PLS-SEM Application in Determining AI-Powered Writing Tools Usage among University Students

Abstract

The growing use of Artificial Intelligence (AI) applications among students has dramatically changed several elements of education, including academic writing. This study analyses the factors that influence students' intentions to use AI-powered writing tools, such as ChatGPT and Perplexity, in their academic writing. The Value-Based Adoption Model (VAM) was applied as a theoretical framework to explore the perceptions of 281 university students regarding the perceived usefulness, enjoyment, technicality, and cost. Data collected through questionnaires were analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM), and the findings concluded that perceived usefulness and enjoyment significantly influence students' intentions to use AI applications, while perceived technicality and cost have no significant influence on the use of AI applications in students’ academic writing. In addition, the study confirms that PLS-SEM model provides better predictive accuracy than a simple linear regression approach. These findings suggest that universities should establish guidelines for appropriate AI use while capitalizing on its benefits for academic progress.