The experience–intention disconnect in AI chatbot feedback: evidence from higher education

Abstract

Generative AI chatbots are increasingly used as interactive formative feedback tools in higher education, yet it remains unclear whether positive feedback experiences translate into students’ intentions for continued academic use. This study examines an instructor-designed generative AI chatbot embedded in a project-based undergraduate digital marketing course. Data were collected from 100 students who interacted with a course-embedded chatbot that provided criteria-based feedback on campaign presentations. Survey measures assessed students’ cognitive perceptions of feedback quality, emotional perceptions of support, and continued-use intention. Multiple and simple regression analyses showed that students evaluated the chatbot feedback positively in terms of clarity, usefulness, guidance, support, respect, and empathy. The findings suggest a possible experience–intention disconnect, defined here not as low continued-use intention, but as a lack of statistically reliable association between positive feedback evaluations and variation in students’ future-use intentions. This pattern should be interpreted cautiously given the marginal reliability of the continued-use intention measure and the limited dispersion of the study variables.