Exploring Factors Influencing Concerns for the Use of Generative Pretraining Transformer (GPT) Variants for the Online Guided Inquiry (OGI) among Higher Education Students

Abstract This study investigates the perceptions of higher education students regarding the use of GPT variants for online guided inquiry (OGI). Historically, technology has supported teaching and learning without significant adverse effects; however, the growing concern about its potential to replace human involvement, particularly in guided inquiry created impetus for this study. A cross-sectional study using mixed-methods approach was conducted with 150 students from Science, Mathematics, and Engineering disciplines at both undergraduate and graduate levels. A nine-construct model was developed and validated using both newly developed and adapted instruments. The results indicated an acceptable model fit (SRMR = 0.12, NFI = 0.89) and explained variance of R2 = 0.461R2. Despite concerns about inaccuracy, misuse, and over-dependence on generative AI, students generally held positive views about its use for OGI. Factors contributing to this positive perception included familiarity with technology (β = 0.346), help-seeking behavior (β = 0.21), perceived enjoyment (β = 0.281), and multitasking abilities (β = 0.25). Conversely, social influences, such as subjective norms, did not significantly impact perceptions. It can be concluded that academic help-seeking, enjoyment derived from its use, and the nature of being multitasking are key factors that encourage students to adopt GPT variants, in contrast to the prevailing subjective norms regarding their use. Consequently, higher education institutions should implement strategies to monitor and regulate the use of generative AI to enhance educational experiences and outcomes.