response frequencies in Table 2, concerning the helpfulness and trustworthiness of feedback from GenAI and teachers.CC-BY-NC-ND · From the PDF
Image 5
CC-BY-NC-ND · From the PDF

Verification

Claims the paper makes

15 Valid · 0 Not valid · 19 Not checked · 86 Own results and interpretations

58 more claims

Create a free account to see all 120 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 32
  • It may state two separate claims as one. 35
  • The quote only partly states this claim. 3

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentt(3327) = -7.17reported p < .001 · recomputed p < .001

    “t(3327) = -7.17,p < .001”

  • Consistentt(3327) = -49.3reported p < .001 · recomputed p < .001

    “t(3327) = -49.30,p < .001”

2 of 2

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 37 references

Cited by

See all 211

Metadata

Authors
Henderson, M., Bearman, M., Chung, J., Fawns, T., Buckingham Shum, S., Matthews, K. E., & de Mello Heredia, J.
Affiliations
  • Monash University (AU)
  • Deakin University (AU)
  • University of Technology Sydney
  • The University of Queensland (AU)
  • Monash University
Corresponding author
Henderson, Michael
Venue
Assessment & Evaluation in Higher Education
Published
2025-05-13
Volume
51
Issue
5
Pages
863-878
ISSN
  • 0260-2938 (ISSN-L)
  • 1469-297X
OpenAlex
W4410317321
Cited by
211
References
34
Citations by year
2026: 170 · 2025: 41
FWCI
174.63
Type
journal article
Language
en
Open access
Licence
Locations
  • Assessment & Evaluation in Higher Education (publishedVersion, cc-by-nc-nd)
  • Figshare (submittedVersion)
  • OPUS - Open Publications of UTS Scholars (University of Technology Sydney) (submittedVersion, cc-by-nc-nd)
Topics
  • Social Sciences › Education › Artificial Intelligence in Education
  • Social Sciences › Education › Student Assessment and Feedback
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
Keywords
generative AI, AI-generated feedback, teacher feedback, student perceptions, trustworthiness, perceived usefulness, university students, feedback quality, student use of AI

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness. https://app.cochat.ai/papers/10.1080/02602938.2025.2502582
The paper
Henderson, M., Bearman, M., Chung, J., Fawns, T., Buckingham Shum, S., Matthews, K. E., & de Mello Heredia, J. (2026). Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness. Assessment & Evaluation in Higher Education. https://doi.org/10.1080/02602938.2025.2502582

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.