Exploring University Students' Acceptance of Generative AI for Writing: An Extended Technology Acceptance Model (TAM 3) Perspective

Abstract

Generative AI presents significant potential for improving students' writing, and its acceptance is influenced by varied factors. The present study aims at exploring the factors by examining six vital constructs of Technology Acceptance Model (TAM) 3 (Venkatesh & Bala, 2008)—academic relevance, output quality, self-efficacy, playfulness, anxiety, and perceived enjoyment—and their impact on perceived usefulness, perceived ease of use, and behavioral intention of students. Using a quantitative research approach, a sample size of 345 university students from Computer Science, Management Science, and Arts and Humanities voluntarily participated in an online cross-sectional survey questionnaire. The data was analyzed by SPSS 23.0 and SmartPLS 4.0 for exploratory and confirmatory factor analysis. Academic relevance (β = .378***) and perceived ease of use (β = .456***) were identified as significant predictors of students' perceived usefulness of (Gen) AI in their writing as compared to the output quality of (Gen) AI, which was found to be an insignificant factor of perceived usefulness. The study also reveals that playfulness (β = 0.332**) and perceived enjoyment (β = 0.449***) were the primary factors influencing students' perceived ease of use, unlike anxiety and self-efficacy, which showed statistically insignificant effects. The findings of the study further suggest that perceived usefulness (β = 0.606***) and ease of use (β = 0.227*) are key enablers of the behavior of students' intention to adopt (Gen) AI in the future.