Leveraging Artificial Intelligence (AI)-Enhanced STEM Cognition-Multi-Directionality of Influence

This research explores the integration of artificial intelligence (AI) in Science, Technology, Engineering, and Mathematics (STEM) cognition, focusing on predictive modeling and performance analysis. The study formulated hypothesis to address gaps identified in the existing literature on AI in educational contexts. Using Bootstrap MGA, Parametric Test, Welch Satterthwaite analyses, and Importance-Performance Map Analysis (IPMA), the research assessed gender differences and predictive importance of latent variables in the structural model. With a sample size of 71 students, the study employed rigorous testing for convergent and discriminant validities in the questionnaire design. Results indicated generally non-significant gender disparities, except for a potential gender-specific nuance in the MMS→ACP pathway. The IPMA highlighted Analogical Comparison Principle (ACP) as a robust predictor (total effect = 1), while Mathematical Cognition (MAS) showed low importance (total effect = 0.05). Mathematical and Computational Algorithms (MCA) emerged as a substantial predictor across gender groups (total effects ranging from 0.5 to 0.59), and Mathematical Modelling and Simulation (MMS) exhibited varying effects.