{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T05:51:43Z","timestamp":1791525103039,"version":"4.3.4"},"reference-count":23,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Large language models (LLMs)-generated texts become increasingly similar to human writing. The purpose of this study was to examine (1) how individuals distinguish between human-written and LLM-generated texts, focusing on their judgment strategies; (2) the relationship between judgment strategies and detection accuracy; and (3) the interaction effects of gender, age, and frequency of large language model (LLM) use.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Participants were asked to judge whether the given texts (e.g., public comments) were written by LLMs or humans and to explain their reasoning using open-ended questions. The public comments were written by humans and generated using seven LLMs (ChatGPT [4o and o1], Claude3.5, Gemini, Microsoft Copilot, Llama3.1, and Perplexity). First, we classified participants\u2019 explanations of reasoning related to judgments into types using BERT-based clustering (i.e., Sentence-BERT and hierarchical cluster analysis).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>BERT-based clustering suggested four interpretable clusters. Based on their characteristic patterns, we interpreted these clusters as four hypothetical judgment strategy types and tentatively proposed that they form a conceptual continuum: (1) Unsubstantiated intuition, (2) Stylistic-Cue, (3) Syntactic-Cue, and (4) Higher-Order Pattern Recognition (HOPR). A hierarchical Bayesian logistic regression analysis revealed that the detection accuracy varied across judgment types, with the HOPR type achieving relatively higher performance than the other strategy types, although its detection accuracy remained near the chance level. Additionally, we identified interaction effects between strategy type and gender; only female participants who used the HOPR strategy showed lower detection accuracy than male participants, although the qualitative interpretation of this difference warrants further empirical investigation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>These findings emphasize the role of judgment strategies and individual differences in discriminating between human and LLM-generated content.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1918939","type":"journal-article","created":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T05:30:24Z","timestamp":1791523824000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["How different judgment strategies affect accuracy in generative AI text discrimination: BERT-based clustering and hierarchical Bayesian logistic regression"],"prefix":"10.3389","volume":"9","author":[{"given":"Wataru","family":"Zaitsu","sequence":"first","affiliation":[{"name":"Faculty of Psychology, Mejiro University","place":["Tokyo, Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mingzhe","family":"Jin","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence, Jilin International Studies University","place":["Changchun, Jilin, China"]},{"name":"Institute of Interdisciplinary Research, Kyoto University of Advanced Science","place":["Kyoto, Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shunichi","family":"Ishihara","sequence":"additional","affiliation":[{"name":"Faculty of Psychology, Mejiro University","place":["Tokyo, Japan"]},{"name":"Speech and Language Laboratory, Australian National University","place":["Canberra, ACT, Australia"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Satoru","family":"Tsuge","sequence":"additional","affiliation":[{"name":"School of Informatics, Daido University","place":["Nagoya, Aichi, Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mitsuyuki","family":"Inaba","sequence":"additional","affiliation":[{"name":"College of Policy Science, Ritsumeikan University","place":["Kyoto, Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,10,9]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1080\/03610927408827101","article-title":"A dendrite method for cluster analysis","volume":"3","author":"Calinski","year":"1974","journal-title":"Commun. Stat."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1037\/h0046743","article-title":"Theory of fluid and crystallized intelligence: a critical experiment","volume":"54","author":"Cattell","year":"1963","journal-title":"J. Educ. Psychol."},{"key":"ref3","doi-asserted-by":"publisher","first-page":"25989","DOI":"10.1038\/s41598-024-76218-y","article-title":"Human intelligence can safeguard against artificial intelligence: individual differences in the discernment of human from AI texts","volume":"14","author":"Chein","year":"2024","journal-title":"Sci. Rep."