Your new experience awaits. Try the new design now and help us make it even better

ORIGINAL RESEARCH article

Front. Comput. Sci.

Sec. Human-Media Interaction

This article is part of the Research TopicArtificial Intelligence for Technology Enhanced LearningView all 5 articles

The Impact of Adaptive Cognitive Diversity and Attention on Discussion Effectiveness in an Intelligent Discussion System

Provisionally accepted
Hongli  GaoHongli Gao1Bing  ZhaoBing Zhao1Xing  HuXing Hu1Chang  LiuChang Liu1Huifang  ChenHuifang Chen1Xiaohan  JiangXiaohan Jiang1Hongxing  ZhangHongxing Zhang1*Huiyu  ZhouHuiyu Zhou2
  • 1School of Psychology, Xinxiang Medical University, Xinxiang, China
  • 2School of Computing and Mathematical Sciences, University of Leicester, Leicester, United Kingdom

The final, formatted version of the article will be published soon.

Introduction: Although group discussion plays a crucial role in collaborative learning, it often falls short of achieving optimal effectiveness. The introduction of conversational agents has the potential to enhance the effectiveness of group discussion; nevertheless, the interaction strategies between conversational agents and human participants remain an issue that requires further investigation. The present study aims to examine how the diverse viewpoints provided by the conversational agent and participants' attention to them affected discussion effectiveness. Methods: This study involved 129 university students who discussed an open-ended question in an adaptive discussion system. A 2 (adaptive cognitive diversity: homogeneity vs. diversity) × 2 (attention: with vs. without instruction) between-subjects design was employed, with an additional control condition. Participants in the experimental conditions interacted with a conversational agent, while those in the control condition discussed in pairs without it. Results and discussion: The results indicated that discussions in the diversity condition exhibited greater breadth, whereas those in the homogeneity condition demonstrated significantly greater depth, suggesting that diverse perspectives promote broader idea exploration, while similar perspectives facilitate deeper elaboration. Compared with the control condition, the diversity-with-instruction demonstrated greater discussion breadth. Participants under the with-instruction condition perceived the conversational agent's viewpoints as obstructing their own idea generation; by contrast, those under the without-instruction condition generated a higher proportion of valid ideas and achieved deeper and better understanding of the discussion topic. These results suggest that attention plays both positive and negative roles in the discussion process. The present study examined the roles of adaptive cognitive diversity and attention in group discussion and explored how manipulating these factors within a human-computer interaction system can shape discussion effectiveness.

Keywords: group learning, Cognitive-Social-Motivational model, artificial intelligence ineducation, Intelligent Tutoring System, conversational agent

Received: 26 Jun 2025; Accepted: 04 Nov 2025.

Copyright: © 2025 Gao, Zhao, Hu, Liu, Chen, Jiang, Zhang and Zhou. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Hongxing Zhang, zhx166666@163.com

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.