Factors Predicting Performance and Motivation in ChatGPT-supported Asynchronous Online Discussion

ChatGPT’s comprehensive, real-time, and profile integration features show significant potential for use in asynchronous online discussions (AOD). However, there is a lack of effective literature on how to integrate ChatGPT into these discussions. Therefore, this study explores factors related to ChatGPT that may predict students’ performance and motivation in AOD. Using the Xiaoya platform, 104 undergraduate students participated in a one-week online case discussion with ChatGPT as an assistant. Data on 10 potential independent variables and 3 dependent variables were collected and analyzed using multiple linear regression. The findings indicate that perceived usefulness and the quality of prompts given to ChatGPT significantly and positively influence AOD motivation. Daily use, interaction frequency, prompt quality, copy rate, and GPA significantly and positively influence the number of posts, whereas perceived usefulness negatively influences it. Additionally, prompt quality, copy rate, and daily use significantly and positively predict post quality, while previous knowledge with ChatGPT negatively predicts post quality. These results highlight the critical factors that educators should consider to enhance the effectiveness of ChatGPT in AOD.