Users’ Satisfaction in Using ChatGPT among University Students in UOW Glenmarie, Malaysia
TL;DR
The study underscores the importance for educators and software developers to focus on factual accuracy and transparency to maximize the positive impact of AI in higher education.
Abstract
Generative AI applications like ChatGPT are transforming how university students conduct research, write, and study. However, because AI models occasionally generate unverified or inaccurate information, student adoption depends heavily on trust. This study evaluates how usefulness, ease of use, and perceived information reliability affect overall student satisfaction with ChatGPT. Surveying 385 undergraduate students at UOW Malaysia, the study applied an expanded Technology Acceptance Model (TAM). The statistical model accounted for 69.1% of the variance in user satisfaction. Results show that Perceived Usefulness is the strongest indicator of satisfaction (β=0.392), followed closely by Information Reliability (β=0.348), and Perceived Ease of Use (β=0.174). While ease of navigation is appreciated, students prioritize functional utility and data accuracy when deciding whether an AI tool is genuinely valuable for their education. The study underscores the importance for educators and software developers to focus on factual accuracy and transparency to maximize the positive impact of AI in higher education.