This study documents accounting students’ perceptions of ChatGPT's educational value by examining how reported integration of ChatGPT into study activities is associated with students’ self-reported motivation and perceived academic performance. Self-Determination Theory (SDT) with the Interactive-Constructive-Active-Passive (ICAP) are used as interpretive frameworks to discuss why students may perceive motivational and learning related benefits, without claiming that underlying SDT needs or ICAP engagement modes were directly measured. All variables were measured via self-report; therefore, the results reflect perceived outcomes and correlational patterns rather than objective learning gains.
A cross-sectional survey of accounting students from multiple universities captured use of ChatGPT, motivational states, and perceived academic performance on Likert-type scales. Measurement and structural models were estimated in AMOS. Model fit and construct validity were assessed using established criteria.
Reported ChatGPT integration was positively associated with self-reported motivation and with perceived academic performance. Motivation was also positively associated with perceived performance and partially mediated the association between ChatGPT integration and perceived performance. Because the design is cross sectional and self-report based, these results indicate perceptions and correlational patterns rather than demonstrated learning gains or causal effects.
Embedding ChatGPT within scaffolded problem sets and formative assessments may help sustain motivation and support students’ perceived conceptual understanding in rule-intensive accounting contexts. Clear academic integrity guidelines and staff development may be important to channel use toward constructive and interactive engagement.
The study offers a theoretically grounded account of how conversational AI is associated with students’ perceived academic benefits in accounting. By linking SDT as an interpretive lens to ICAP informed engagement pathways (rather than directly measured ICAP levels), it helps explain how any perceived benefits of ChatGPT may be realised chiefly through motivational mechanisms rather than content delivery alone.
