Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 26Artificial Intelligence in Healthcare and EducationArtificial Intelligence in EducationStudent Assessment and Feedback

The DOI matches its Crossref record; 11 of the 11 references it cites were found in a registry or an index.

Dental Students’ Learning Experience: Artificial Intelligence vs Human Feedback on Assignments

Open access
Jayawardena, Chantha K., Gunathilake, Yuvika, Ihalagedara, Dhammika·2025·International Dental Journal
DOI 10.1016/j.identj.2024.12.022PMC PMC11806320
Open this paper in CoChat — free account

Abstract

OBJECTIVE: This study evaluated the effectiveness of an AI-based tool (ChatGPT-4) (AIT) vs a human tutor (HT) in providing feedback on dental students' assignments. METHODS: A total of 194 answers to two histology questions were assessed by both tutors using the same rubric. Students compared feedback from both tutors and evaluated its accuracy against a standard rubric. Students' perceptions were collected on five dimensions of feedback quality. A subject expert also evaluated feedback provided by the two tutors for 40 randomly selected answers. RESULTS: = 9.01, P < .05). In contrast, expert evaluation highlighted that AIT scored higher in identifying mistakes, with significant differences in clarity (W = 40.5, P < .001) and suggestions for improvement (W = 96.5, P < .001). CONCLUSION: AIT demonstrates significant potential to complement HT by providing detailed feedback in a shorter timeframe. While students did not perceive differences in feedback quality, expert analysis identified AIT as superior in clarity and suggestions for improvement.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/j.identj.2024.12.022

Cites

11 of 11 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 The figure outlines the data collection and analysis process for comparing feedback from human tutors (HT) and AI tutors (AIT). A total of 194 answers were collected from an assignment. Both HT and AIT evaluated these answers using the same rubric. Students compared the feedback from both tutors and, 2 days later, assessed this feedback against the standard rubric. Students’ opinions were collected through an online survey. Prompts were analysed for cases where the grade difference between students’ self-evaluated scores and AIT scores was 15% or less. Expert analysis was performed to evaluate the feedback quality of both tutors. The Wilcoxon ranked test and paired t test were applied for statistical comparisons.
Fig. 1 The figure outlines the data collection and analysis process for comparing feedback from human tutors (HT) and AI tutors (AIT). A total of 194 answers were collected from an assignment. Both HT and AIT evaluated these answers using the same rubric. Students compared the feedback from both tutors and, 2 days later, assessed this feedback against the standard rubric. Students’ opinions were collected through an online survey. Prompts were analysed for cases where the grade difference between students’ self-evaluated scores and AIT scores was 15% or less. Expert analysis was performed to evaluate the feedback quality of both tutors. The Wilcoxon ranked test and paired t test were applied for statistical comparisons.CC BY-NC-ND · PubMed Central
Fig. 2 The figure illustrates the process undertaken to ensure ChatGPT's understanding of the rubric and to clarify any doubts. This conversation demonstrates the interaction with ChatGPT to confirm its grasp of the assessment criteria and the specific instructions provided for evaluating student assignments.
Fig. 2 The figure illustrates the process undertaken to ensure ChatGPT's understanding of the rubric and to clarify any doubts. This conversation demonstrates the interaction with ChatGPT to confirm its grasp of the assessment criteria and the specific instructions provided for evaluating student assignments.CC BY-NC-ND · PubMed Central
Fig. 3 The figure illustrates a comparative analysis of students’ perceptions of feedback provided by Human tutors (HT) and AI tutors (AIT) on their assignments. Subfigures (A-C) depict the students’ views on grading accuracy, clarity of feedback, and relevance of feedback to the rubric criteria, respectively. The broken lines in these subfigures represent the proportion of students who rated these aspects as satisfactory, exceeding 50% for each tutor. Subfigures (D-F) show the perceived usefulness of feedback in pointing out mistakes, the helpfulness of feedback in fostering critical thinking and self-reflection, and the comfort level with each tutor. The broken lines in these subfigures highlight the percentage of students who found the feedback from the respective tutors to be helpful and very helpful in pointing out mistakes, promoting critical thinking and self-evaluation, and felt comfortable and very comfortable with the tutor. Percentages were calculated based on the number of valid responses (N).
Fig. 3 The figure illustrates a comparative analysis of students’ perceptions of feedback provided by Human tutors (HT) and AI tutors (AIT) on their assignments. Subfigures (A-C) depict the students’ views on grading accuracy, clarity of feedback, and relevance of feedback to the rubric criteria, respectively. The broken lines in these subfigures represent the proportion of students who rated these aspects as satisfactory, exceeding 50% for each tutor. Subfigures (D-F) show the perceived usefulness of feedback in pointing out mistakes, the helpfulness of feedback in fostering critical thinking and self-reflection, and the comfort level with each tutor. The broken lines in these subfigures highlight the percentage of students who found the feedback from the respective tutors to be helpful and very helpful in pointing out mistakes, promoting critical thinking and self-evaluation, and felt comfortable and very comfortable with the tutor. Percentages were calculated based on the number of valid responses (N).CC BY-NC-ND · PubMed Central
Fig. 4 Pictures (A-E) illustrate samples of different feedback styles used by HT and AIT. HT noted correct points by underlining text (red lines) and using ✓ marks (A and B) and indicated missing information with short handwritten phrases. Motivational elements are shown on some occasions (B). Pictures (A) and (C) show that HT and AIT gave scores of 64 and 57.7, respectively, for the same answer. AIT begins feedback with correct points, followed by missing points, presented as short paragraphs (C) or lists (D). AIT also included detailed personalised feedback (E) in addition to overall feedback on some occasions.
Fig. 4 Pictures (A-E) illustrate samples of different feedback styles used by HT and AIT. HT noted correct points by underlining text (red lines) and using ✓ marks (A and B) and indicated missing information with short handwritten phrases. Motivational elements are shown on some occasions (B). Pictures (A) and (C) show that HT and AIT gave scores of 64 and 57.7, respectively, for the same answer. AIT begins feedback with correct points, followed by missing points, presented as short paragraphs (C) or lists (D). AIT also included detailed personalised feedback (E) in addition to overall feedback on some occasions.CC BY-NC-ND · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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?

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Show all 11 references

Cited by

See all 26

Related

Metadata

Authors
Jayawardena, C. K., Gunathilake, Y., & Ihalagedara, D.
Affiliations
University of Peradeniya (LK)
Corresponding author
Jayawardena, Chantha K.
ORCID
Jayawardena, Chantha K.
Venue
International Dental Journal
Publisher
Elsevier BV
Published
2025-02
Volume
75
Issue
1
Pages
100-108
ISSN
0020-6539 (ISSN-L)
DOI
10.1016/j.identj.2024.12.022
PMC
PMC11806320
OpenAlex
W4406244152
Cited by
26
References
11
Citations by year
2026: 19 · 2025: 7
FWCI
6.26
Type
journal article
Language
en
Open access
Licence
Locations
  • International Dental Journal (publishedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
University of Peradeniya
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Social Sciences › Education › Artificial Intelligence in Education
  • Social Sciences › Education › Student Assessment and Feedback
Keywords
ChatGPT-4, automated feedback, dental students, feedback quality, student perceptions, rubric-based assessment
MeSH terms
Formative Feedback, Artificial Intelligence, Education, Dental, Educational Measurement, Feedback, Female, Humans, Learning, Male, Students, Dental

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.

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

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.