SUPPORTING CASE ANALYSIS THROUGH GUIDED LLM PROMPTING IN Online CASE-BASED MARKETING COURSES: AN EXPLORATOYR CLASSROOM-BASED COMPARISON

The Case Method trains students to define managerial problems, compare alternatives, and defend recommendations under uncertainty. Preserving these processes online can be difficult, and large language models (LLMs) further complicate this challenge by enabling students to generate “polished” recommendations without constructively engaging in the Case Method. Using the ICAP framework as a design lens, we developed a guided LLM prompting marketing case assignment requiring students to evaluate multiple LLM-generated alternatives before synthesizing their own recommendation. We then conducted a classroom-based comparison across three online sections: asynchronous with guided LLM intervention, asynchronous without guided LLM intervention, and hybrid without guided LLM intervention but with synchronous discussion and debrief. Students in the guided LLM prompting format self-reported higher perceived self-directed learning than students in the two comparison formats, although uncertainty around the effect-size estimate warrants cautious interpretation. No significant omnibus differences emerged for perceived collaborative learning, problem solving, or critical thinking, indicating that guided LLM prompting should not be assumed to improve all dimensions of case-based learning. Findings suggest that this assignment may be most useful as a targeted pedagogical structure for supporting independent case analysis and offer educators a practical framework for aligning LLM use with specific Case Method learning processes.