},{"key":"ref4","first-page":"7282","author":"Clark","year":"2021"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"101426","DOI":"10.1016\/j.xcrp.2023.101426","article-title":"Distinguishing academic science writing from humans or ChatGPT with over 99% accuracy using off-the-shelf machine learning tools","volume":"4","author":"Desaire","year":"2023","journal-title":"Cell Rep. Phys. Sci."},{"key":"ref6","author":"Devlin","year":"2019"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"100209","DOI":"10.1016\/j.caeai.2024.100209","article-title":"Do teachers spot AI? Evaluating the detectability of AI-generated texts among student essays","volume":"6","author":"Fleckenstein","year":"2024","journal-title":"Comput. Educ. Artif. Intell."},{"key":"ref8","volume-title":"Clustering Algorithms","author":"Hartigan","year":"1975"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"e2208839120","DOI":"10.1073\/pnas.2208839120","article-title":"Human heuristics for AI-generated language are flawed","volume":"120","author":"Jakesch","year":"2023","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"ref10","volume-title":"Thinking, Fast and Slow","author":"Kahneman","year":"2011"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"106553","DOI":"10.1016\/j.chb.2020.106553","article-title":"Artificial intelligence versus Maya Angelou: experimental evidence that people cannot differentiate AI-generated from human-written poetry","volume":"114","author":"K\u00f6bis","year":"2021","journal-title":"Comput. Hum. Behav."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1111\/j.1460-2466.2000.tb02833.x","article-title":"The limited capacity model of mediated message processing","volume":"50","author":"Lang","year":"2000","journal-title":"J. Commun."},{"key":"ref13","author":"MacQueen","year":"1967"},{"key":"ref14","year":"2022"},{"key":"ref15","article-title":"Sentence-BERT: sentence embeddings using Siamese BERT-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)","author":"Reimers","year":"2019"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"53","DOI":"10.1016\/0377-0427(87)90125-7","article-title":"Silhouettes: a graphical aid to the interpretation and validation of cluster analysis","volume":"20","author":"Rousseeuw","year":"1987","journal-title":"J. Comput. Appl. Math."},{"key":"ref17","author":"Russell","year":"2025"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"2573","DOI":"10.1161\/STROKEAHA.124.045012","article-title":"Reviewer experience detecting and judging human versus artificial intelligence content: the stroke journal essay contest","volume":"55","author":"Silva","year":"2024","journal-title":"Stroke"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"411","DOI":"10.1111\/1467-9868.00293","article-title":"Estimating the number of clusters in a data set via the gap statistic","volume":"63","author":"Tibshirani","year":"2001","journal-title":"J. R. Stat. Soc. B Stat. Methodol."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"20220158","DOI":"10.1515\/opis-2022-0158","article-title":"The effectiveness of software designed to detect AI-generated writing: a comparison of 16 AI text detectors","volume":"7","author":"Walters","year":"2023","journal-title":"Open Inf. Sci."},{"key":"ref21","doi-asserted-by":"publisher","first-page":"26","DOI":"10.1007\/s40979-023-00146-z","article-title":"Testing of detection tools for AI-generated text","volume":"19","author":"Weber-Wulff","year":"2023","journal-title":"Int. J. Educ. Integr."},{"key":"ref22","doi-asserted-by":"publisher","first-page":"e0288453","DOI":"10.1371\/journal.pone.0288453","article-title":"Distinguishing ChatGPT(\u22123.5, \u22124)-generated and human-written papers through Japanese stylometric analysis","volume":"18","author":"Zaitsu","year":"2023","journal-title":"PLoS One"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"e0335369","DOI":"10.1371\/journal.pone.0335369","article-title":"Stylometry can reveal artificial intelligence authorship, but humans struggle: a comparison of human and seven large language models in Japanese","volume":"20","author":"Zaitsu","year":"2025","journal-title":"PLoS One"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1918939\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T05:30:26Z","timestamp":1791523826000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1918939\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,9]]},"references-count":23,"alternative-id":["10.3389\/frai.2026.1918939"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1918939","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,9]]},"article-number":"1918939"}}