Academic Writing with GPT-3.5 (ChatGPT): Reflections on Practices, Efficacy and Transparency
Open accessAbstract
The debate around the use of GPT-3.5 has been a popular topic among academics since the release of ChatGPT. Whilst some have argued for the advantages of GPT-3.5 in enhancing academic writing, others have raised concerns such as plagiarism, the spread of false information, and ecological issues. The need for finding ways to use GPT-3.5 models transparently has been voiced, and suggestions have been made on social media as to how to use GPT-3.5 models in a smart way. Nevertheless, to date, there is a lack of literature which clearly outlines how to use GPT-3.5 models in academic writing, how effective they are, and how to use them transparently. To address this, I conducted a personal experience experiment with GPT-3.5, specifically by using text-davinci-003 model of OpenAI, for writing this article. I identified six ways of using GPT-3.5: Chunk Stylist, Bullet-to-Paragraph, Talk Textualizer, Research Buddy, Polisher and Rephraser. I reflected on their efficacy, and commented on their potential impact on writing ethics. Additionally, I provided a comprehensive document which shows the prompts I used, results I got from GPT-3.5, the final edits and visually compares those by showing the differences in percentage.
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.
Integrity signals
Registry
Cites
Statistics
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.
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?
In the General context, systems commonly provided suggestions for revisions which could be integrated directly into the writing area, aiding efficiency and idea exploration in the text.
“Systems commonly provided suggestions for revisions which could be integrated directly into the writing area [14,18,75], aiding efficiency and idea exploration in the text.”
From Co-Writing with AI, on Human Terms: Aligning Research with User Demands Across the Writing Process (2025) · cites this paper as [14]
Section 4.2, “Research Buddy” (page unavailable), says GPT-3.5 can suggest ideas and help writers explore them efficiently; Section 4.1 describes using it to produce text quickly. But the paper describes one author’s use of GPT-3.5, not systems commonly offering revision suggestions that can be integrated directly into a writing area. Section 5.6 proposes GPT-supported text editors as a future tool rather than reporting that feature in existing systems. (LLM, conf=0.84)
Experimental — accuracy not yet measuredStudies found that suggestion length is inversely associated with perceived ownership of the text.
“Other studies have examined the impact of suggestion length or quantity of suggestions on user experience and acceptance [14,29,36], finding that suggestion length is inversely associated with perceived ownership of the text.”
From Co-Writing with AI, on Human Terms: Aligning Research with User Demands Across the Writing Process (2025) · cites this paper as [14]
Section 5.4, “Transparency, Agency and Biases” (page unavailable), reports the author’s uncertainty about their agency over some GPT-3.5-generated text. It does not examine suggestion length or report an inverse association between length and perceived ownership. (LLM, conf=0.78)
Experimental — accuracy not yet measuredSome systems present suggestions directly in the user's workspace, creating a more integrated experience but potentially blurring boundaries between user and AI contributions.
“Some systems present suggestions directly in the user's workspace [8,14,18], creating a more integrated experience but potentially blurring boundaries between user and AI contributions.”
From Co-Writing with AI, on Human Terms: Aligning Research with User Demands Across the Writing Process (2025) · cites this paper as [14]
Section 5.4 (page not provided) describes uncertainty about the author's agency over GPT-generated text, supporting the concern about blurred contributions. Section 5.6 proposes GPT-supported text editors that show prompts, outputs, and edits, but does not say that existing systems present suggestions directly in the user's workspace or that this creates a more integrated experience. (LLM, conf=0.82)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
ChatGPT is fun, but not an author
Thorp, H. H. (2023) · Science · doi:10.1126/science.adg7879 · Cited by 1,437 · see citing works
Plagiarism in the age of massive generative pre-trained transformers (GPT-3)
Dehouche, N. (2021) · Ethics in Science and Environmental Politics · doi:10.3354/esep00195 · Cited by 296 · see citing works
HS Kumar, A. (2023) · Biology, Engineering, Medicine and Science Reports · doi:10.5530/bems.9.1.5 · Cited by 165 · cited 1× · see citing works
“Another study assessed the efficacy of ChatGPT for Biomedical Writing by giving short prompts [20], and found that the generated text lacks depth, yet ChatGPT has an enormous potential to be used in academic writing with improvements.”
Background
Flagged first, then by citations
Research works printed but not linked: 22
- 2020. Next Chapter in Artificial Writing. Nature Machine Intelligence 2, 8 (Aug 2020), 419-419. https://doi.org/10.1038/s42256-020-0223-0 · DOI
- 2021. Compare Text Online -Show Text Difference. https://countwordsfree. com/comparetexts [] 2023. Nature 613, 7945 (Jan 2023), 612-612. https://doi.org/10.1038/d41586-023- 00191-1 · DOI · Paper page
- Susanne Bødker. 2006. When second wave HCI meets third wave challenges. In Proceedings of the 4th Nordic conference on Human-computer interaction: changing roles. 1-8.
- Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. Language models are few-shot learners. Advances in neural information processing systems 33 (2020), 1877-1901.
- Oğuz'Oz' Buruk. 2023. Academic Writing with GPT-3.5: Reflections on Practices, Efficacy and Transparency. arXiv preprint arXiv:2304.11079 (2023). · arXiv
- Oğuz 'Oz' Buruk, Louise Petersen Matjeka, and Florian 'Floyd' Mueller. 2023. Towards Designing Playful Bodily Extensions: Learning from Expert Interviews. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI '23). Association for Computing Machinery, New York, NY, USA, Article 233, 20 pages. https://doi.org/10.1145/3544548.3581165 · DOI
- Ke-Li Chiu, Annie Collins, and Rohan Alexander. 2021. Detecting hate speech with gpt-3. arXiv preprint arXiv:2103.12407 (2021). · arXiv
- Vanya Cohen and Aaron Gokaslan. 2020. OpenGPT-2: Open Language Models and Implications of Generated Text. XRDS 27, 1 (Sep 2020), 26-30. https: //doi.org/10.1145/3416063 · DOI
- Audrey Desjardins and Aubree Ball. 2018. Revealing tensions in autobiographical design in HCI. In proceedings of the 2018 designing interactive systems conference. 753-764.
