Critical reviewResearch writingAll disciplines

ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024)

An early systematic literature review mapped how ChatGPT was studied in universities during its first year. This critical review explains what the review found, tests how far its methods support those findings, and shows doctoral researchers how to use it, and how to do better in their own reviews.

What you will learn

  • Summarise the aims, methods and main findings of Baig and Yadegaridehkordi (2024)
  • Evaluate the search, screening and quality appraisal of a systematic literature review
  • Explain why counting significant constructs across adoption studies is not the same as estimating their effect
  • Position the review within wider syntheses of artificial intelligence in higher education
  • Cite an early-stage review accurately in your own literature review chapter

What you will unlearn

  • A systematic review is reliable simply because it calls itself systematic.Its reliability depends on how the search, screening and appraisal were done and reported, which you need to check.
  • A quality assessment step means weak studies were filtered out.Quality criteria can measure relevance or reporting rather than risk of bias, so read what was actually scored.
  • The constructs reported most often are the strongest predictors of adoption.A frequency count shows what studies tested and found significant, not how large the effects were.
  • Findings about ChatGPT from 2023 still describe how students use generative AI today.The tools, policies and user habits have changed quickly, so early findings need to be dated and qualified.

Skills you will practise

  • Critical appraisal of systematic reviews
  • Search and screening evaluation
  • Technology acceptance theory
  • Evidence synthesis
  • Literature review writing
  • Responsible use of generative AI

The source and why it matters

The paper under review is Baig and Yadegaridehkordi (2024), "ChatGPT in the higher education: A systematic literature review and research challenges", published open access in the International Journal of Educational Research (volume 127, article 102411). Both authors are based at CQUniversity in Sydney. The review sets out to explore the trends, adoption measures, applications and current limitations of research on ChatGPT in higher education.

It matters to doctoral researchers for two reasons. First, it is widely cited as a map of the first wave of ChatGPT studies, so many literature reviews on generative artificial intelligence (AI) in education will meet it. Second, it is a useful teaching case: a review conducted quickly on a fast-moving topic, where the strengths and weaknesses of systematic review methods are easy to see. Reading it critically is good practice for appraising any secondary study you intend to rely on.

ChatGPT in higher education
Figure 1. Baig and Yadegaridehkordi (2024).

What the review did and found

The authors followed the systematic literature review guidelines of Kitchenham and Charters (2007), which were developed for software engineering. They searched publisher databases including Emerald, ERIC, MDPI, SAGE, Elsevier, SpringerLink, Frontiers, PLOS ONE, Wiley and Taylor & Francis, added a manual Google Scholar search, and included English-language research articles on ChatGPT in higher education published between January 2023 and mid-January 2024. Reviews, editorials, conference papers and studies of school or pre-university education were excluded. After screening and a quality assessment, 57 articles were analysed (Baig & Yadegaridehkordi, 2024).

The review was organised around four questions. The table summarises what it reported for each.

Research questions and findings of Baig and Yadegaridehkordi (2024), summarised
Research questionMain findings reported
What are the trends in ChatGPT research in higher education?Rapid growth through 2023 (a peak of 12 articles in September 2023); studies from more than 25 countries; quantitative designs most common (49%), then qualitative (18%) and mixed methods (3%), with 30% not stating a method; surveys the main data collection tool (44%)
How have adoption and use been measured?Studies grouped into post-adoption (continued use), intention to use, and acceptance and use; the Technology Acceptance Model (TAM), UTAUT and UTAUT2 most often applied; social influence, perceived usefulness, effort expectancy and facilitating conditions most often reported as significant
How is ChatGPT applied?Academic staff: designing assignments and examinations, supervision, lesson planning, curriculum development. Students: language learning, writing support, coding help, personalised learning. Researchers: idea generation, data analysis, writing and editing. Non-academic staff: content generation and administrative support
What are the limitations and future directions?Few studies of staff and non-academic users; heavy reliance on student surveys; a need for qualitative and mixed methods; concerns about privacy and plagiarism needing deeper study; uneven geographic coverage

The screening process can be reconstructed from the paper as a simple flow.

