AI-powered platforms in STEM education, insights from UTAUT model: PLS-SEM and artificial neural networks hybrid analysis

Abstract

This investigation examines the effect of AI-powered technologies on STEM cognition by means of the unified theory of acceptance and use of technology (UTAUT) model. A sample of 160 respondents from diverse academic backgrounds was analysed by partial least squares structural equation modelling (PLS-SEM) to test the UTAUT model. Furthermore, artificial neural networks with a multilayer perceptron (MLP) architecture were employed to validate the outcomes and forecast the relationships amongst the UTAUT constructs. The examination focused on the core UTAUT constructs: performance expectancy, EE, SI, facilitating conditions, and their influence on the behavioural intention and actual use of AI tools in STEM education. The outcomes indicated that performance expectancy and facilitating conditions are significant predictors of behavioural intention (BI), signifying the perceived usefulness of AI tools and the existence of supportive didactic settings are critical for their adoption. Nevertheless, the conjectured moderating effects of user gender and institutional role were not strongly supported. However, the hypothesized moderating effects of user gender and institutional role on these relationships were not strongly supported, inferring a more universal applicability of AI technologies across demographic sets. These outcomes highlight the necessity of integrating AI tools into educational practices in a way that is collectively available and operative, nurturing wide-ranging educational environment that aids all users irrespective of demographic variances.