Paper Type
ERF
Abstract
Higher education students often struggle to navigate dispersed course materials, increasing search effort and extraneous cognitive load. We introduce CourseGPT, a course-bounded generative AI assistant built on ChatGPT's custom GPT feature and grounded exclusively in instructor-provided materials. In an exploratory pilot evaluation conducted in an undergraduate information systems course, we examined changes in student information-seeking behaviors and perceptions before and after mid-semester introduction of the tool. Using pre- and post-introduction questionnaires (n=52 paired responses), we observed reductions in self-reported search time, effort and perceived mental load, alongside an unexpected decline in perceived understanding of course policies that we interpret as preliminary evidence of a cognitive offloading trade-off. Findings are consistent with Information Foraging Theory and Cognitive Load Theory and motivate a scaled comparison group study.
Paper Number
1452
Recommended Citation
Mills, Stefan A.; Wang, Xuan; and Owusu, Gabriel, "Reducing Routine Friction: The Impact of Course-Specific Generative AI on Student Information Seeking" (2026). AMCIS 2026 Proceedings. 12.
https://aisel.aisnet.org/amcis2026/sig_ed/sig_ed/12
Reducing Routine Friction: The Impact of Course-Specific Generative AI on Student Information Seeking
Higher education students often struggle to navigate dispersed course materials, increasing search effort and extraneous cognitive load. We introduce CourseGPT, a course-bounded generative AI assistant built on ChatGPT's custom GPT feature and grounded exclusively in instructor-provided materials. In an exploratory pilot evaluation conducted in an undergraduate information systems course, we examined changes in student information-seeking behaviors and perceptions before and after mid-semester introduction of the tool. Using pre- and post-introduction questionnaires (n=52 paired responses), we observed reductions in self-reported search time, effort and perceived mental load, alongside an unexpected decline in perceived understanding of course policies that we interpret as preliminary evidence of a cognitive offloading trade-off. Findings are consistent with Information Foraging Theory and Cognitive Load Theory and motivate a scaled comparison group study.
Comments
SIG ED