Exploring Learners' Interactions with GenAI Agents in Educational Games: Typologies and Emotional Factors in Human-Computer Interaction
Abstract
This study explores and compares learners' interactions with Generative Artificial Intelligence (GenAI) within an educational game context. Data from 365 dialogues were collected from 76 K12 students at a middle school in Suzhou, China, who had conversations with GenAI agents in an educational game environment. Drawing upon a framework adapted from the classification system of learners' online asynchronous discussions, six categories were utilized: social and affective (ES), information (Inf), questions (Q&A), ideas (ID), agency (EA), and meta-recapitulation (MA). The study revealed several important findings. Firstly, students' responses contained less information (Inf) and meta-recapitulation (MA) due to the limitations of GenAI in providing authoritative and external resources of information. Secondly, players’ social and affective outputs (ES) were associated with mixed emotions, with the majority of emotions experienced by the students being negative (82.02%), and the reasons behind these negative emotions include mistrust of GenAI during or prior to the game, GenAI misinterpreting players’ ideas, a mismatch with players’ expectations, and GenAI repeatedly outputting the same text. Overall, this study sheds light on the complex dynamics of learners' dialogues with GenAI in educational environment, highlighting the learning opportunities and challenges for improving AI-driven educational experiences.
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?
A second focus is emotion recognition and regulation in virtual environments, including the development of prosocial behavior and empathy.
“Asecond focus is emotion recognition and regulation in virtual environments, including the development of prosocial behavior and empathy [121,122].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [122]
In science and outreach games, AI operates as a non-player character or assistant, providing stepwise hints, rule explanations, and plot advancement.
“In science and outreach games, AI operates as a non-player character or assistant, providing stepwise hints, rule explanations, and plot advancement.”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [122]
Studies document limited understanding of model training and datasets, which impairs students' ability to detect errors or deceptive content and leads to weak knowledge/analysis or no significant improvement.
“Studies document limited understanding of model training and datasets [112,122,135,136], which impairs students' ability to detect errors or deceptive content [105,133] and leads to weak knowledge/analysis [109] or no significant improvement [98].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [122]
Motivation gains are not guaranteed: several studies find no significant effects or only short-lived novelty bumps that taper over time.
“Motivation gains are not guaranteed: several studies find no significant effects [109] or only short-lived novelty bumps that taper over time [82,122].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [122]
The citation lists haven’t been fetched yet
References 19 · Cited by 7
- Authors
- Chen, Z., Xiong, Z., Ruan, X., Jiang, S., Wei, W., & Fang, K.
- Affiliations
- Macao Polytechnic University (MO)
- Tsinghua Shenzhen International Graduate School (CN)
- Venue
- 2024 4th International Conference on Educational Technology (ICET)
- Publisher
- IEEE
- Published
- 2024-09-13
- Pages
- 136-140
- OpenAlex
- W4407215353
- Cited by
- 7
- References
- 19
- Citations by year
- 2026: 3 · 2025: 4
- FWCI
- 2.02
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- International Conference on Educational Technology (ICET) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
- Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
- Social Sciences › Psychology › Developmental and Educational Psychology › Educational Games and Gamification
- Keywords
- game-based learning, affective learning, K-12 students, junior high school students, human-AI interaction, learner agency
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
