Artificial Intelligence in University Students: Analyzing ChatGPT’s Academic Engagement and Usage Frequency

Abstract

Higher university students are doubtful about how to use ChatGPT for academic purposes effectively due to the lack of clear guidance and other issues such as academic dishonesty. This will discourage students from fully engaging in it. In current time, the skill becomes increasingly important in various careers, especially as AI tools become more integrated into day-to-day work. This research takes a survey on quantitative study approach to identify students’ academic engagement on ChatGPT for academic purposes. Samples of 429 have been collected among university students from a reputed private university in Malaysia. Descriptive analysis of SPSS version 29.0 is utilized to identify the mean, percentage and standard deviation. Meanwhile the Independent T Test and Spearman Rho’s correlation analysis were used to test the relationship between variables. Study revealed 30% students used ChatGPT for academic purposes weekly, 17% used monthly, 16% daily, 29% rarely, and interestingly 7% never used it. Mann-Whitney U Test analysis showed that there are no differences in academic engagement between males (M=3.59, SD=1.01) and females (M=3.38, SD=.99), U= 16126.5, Z = -1.70, p = >.05. There is no significant difference in ChatGPT academic engagement between conventional students (M=3.47, SD=.89) and Modular students (M=3.42, SD=1.06), U=21683.5, Z= -.265, p = >.05. A Kruskal-Wallis H test analysis showed no difference of ChatGPT usage across level of study (PhD, Master, Bachelor and Diploma), H (4) =7.120, p=>.05 indicated level of studies or depth studies do not influence the frequency of ChatGPT usage. Overall students’ academic engagement is in medium (M=3.44, SD=1.01M, N=429). T Stakeholders, policy makers, academicians, institutions and instructors must create a well-informed guideline of using this tool for students to connect effectively with the tool in learning and performing academic tasks.