A Hybrid SEM-ANN Model for Post-Pandemic Adoption of Online Platform Using Information and Technology Acceptance Constructs

Abstract

Using the TAM and the information attributes of Information Richness and Information Quality, we study Gulf region universities that make use of online resources during the pandemic. We then employ a dual-stage analytical design to determine their future behavioral intentions. Core behavioral aspects were systemically tapped by use of a cross-sectional survey instrument design. Out of the 638 students surveyed, PU, PEOU, IR, IQ, and BI were sampled (data collected October 25 25- December 25, 2021). Following the basic IR→BI and IQ→BI, the structural model for PLS-SEM (SmartPLS) was adjusted and possesses marked later validation and nonlinear profound artificial neural network rank effects (2 hidden layers, 10-fold cross-validation). The PLS-SEM Model Explained 63.2 on the variance of BI and all the paths IR→BI PU→BI PEOU→BI IQ were hypothesized (IR→BI β = 0.721; PU→BI β = 0.651; PEOU→BI β = 0.495; IQ→BI β = 0.488; p <0.05), then β value predictions, relevant and plausible from the ANN model achieved low, around 129 and 133 prediction error train and test and powered up R^2 prediction value of 0.83. The sensitivity of the predictors, ranked as IR > PEOU > PU > IQ, reveals that the outcomes of the three tiers of Situation Awareness (perception, understanding, projection) indicate that content characterized by low structure, high usability, and substantial, yet poor quality, with enduring post-pandemic usage intentions, was paramount. The contribution of the study includes (i) the first validated hybrid SEM-ANN protocol for education technology acceptance, (ii) proof of the primacy of IR in sustaining BI, and (iii) a pragmatic policy for universities: invest in richer content objects (interactive/multimedia), effortless interfaces, and sustained information governance to maintain engagement even outside of crisis contexts.