  1. 1209 records from database searches
  2. 242 after filtering by publication type
  3. 333 after inclusion and exclusion criteria
  4. 463 after adding a manual Google Scholar search
  5. 557 after removing six studies rated weak
Figure 2. Study selection in Baig and Yadegaridehkordi (2024), redrawn from the counts reported in the paper.

Strengths

  • Timely and transparent scope. The search strings, databases, dates and inclusion criteria are stated, so a reader can see exactly which slice of the literature is covered.
  • Broad view of stakeholders. By separating applications for academic staff, students, researchers and non-academic staff, the review moves beyond the student-only focus of many single studies and exposes how little was known about staff use.
  • Honest about the evidence base. Reporting that 30% of included studies did not state a method, and that surveys dominated, gives readers the information they need to discount weak findings.
  • Useful research agenda. The closing challenges (more qualitative work, more stakeholders, privacy and plagiarism) are specific enough to guide new projects, including doctoral ones.

Limitations and methodological issues

What the quality assessment measured

Each study was scored on four criteria: whether it explains ChatGPT, addresses higher education, is a research article and states its method (Baig & Yadegaridehkordi, 2024). These are useful relevance and reporting checks, but they do not appraise sampling, measurement validity, analysis or risk of bias. One criterion repeats an inclusion rule. The step therefore removed poorly reported papers rather than methodologically weak ones, and it is notable that studies with no stated method still passed. Readers should not treat "after quality assessment" as meaning the 57 studies are of sound quality.

Search and selection

Searching publisher platforms rather than bibliographic databases such as Scopus or Web of Science, and then adding about 30 articles from Google Scholar, makes the search hard to reproduce. The paper does not report whether two reviewers screened and scored studies independently, and it does not follow the PRISMA 2020 reporting checklist (Page et al., 2021), which would have documented reasons for exclusion at each stage. These are common gaps: a tertiary review of AI in higher education found that 65% of the evidence syntheses it examined were of critically low to medium quality, with frequent omissions in database searching, quality assessment and inter-rater reliability (Bond et al., 2024).

Counting constructs is not synthesis

The finding that social influence, perceived usefulness, effort expectancy and facilitating conditions are the leading adoption factors is based on how many studies reported each construct as significant. Such vote counting ignores sample size and effect size, and it is shaped by which constructs the original studies chose to test. Because most studies applied TAM (Davis, 1989), UTAUT (Venkatesh et al., 2003) or UTAUT2 (Venkatesh et al., 2012), the constructs of those models were bound to dominate. Most data were also self-reported intentions from students, and intention does not reliably translate into use. Bagozzi (2007) argued that TAM's parsimony leaves out much of what drives technology use, a caution that applies with force to a tool whose use is shaped by assessment rules, integrity policies and discipline norms.

A fast-moving object

The window from January 2023 to mid-January 2024 covers the first year of public access to ChatGPT, when models, institutional policies and student habits were changing month by month. Excluding conceptual papers and editorials, defensible for an empirical review, also omitted much of the early debate on assessment and integrity. The findings are best read as a dated snapshot rather than a stable description.

How the review sits in the wider literature

Earlier work on AI in higher education already warned that research was dominated by computer science and quantitative methods, with little critical reflection on risks and weak links to pedagogical theory (Zawacki-Richter et al., 2019). Baig and Yadegaridehkordi's findings suggest the first wave of ChatGPT studies repeated that pattern. A contemporaneous scoping review of 69 studies reached broadly similar conclusions about uses by teachers, students and researchers, but found that only 27% of its studies were based on empirical data and that most came from high-income countries (Ansari et al., 2024). Read together, the two reviews show how much early claims rested on commentary and small surveys.

