Researchers at the HJ Institute of Technology and Management (HJITM) have published a study analyzing the factors that determine university students’ ChatGPT usage behavior from the perspectives of both benefits and concerns in the International Journal of Educational Technology in Higher Education (IJETHE, SSCI), an international journal in the field of educational technology.

The researchers surveyed 273 university students and analyzed the data using partial least squares structural equation modeling (PLS-SEM). They examined how ChatGPT’s self-learning ability and personalization features lead through knowledge acquisition and application, novelty and personal performance to benefits and intention to use, while also analyzing how barriers such as privacy concerns, technophobia and guilt, along with personal innovativeness, affect intention to use and actual behavior. All 16 hypotheses were supported.

The analysis showed that ChatGPT’s self-learning ability strongly increased knowledge acquisition (β=0.663) and knowledge application (β=0.664), and knowledge application (β=0.472) and knowledge acquisition (β=0.269) led to perceived personal performance in academic work. Personalization increased novelty (β=0.458) and benefits (β=0.216), and novelty also increased benefits (β=0.495). The factor with the strongest effect on intention to use was personal performance (β=0.432).

The barriers, however, were also clear. Privacy concerns (β=−0.191), guilt (β=−0.157) and technophobia (β=−0.101) all lowered intention to use. Yet innovativeness, the tendency to adopt new technologies early, increased both intention to use (β=0.128) and actual behavior (β=0.126), and intention to use led strongly to actual usage behavior (β=0.550). This means that students who experience academic gains end up using the tool despite their concerns.

The study is significant in explaining university students’ use of AI chatbots by combining benefits and psychological barriers in a single model. It suggests that universities and instructors should design assignments so that AI use leads to learning outcomes, while clarifying privacy standards and permitted uses to reduce students’ concerns and guilt.

HJITM continues to conduct research on the use of generative AI in education.