Does ChatGPT Write Like a Student? Engagement Markers in Argumentative Essays
Abstract
ChatGPT has created considerable anxiety among teachers concerned that students might turn to large language models (LLMs) to write their assignments. Many of these models are able to create grammatically accurate and coherent texts, thus potentially enabling cheating and undermining literacy and critical thinking skills. This study seeks to explore the extent LLMs can mimic human-produced texts by comparing essays by ChatGPT and student writers. By analyzing 145 essays from each group, we focus on the way writers relate to their readers with respect to the positions they advance in their texts by examining the frequency and types of engagement markers. The findings reveal that student essays are significantly richer in the quantity and variety of engagement features, producing a more interactive and persuasive discourse. The ChatGPT-generated essays exhibited fewer engagement markers, particularly questions and personal asides, indicating its limitations in building interactional arguments. We attribute the patterns in ChatGPT’s output to the language data used to train the model and its underlying statistical algorithms. The study suggests a number of pedagogical implications for incorporating ChatGPT in writing instruction.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
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?
The launch of ChatGPT is widely regarded as a milestone in the mainstream adoption of GenAI.
“Although ChatGPT was not the first publicly available AI chatbot, we used its launch as the cut-off point because it is widely regarded as a milestone in the mainstream adoption of GenAI 22,23,24,25 .”
From Style, sentiment, and quality of undergraduate writing in the AI era: A cross-sectional and longitudinal analysis of 4,820 authentic empirical reports (2025) · cites this paper as [25]
Cites
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · see citing works
ChatGPT: Bullshit spewer or the end of traditional assessments in higher education?
(2023) · Journal of Applied Learning & Teaching · doi:10.37074/jalt.2023.6.1.9 · Cited by 1,806 · cited 2× · see citing works
“For decades, conversational agents, commonly referred to as chatbots, have been applied to support teaching of foreign languages (Ji et al., 2023; Rudolph et al., 2023).”
ChatGPT for Language Teaching and Learning
Kohnke, L., Moorhouse, B. L., & Zou, D. (2023) · RELC Journal · doi:10.1177/00336882231162868 · Cited by 1,371 · cited 1× · see citing works
“For some, the language modeling capabilities of ChatGPT can help scaffold students’ language study (e.g., Kasneci et al., 2023; Kohnke et al., 2023).”
Using ChatGPT for second language writing: pitfalls and potentials
Barrot, J. S. (2023) · Assessing Writing · doi:10.1016/j.asw.2023.100745 · Cited by 770 · cited 1× · see citing works
“Its potential to assist writing rests on a capacity to generate writing that is “typi-cally coherent and grammatically correct” (Barrot, 2023, p. 2) and to refine the tone and style of a text (Ji et al., 2023).”
The opportunities and challenges of ChatGPT in education
Adeshola, I., & Adepoju, A. P. (2024) · Interactive Learning Environments · doi:10.1080/10494820.2023.2253858 · Cited by 517 · cited 2× · see citing works
“Although tools such as GPTZero and AICheatCheck have been developed to detect AI involvement in writing, these seem currently unable to make a reliable distinction (Adeshola & Adepoju, 2024; Scarfe et al., 2024).”
Bringing in the Reader: Addressee Features in Academic Articles
HYLAND, K. (2001) · Written Communication · doi:10.1177/0741088301018004005 · Cited by 375 · see citing works
Self-citationCasal, J. E., & Kessler, M. (2023) · Research Methods in Applied Linguistics · doi:10.1016/j.rmal.2023.100068 · Cited by 261 · see citing works
What do they mean? Questions in academic writing
HYLAND, K. (2002) · Text - Interdisciplinary Journal for the Study of Discourse · doi:10.1515/text.2002.021 · Cited by 201 · cited 2× · see citing works
“Questions, then, are the strategy of dialogic involvement par excellence, serving up an invitation for readers to orientate themselves in a certain way to the argument presented and to enter a frame of discourse where they can be led to the writer’s view-point (Hyland, 2002).”
