Unraveling Downstream Gender Bias from Large Language Models: A Study on AI Educational Writing Assistance

被引:0
|
作者
Wambsganss, Thiemo [1 ]
Su, Xiaotian [2 ]
Swamy, Vinitra [2 ]
Neshaei, Seyed Parsa [2 ]
Rietsche, Roman [1 ]
Kaser, Tanja [2 ]
机构
[1] Bern Univ Appl Sci, Bern, Switzerland
[2] Ecole Polytech Fed Lausanne, Lausanne, Switzerland
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Large Language Models (LLMs) are increasingly utilized in educational tasks such as providing writing suggestions to students. Despite their potential, LLMs are known to harbor inherent biases which may negatively impact learners. Previous studies have investigated bias in models and data representations separately, neglecting the potential impact of LLM bias on human writing. In this paper, we investigate how bias transfers through an AI writing support pipeline. We conduct a large-scale user study with 231 students writing business case peer reviews in German. Students are divided into five groups with different levels of writing support: one classroom group with featurebased suggestions and four groups recruited from Prolific - a control group with no assistance, two groups with suggestions from finetuned GPT-2 and GPT-3 models, and one group with suggestions from pre-trained GPT-3.5. Using GenBit gender bias analysis, Word Embedding Association Tests (WEAT), and Sentence Embedding Association Test (SEAT) we evaluate the gender bias at various stages of the pipeline: in model embeddings, in suggestions generated by the models, and in reviews written by students. Our results demonstrate that there is no significant difference in gender bias between the resulting peer reviews of groups with and without LLM suggestions. Our research is therefore optimistic about the use of AI writing support in the classroom, showcasing a context where bias in LLMs does not transfer to students' responses(1).
引用
收藏
页码:10275 / 10288
页数:14
相关论文
共 50 条
  • [41] Learning from Mistakes via Cooperative Study Assistant for Large Language Models
    Wang, Danqing
    Li, Lei
    2023 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2023), 2023, : 10667 - 10685
  • [42] Writing medical papers using large-scale language models: a perspective from the Japanese Journal of Radiology
    Nakaura, Takeshi
    Naganawa, Shinji
    JAPANESE JOURNAL OF RADIOLOGY, 2023, 41 (05) : 457 - 458
  • [43] Writing medical papers using large-scale language models: a perspective from the Japanese Journal of Radiology
    Takeshi Nakaura
    Shinji Naganawa
    Japanese Journal of Radiology, 2023, 41 : 457 - 458
  • [44] Exploring Radiology Postgraduate Students' Engagement with Large Language Models for Educational Purposes: A Study of Knowledge, Attitudes, and Practices
    Sarangi, Pradosh Kumar
    Panda, Braja Behari
    Sanjay, P.
    Pattanayak, Debabrata
    Panda, Swaha
    Mondal, Himel
    INDIAN JOURNAL OF RADIOLOGY AND IMAGING, 2025, 35 (01): : 35 - 42
  • [45] The Promises and Pitfalls of Large Language Models as Feedback Providers: A Study of Prompt Engineering and the Quality of AI-Driven Feedback
    Jacobsen, Lucas Jasper
    Weber, Kira Elena
    AI, 2025, 6 (02)
  • [46] Clinical Knowledge and Reasoning Abilities of AI Large Language Models in Anesthesiology: A Comparative Study on the American Board of Anesthesiology Examination
    Angel, Mirana C.
    Rinehart, Joseph B.
    Cannesson, Maxime P.
    Baldi, Pierre
    ANESTHESIA AND ANALGESIA, 2024, 139 (02): : 349 - 356
  • [47] Deus Ex Machina and Personas from Large Language Models: Investigating the Composition of AI-Generated Persona Descriptions
    Salminen, Joni
    Liu, Chang
    Pian, Wenjing
    Chi, Jianxing
    Hayhanen, Essi
    Jansen, Bernard J.
    PROCEEDINGS OF THE 2024 CHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYTEMS, CHI 2024, 2024,
  • [48] 'No, Alexa, no!': designing child-safe AI and protecting children from the risks of the 'empathy gap' in large language models
    Kurian, Nomisha
    LEARNING MEDIA AND TECHNOLOGY, 2024,
  • [49] Right to Explanation in Large Language Models: Lessons from the European Union AI Act and General Data Protection Regulation
    Gorski, Lukasz
    Ramakrishna, Shashishekar
    IT PROFESSIONAL, 2025, 27 (01) : 34 - 40
  • [50] A Comparative Study of Responses to Retina Questions from Either Experts, Expert-Edited Large Language Models, or Expert-Edited Large Language Models Alone
    Tailor, Prashant D.
    Dalvin, Lauren A.
    Chen, John J.
    Iezzi, Raymond
    Olsen, Timothy W.
    Scruggs, Brittni A.
    Barkmeier, Andrew J.
    Bakri, Sophie J.
    Ryan, Edwin H.
    Tang, Peter H.
    Parke, D. Wilkin., III
    Belin, Peter J.
    Sridhar, Jayanth
    Xu, David
    Kuriyan, Ajay E.
    Yonekawa, Yoshihiro
    Starr, Matthew R.
    OPHTHALMOLOGY SCIENCE, 2024, 4 (04):