- Vijay Kumar Jain, Karthikeyan P Iyengar, and Raju Vaishya. 2022. Is the English language a barrier to the non-English-speaking authors in academic publishing? Postgraduate Medical Journal 98, 1157 (2022), 234-235. https://doi.org/10.1136/ postgradmedj-2020-139243 · DOI
- Tiffany H Kung, Morgan Cheatham, Arielle Medinilla, ChatGPT, Czarina Sillos, Lorie De Leon, Camille Elepano, Marie Madriaga, Rimel Aggabao, Giezel Diaz- Candido, et al. 2022. Performance of ChatGPT on USMLE: Potential for AI- Assisted Medical Education Using Large Language Models. medRxiv (2022), 2022-12.
- Andrés Lucero, Audrey Desjardins, Carman Neustaedter, Kristina Höök, Marc Hassenzahl, and Marta E Cecchinato. 2019. A sample of one: First-person research methods in HCI. In Companion Publication of the 2019 on Designing Interactive Systems Conference 2019 Companion. 385-388.
- Kris McGuffie and Alex Newhouse. 2020. The radicalization risks of GPT-3 and advanced neural language models. arXiv preprint arXiv:2009.06807 (2020). · arXiv
- Albert Meroño-Peñuela, Dayana Spagnuelo, and GPT-2. 2020. Can a Transformer Assist in Scientific Writing? Generating Semantic Web Paper Snippets with GPT- 2. In The Semantic Web: ESWC 2020 Satellite Events: ESWC 2020 Satellite Events, Heraklion, Crete, Greece, May 31-June 4, 2020, Revised Selected Papers 17. Springer, 158-163.
- Carman Neustaedter and Phoebe Sengers. 2012. Autobiographical Design in HCI Research: Designing and Learning through Use-It-Yourself. In Proceedings of the Designing Interactive Systems Conference (DIS '12). Association for Computing Ma- chinery, New York, NY, USA, 514-523. https://doi.org/10.1145/2317956.2318034 · DOI
- June Oshiro, Suzanne L. Caubet, Kelly E. Viola, and Jill M. Huber. 2020. Going Beyond "Not Enough Time": Barriers to Preparing Manuscripts for Academic Medical Journals. Teaching and Learning in Medicine 32, 1 (2020), 71-81. https: //doi.org/10.1080/10401334.2019.1659144 · DOI · Paper page
- Corina Sas. 2019. First person HCI research: Tapping into designers' tacit ex- periences. In ACM Designing Interactive Systems Conference: 1st Person Research Methods in HCI Workshop.
- Jesse Schell. 2014. The Art of Game Design: A book of lenses. AK Peters/CRC Press.
- Tobias Schmohl, Alice Watanabe, Nadine Fröhlich, Dominikus Herzberg, et al. 2020. How Artificial Intelligence can improve the Academic Writing of Students. In Conference Proceedings. The Future of Education 2020.
- Chris Stokel-Walker and Richard Van Noorden. 2023. What ChatGPT and gen- erative AI mean for science. Nature 614, 7947 (Feb 2023), 214-216. https: //doi.org/10.1038/d41586-023-00340-6 · DOI · Paper page
- Angel Urbina-Garcia. 2020. What do we know about university academics' mental health? A systematic literature review. Stress and Health 36, 5 (Jul 2020), 563-585. https://doi.org/10.1002/smi.2956 · DOI
- Jan Philip Wahle, Terry Ruas, Frederic Kirstein, and Bela Gipp. 2022. How Large Language Models are Transforming Machine-Paraphrase Plagiarism. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, Abu Dhabi, United Arab Emirates, 952-963. https://aclanthology.org/2022.emnlp-main.62
Cited by
Sorted by citations
Lee, H.-P. (., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713778 · Cited by 490 · 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 · see citing works
Co-Writing with AI, on Human Terms: Aligning Research with User Demands Across the Writing Process
Reza, M., Thomas-Mitchell, J., Dushniku, P., Laundry, N., Williams, J. J., & Kuzminykh, A. (2025) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3757566 · Cited by 37 · see citing works
“…processes and contexts–for example, strong representation of Generating and Translating activities in Creative and Academic settings, and limited focus on Monitoring across all contexts. process across them; we also report how many papers 1 were coded into each writing context, as shown in Table 1.”
- Authors
- Buruk, O. '.
- Affiliations
- Tampere University (FI)
- Corresponding author
- Buruk, Oğuz 'Oz'
- ORCID
- Buruk, Oğuz 'Oz'
- Venue
- 26th International Academic Mindtrek Conference
- Publisher
- ACM
- Published
- 2023-10-03
- Pages
- 144-153
- OpenAlex
- W4388220988
- Cited by
- 46
- References
- 3
- Citations by year
- 2026: 7 · 2025: 16 · 2024: 22 · 2023: 1
- FWCI
- 11.83
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- International Academic Mindtrek Conference (Mindtrek) (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion, cc-by)
- Tampere University Institutional Repository (Tampere University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Topics
- Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Academic Writing and Publishing
- Social Sciences › Safety Research › Academic integrity and plagiarism
- Keywords
- GPT-3.5, ChatGPT, academic writing, large language models, text generation, prompt engineering, ethics of writing, transparency, plagiarism, AI-assisted writing, text editing
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
