Using Diffusion Theory of Innovation to Investigate Perceptions of STEM and Non-STEM Students' Adoption of Chatbot Systems in Higher Education: A Multiple Group Analysis

Chatbot systems have emerged as potent tools to bridge the divide between the expectations of 21st-century students and the conventional educational framework. Utilizing the Diffusion Theory of Innovation, this research explores the perceptions and adoption of chatbot systems in higher education, considering both STEM and non-STEM students. The study delves into factors such as relative advantages (RA), compatibility (CP), trialability (TR), observability (OB), and complexity (CO) in shaping the intention to adopt chatbots (IN). Drawing insights from 842 higher education students (445 STEM, 397 non-STEM), Partial least square structural equation modeling analysis through SmartPLS software yielded significant findings. Specifically, STEM students exhibited a notable positive relationship between complexity and the intention to adopt chatbots, a pattern not mirrored among non-STEM students. Conversely, observability significantly influenced the intention to adopt chatbots for non-STEM students but did not hold the same significance for STEM disciplines. This study underscores the necessity of considering discipline-specific factors when implementing chatbot systems in higher education. Recognizing nuanced differences in the impact of complexity and compatibility among STEM and non-STEM students is crucial for institutions aiming to enhance the acceptance and effectiveness of these systems, thereby enriching the overall educational experience. The study discussed the findings, highlighted the limitations, and suggested future direction.