The role of epistemic beliefs in predicting ChatGPT adoption and avoidance in higher education

Abstract

AI-based tools such as ChatGPT are currently reshaping the landscape of higher education and significantly impacting student learning and academic achievement. With such powerful tools, individuals might get the impression of being guided by an omniscient epistemic authority that can be trusted absolutely and has the right answer to every question. However, little empirical research has yet addressed how these epistemic beliefs are associated with the adoption or avoidance of ChatGPT in higher education. Based on the theoretical frameworks of Hofer and Pintrich (1997) as well as Schommer (1990), we conducted a questionnaire study to assess epistemic beliefs about generative artificial intelligence and its adoption or avoidance. The sample consisted of 185 students (74.6% female) with an average age of 22.50 years from a German university. The questionnaire assessed the source and certainty of knowledge, the justification of knowledge claims, and the speed of knowledge acquisition. Results show that students' epistemic beliefs strongly predict their adoption of ChatGPT and moderately predict their avoidance of it. Epistemic beliefs thus emerge as an important factor in explaining learners' engagement with generative AI in higher education.

Ähnliche Arbeiten