Self-citationAI-generated vs human-authored texts: A multidimensional comparison
Berber Sardinha, T. (2024) · Applied Corpus Linguistics · doi:10.1016/j.acorp.2023.100083 · Cited by 139 · see citing works
Generative AI for corpus approaches to discourse studies: A critical evaluation of ChatGPT
Curry, N., Baker, P., & Brookes, G. (2024) · Applied Corpus Linguistics · doi:10.1016/j.acorp.2023.100082 · Cited by 102 · cited 1× · see citing works
“Curry et al. (2024), for example, observed that ChatGPT sometimes fabricated questions in their corpus by adding question marks or question tags to declarative statements, resulting in inaccurate questions.”
Does ChatGPT Argue Like Students? Bundles in Argumentative Essays
Jiang, F. (., & Hyland, K. (2025) · Applied Linguistics · doi:10.1093/applin/amae052 · Cited by 76 · cited 2× · see citing works
“Interestingly, Jiang and Hyland (2024) similarly identified significantly fewer 3-word stance bundles (e.g., it is possible, this never is, in my opinion ) in the ChatGPT essays than used by human writers.”
Self-citationHow ChatGPT is transforming the postdoc experience
Nordling, L. (2023) · Nature · doi:10.1038/d41586-023-03235-8 · Cited by 71 · cited 2× · see citing works
“Yet it is seen by students and researchers as most useful when assisting with writing: to plan, write, edit, and polish academic texts (Ingley & Pack, 2023; Nordling, 2023).”
Rhetorical distinctions: Comparing metadiscourse in essays by ChatGPT and students
Jiang, F. (., & Hyland, K. (2025) · English for Specific Purposes · doi:10.1016/j.esp.2025.03.001 · Cited by 44 · cited 1× · see citing works
“This “audience blindness” can result in output that, while grammatically conventional and topically relevant, may lack the nuanced involvement features that characterize effective human writing (Markey et al., 2024; Jiang & Hyland, 2025).”
Self-citation
Flagged first, then by citations
Cited by
Sorted by citations
Rhetorical distinctions: Comparing metadiscourse in essays by ChatGPT and students
Jiang, F. (., & Hyland, K. (2025) · English for Specific Purposes · doi:10.1016/j.esp.2025.03.001 · Cited by 44 · see citing works
Self-citationMak, M. H. C., & Walasek, L. (2025) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2025.100507 · Cited by 1 · see citing works
“Future work should therefore investigate the extent to which GenAI tools influence the epistemic diversity of student writing (Jiang & Hyland, 2025).”
- Authors
- Jiang, F. (., & Hyland, K.
- Affiliations
- Beihang University (CN)
- University of East Anglia (GB)
- Corresponding author
- Jiang, Feng (Kevin)
- Venue
- Written Communication
- Publisher
- SAGE Publications
- Published
- 2025-04-30
- Volume
- 42
- Issue
- 3
- Pages
- 463-492
- ISSN
- 0741-0883 (ISSN-L)
- 1552-8472
- OpenAlex
- W4410632103
- Cited by
- 52
- References
- 13
- Citations by year
- 2026: 34 · 2025: 18
- FWCI
- 517.84
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Written Communication (publishedVersion)
- UEA Digital Repository (University of East Anglia) (acceptedVersion)
- Topics
- Social Sciences › Arts and Humanities › Literature and Literary Theory › Discourse Analysis in Language Studies
- Social Sciences › Arts and Humanities › Philosophy › Linguistics and Discourse Analysis
- Social Sciences › Arts and Humanities › Literature and Literary Theory › Linguistics and Education Research
- Keywords
- ChatGPT, large language models, argumentative essay, engagement markers, student writing, AI-generated text, reader engagement, rhetorical questions, persuasive discourse, writing instruction, academic integrity
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