On integrity, the review records plagiarism and privacy as concerns but does not analyse them in depth. Cotton et al. (2024), one of the most cited papers in its corpus, set out practical institutional responses such as policies, training and assessment redesign; a doctoral reader looking for guidance on integrity should go to such sources directly. For research degrees specifically, interview evidence from Australian postgraduate research students suggests that ChatGPT is already changing the division of labour in supervision (Dai et al., 2023), an area the review touches only briefly under supervision tasks. Related Academy reviews look at supervising PhD students with AI and at a hands-on doctoral course on responsible AI.

What it means for doctoral researchers

If you cite this review, cite it for what it can support: a description of what was studied in 2023, by whom and with which methods, and a list of open problems. Avoid citing it as evidence that particular factors drive adoption, or that ChatGPT improves learning, because it did not test either.

If you plan your own review, the paper shows what to add. The table contrasts what Baig and Yadegaridehkordi reported with what a stronger doctoral review would report.

Reporting a systematic review: the reviewed paper compared with good doctoral practice
Review elementIn Baig and Yadegaridehkordi (2024)What to report in your own review
Review typeSystematic literature review following Kitchenham and ChartersName the review type and justify it; a mapping or scoping design may fit an emerging field better (Gough et al., 2012)
SearchPublisher platforms plus manual Google Scholar searchFull search strings for named bibliographic databases, search dates, and how supplementary searching was done
ScreeningCounts reported; reviewer procedure not describedA PRISMA 2020 flow diagram with reasons for exclusion, and independent double screening of at least a sample
Quality appraisalFour relevance and reporting criteriaA recognised appraisal tool suited to the designs included, applied by two reviewers
SynthesisFrequency counts of significant constructsEffect sizes or a structured narrative synthesis that weighs study quality and sample size
CurrencySearch to January 2024Date the evidence explicitly and update the search before submission

Worked example (illustrative)

A hypothetical doctoral student studying how engineering lecturers use generative AI drafts this sentence: "Social influence is the strongest predictor of ChatGPT adoption in universities (Baig and Yadegaridehkordi, 2024)." The claim overstates the source. A more accurate version reads: "In an early systematic review of 57 studies, mostly student surveys, social influence was the construct most often reported as significant (Baig and Yadegaridehkordi, 2024); evidence on academic staff was scarce, which this study addresses." The revision states the evidence base, keeps the claim within what the review measured, and turns the gap into a justification for the student's own project.

Common mistakes when using systematic reviews of generative AI

  • Citing a review's frequency counts as if they were pooled effect sizes.
  • Assuming that a "quality assessment" step guarantees methodological quality without reading the criteria.
  • Treating findings about one model version and one year as a stable description of generative AI in education.
  • Generalising student survey findings to academic staff, supervisors or non-academic staff.
  • Copying a review's research gaps into your proposal without checking whether later studies have filled them.
  • Relying on one review when other syntheses of the same period reached different or more cautious conclusions.

Key takeaways

  • Baig and Yadegaridehkordi (2024) mapped 57 studies of ChatGPT in higher education from 2023 to early 2024 by trend, adoption measure, application and limitation.
  • Its quality criteria assessed relevance and reporting, so the included studies were not appraised for risk of bias.
  • The leading adoption factors it reports are frequency counts within TAM and UTAUT studies, not estimates of effect (Davis, 1989; Venkatesh et al., 2003).
  • Wider syntheses confirm a field dominated by surveys, high-income settings and weak review methods (Bond et al., 2024; Ansari et al., 2024).
  • Doctoral researchers should cite it as a dated map of early research and use PRISMA 2020 to report their own reviews (Page et al., 2021).
Your next step

Want to apply this to your own study?

