CoChat

Integrating Generative AI into public health education: a mixed methods evaluation of student experiences

Open access
Mengesha, Zelalem, Estellore, Jennifer, Smith, Jennifer, George, Mathew Sunil, McFarlane, Rosemary A., Parmar, Jinal · 2026 · Frontiers in Public Health

The DOI matches its Crossref record; 41 of the 49 references it cites were found in a registry or an index; 7 of the 23 sources its claims cite were checked for support.

Abstract

Introduction: As Generative Artificial Intelligence (GenAI) continues to influence pedagogical practices in higher education, prevailing discourse has largely focused on issues of assessment integrity, often overlooking student perspectives. Yet, understanding student voices is essential to ensure that AI-enhanced teaching and assessment remain student centred, equitable, and aligned with intended learning outcomes. This study aimed to explore postgraduate health science students’ perceptions and experiences of the structured integration of GenAI into teaching and assessment within the Epidemiology and Principles of Research unit at the University of Canberra. Methods: A mixed-methods study was conducted using pre- and post-intervention surveys administered to all enrolled students, with 78 participants completing the evaluation. The intervention involved the scaffolded integration of GenAI into a critical appraisal assessment. Quantitative data were analysed using descriptive statistics and Fisher’s exact test to assess changes over time. Qualitative data from open-ended responses were analysed using thematic analysis following Braun and Clarke’s six-phase approach. Results: Students reported statistically significant improvements in their understanding of GenAI and perceived ability to use it effectively ( p < 0.001). Thematic analysis identified five key themes: tensions between automation and authentic learning; prompt literacy as a new academic skill; GenAI as a support for metacognitive engagement; ethical ambiguity and cognitive dissonance; and future-oriented learning. Conclusion: The findings highlight the importance of thoughtful pedagogical designs that center student voice, support critical engagement, and prepare learners for responsible and reflective use in future professional practice.

Overview

CoChat Summary

TL;DR

This mixed-methods evaluation of 78 postgraduate health science students found that scaffolded integration of GenAI into an epidemiology critical appraisal assessment was associated with significant gains in self-reported GenAI understanding, familiarity, and effective use. Qualitative themes highlighted tensions between automation and authentic learning, prompt literacy, metacognitive engagement, ethical ambiguity, and future-oriented learning.

Key findings

  • Among 78 participants (71% response rate from 110 invited), statistically significant improvements occurred in general understanding of GenAI (moderate to extensive understanding rose from 53.4% to 88.0%), familiarity with GenAI tools (“familiar” or “very familiar” rose from 21.7% to 67.1%), and perceived ability to use GenAI effectively (“effective” or “very effective” rose from 25.0% to 55.4%), all p < 0.001.
  • Concerns regarding GenAI use increased significantly, from 43.3% pre-intervention to 56.8% post-intervention (p < 0.001), and interest in receiving more GenAI training also increased from 83.0% to 87.1% (p = 0.002).
  • Perceived usefulness, ease of use, feelings about integration, and belief that GenAI can improve epidemiological research did not change significantly (p = 0.760, 0.699, 0.239, and 0.753, respectively).
  • Thematic analysis identified five key themes: tensions between automation and authentic learning; prompt literacy as a new academic skill; GenAI as a support for metacognitive engagement; ethical ambiguity and cognitive dissonance; and future-oriented learning.

Methods

  • A convergent mixed-methods study was conducted in the Epidemiology and Principles of Research postgraduate unit at the University of Canberra in 2025; all enrolled students (n = 110) were invited, and 78 completed the evaluation.
  • The intervention scaffolded GenAI (Copilot) into a critical appraisal assessment worth 40% of the final grade, with a 10% critical reflection, using a 2-hour tutorial informed by Salmon’s five-stage model, RTSI prompt writing, and the Joanna Briggs Institute appraisal tool.
  • Quantitative data were analysed in Stata 16 using descriptive statistics and Fisher’s exact test; qualitative open-ended responses were analysed using Braun and Clarke’s six-phase thematic analysis in NVivo 15 with inductive coding.

Limitations

  • The paper states that the evaluation relied on survey data, including open-ended responses, and did not include structured baseline measures of students’ prior engagement with conventional critical appraisal tasks or their conceptions of authentic learning, limiting direct before-and-after comparison of learning processes.
  • The paper states that future research should use more in-depth qualitative approaches such as focus groups or interviews and include baseline and comparative measures.
  • Evident from the text: the study was conducted in a single unit and single cohort at one university, with no priori sample size calculation; the abstract reports five themes while the Results section says “six major themes” but lists only five; and professional preparedness was not directly measured, as the paper itself notes.

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

Integrity signals

Registry

Crossref · 10.3389/fpubh.2026.1825937

Cites

41 of 49 references verified0 citation errors · 8 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

5 Valid · 1 Not valid · 13 Not checked · 56 Own results and interpretations

27 more claims

Create a free account to see all 75 claims and to verify the 2 references still open.

See all claims — free account
  • Probably not a substantial claim. 22
  • It may state two separate claims as one. 28
  • The quote only partly states this claim. 1

Create a free account to see these claims.

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.

2 of 2

Citations

Cites

Show all 49 references

Metadata

Authors
Mengesha, Z., Estellore, J., Smith, J., George, M. S., McFarlane, R. A., & Parmar, J.
Affiliations
  • University of Canberra (AU)
  • Learning and Teaching, University of Canberra
  • Discipline of Public Health, University of Canberra
Venue
Frontiers in Public Health
Published
2026-07-15
Volume
14
Pages
1825937-1825937
ISSN
2296-2565 (ISSN-L)
OpenAlex
W4415790800
Cited by
1
References
49
Citations by year
2026: 1
FWCI
2.44
Type
journal article
Language
en
Open access
Locations
  • Frontiers in Public Health (publishedVersion, cc-by)
  • Research Square (acceptedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • PubMed Central (submittedVersion, cc-by)
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Health Sciences › Medicine › Family Practice › Clinical Reasoning and Diagnostic Skills
  • Social Sciences › Business, Management and Accounting › Marketing › AI in Service Interactions
Keywords
generative AI, public health education, student perceptions, prompt literacy, metacognition, AI ethics, assessment integrity, artificial intelligence in higher education
MeSH terms
Academia, Generative Artificial Intelligence, Female, Humans, Male, Public Health, Surveys and Questionnaires

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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (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: Integrating Generative AI into public health education: a mixed methods evaluation of student experiences. https://app.cochat.ai/papers/10.3389/fpubh.2026.1825937
The paper
Mengesha, Z., Estellore, J., Smith, J., George, M. S., McFarlane, R. A., & Parmar, J. (2026). Integrating Generative AI into public health education: a mixed methods evaluation of student experiences. Frontiers in Public Health. https://doi.org/10.3389/fpubh.2026.1825937

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.