Skip to main content

Presentation

·466 words·3 mins· loading · loading ·
Xuanyu Chen
Author
Xuanyu Chen
M.Phil Student @Fudan University
Table of Contents

2026
#

  • AERA Annual Meeting 2026, Los Angeles, California, USA AERA “Learners’ Collaboration With AI in Enhancing Peer Feedback in Argumentative Writing: An Exploratory Study” (Submitted, with Wenli Chen, Qianru Lyu, Lishan Zheng). Human-AI Collaboration
    Details

    Peer feedback is an effective pedagogical strategy for improving argumentative writing, yet the integration of generative AI as a collaborator into this process remains underexplored. This study examines how AI-directed (explicit revisions) and AI-assisted (implicit suggestions and encouragement) peer feedback affect argumentative writing and feedback characteristics through an experiment among 90 university students. Content analysis revealed that AI-supported peer feedback enhances elaboration of writing. Epistemic Network Analysis showed that AI-assisted condition fostered integration of cognitive (problem identification, elaboration) and affective (mitigation, hedging) elements, whereas AI-directed condition yielded more structured, task-focused revisions. These findings highlight the impacts of AI support on peer feedback characteristics and argumentative writing, underscoring the importance of cognitive efficiency and socio-emotional engagement in AI-supported peer feedback.

    Automation figure

    AERA: American Educational Research Association.

2025
#

  • ISLS Annual Meeting 2025, Helsinki, Finland ISLS

    ISLS: International Society of the Learning Sciences.

  • AERA Annual Meeting 2025, Denver, Colorado, USA AERA Poster (Accepted): “Generative Artificial Intelligence and Job Displacement: A Perspective from College Entrance Examination” (with Peikang Zhang). Economics of Education

    Details

    As global economy transitions into digital era, generative AI is sparking concerns over job displacement and triggering risk-averse decision-making among college entrance exam candidates regarding their major choices, to whom existing research pays inadequate attention. This study introduces a novel perspective by using O*NET, NCEE, and population surveys to calculate major-level employment substitution rates. We reveal that the emergence of ChatGPT notably decreased admission scores due to heightened job displacement risks, particularly in STEM majors and Tier 1 colleges, and further investigate its demand-supply mechanism in job postings and graduate surveys. Our findings withstand extensive robustness checks, offering new insights into the dynamics of education and labor markets, enhancing human capital theory, and informing policies for optimizing educational resource allocations.

    Automation figure

2024
#

  • BERA Annual Meeting and WERA Focal Meeting 2024, Manchester, UK BERA Individual Paper, Session Chair: “Research on the Influence Mechanism of Self-Regulated Learning Behaviors on Second Language Learning”. Learning Analytics

    BERA: British Educational Research Association;
    WERA: World Educational Research Association.

2021
#

  • International Conference of AITELL 2021, Shanghai, China AITELL Individual Paper: “The Effect of Study Duration and Learning Initiative on Student Completion of Online Courses”. Learning Analytics

    AITELL: Artificial Intelligence and Technology-Enhanced Language Learning.