  1. 1Tell us your topic, data and deadline
  2. 2A subject mentor reviews your work with you
  3. 3You get a plan and a quote
Literature review

References

  1. Ansari, A. N., Ahmad, S., & Bhutta, S. M. (2024). Mapping the global evidence around the use of ChatGPT in higher education: A systematic scoping review. Education and Information Technologies, 29(9), 11281–11321. https://doi.org/10.1007/s10639-023-12223-4
  2. Bagozzi, R. P. (2007). The legacy of the technology acceptance model and a proposal for a paradigm shift. Journal of the Association for Information Systems, 8(4), 244–254. https://doi.org/10.17705/1jais.00122
  3. Baig, M. I., & Yadegaridehkordi, E. (2024). ChatGPT in the higher education: A systematic literature review and research challenges. International Journal of Educational Research, 127, Article 102411. https://doi.org/10.1016/j.ijer.2024.102411
  4. Bond, M., Khosravi, H., De Laat, M., Bergdahl, N., Negrea, V., Oxley, E., Pham, P., Chong, S. W., & Siemens, G. (2024). A meta systematic review of artificial intelligence in higher education: A call for increased ethics, collaboration, and rigour. International Journal of Educational Technology in Higher Education, 21, Article 4. https://doi.org/10.1186/s41239-023-00436-z
  5. Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. Innovations in Education and Teaching International, 61(2), 228–239. https://doi.org/10.1080/14703297.2023.2190148
  6. Dai, Y., Lai, S., Lim, C. P., & Liu, A. (2023). ChatGPT and its impact on research supervision: Insights from Australian postgraduate research students. Australasian Journal of Educational Technology, 39(4), 74–88. https://doi.org/10.14742/ajet.8843
  7. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
  8. Gough, D., Thomas, J., & Oliver, S. (2012). Clarifying differences between review designs and methods. Systematic Reviews, 1, Article 28. https://doi.org/10.1186/2046-4053-1-28
  9. Kitchenham, B., & Charters, S. (2007). Guidelines for performing systematic literature reviews in software engineering (EBSE Technical Report EBSE-2007-01). Keele University and Durham University. https://www.scirp.org/reference/ReferencesPapers?ReferenceID=1555797
  10. Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., ... Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71
  11. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478. https://doi.org/10.2307/30036540
  12. Venkatesh, V., Thong, J. Y. L., & Xu, X. (2012). Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology. MIS Quarterly, 36(1), 157–178. https://doi.org/10.2307/41410412
  13. Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education: Where are the educators? International Journal of Educational Technology in Higher Education, 16, Article 39. https://doi.org/10.1186/s41239-019-0171-0

How to cite this article

Choose your referencing style, then copy. Check your university's or journal's guide for local variations.

PhD Research Lab Academy. (2026, January 27). ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024). PhD Research Lab. https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education/

PhD Research Lab Academy. “ChatGPT in Higher Education: A Critical Review of Baig and Yadegaridehkordi (2024).” PhD Research Lab, 27 Jan. 2026, www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education. Accessed [access date].

PhD Research Lab Academy (2026) ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024). PhD Research Lab. Available at: https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education/ (Accessed: [access date]).

PhD Research Lab Academy. “ChatGPT in Higher Education: A Critical Review of Baig and Yadegaridehkordi (2024).” PhD Research Lab. January 27, 2026. https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education.

PhD Research Lab Academy, ‘ChatGPT in Higher Education: A Critical Review of Baig and Yadegaridehkordi (2024)’ (PhD Research Lab, 27 January 2026) <https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education> accessed [access date].

PhD Research Lab Academy. ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024) [Internet]. PhD Research Lab; 2026 Jan 27 [cited [access date]]. Available from: https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education/

PhD Research Lab Academy, “ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024),” PhD Research Lab, Jan. 27, 2026. [Online]. Available: https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education. [Accessed: [access date]].

@online{critical_review_chatgpt_higher_education_2026,
  author = {{PhD Research Lab Academy}},
  title = {{ChatGPT in higher education: a critical review of Baig and Yadegaridehkordi (2024)}},
  organization = {PhD Research Lab},
  date = {2026-01-27},
  url = {https://www.phdassistance.com/academy/critical-review/critical-review-chatgpt-higher-education},
  urldate = {[access date]}
}

This article is for learning. Use it to understand the method and cite the original sources above in your own work; do not copy or submit it as your own.

ChatGPT in higher educationNeed help applying it to your study?
Literature review