AI-Augmented Advising: A Comparative Study of GPT-4 and Advisor-Based Major Recommendations

Abstract

Choosing an undergraduate major is an important decision that impacts academic and career outcomes. In this work, we investigate augmenting personalized human advising for major selection using a large language model (LLM), GPT-4. Through a three-phase survey, we compare GPT suggestions and responses for undeclared first-and second-year students (n = 33) to expert responses from university advisors (n = 25). Undeclared students were first surveyed on their interests and goals. These responses were then given to both campus advisors and GPT to produce a major recommendation for each student. In the case of GPT, information about the majors offered on campus was added to the prompt. Overall, advisors rated the recommendations of GPT to be highly helpful (4.0 out of 5 on its explanation for the recommendation and 3.8 on its answers to individual student questions) and agreed with its recommendations 33% of the time. Additionally, we observe more agreement with AI’s major recommendations when advisors see the AI recommendations before making their own. However, this result was not statistically significant. We categorize qualitative feedback from advisors with an affinity diagram and outline five design implications for future AI-assisted academic advising systems. The results provide a first signal as to the viability of LLMs for personalized major recommendation and shed light on the promise and limitations of AI for advising support.