DETERMINING FACTORS IN THE UTILIZATION OF ARTIFICIAL INTELLIGENCE: PERCEPTIONS AND BEHAVIORS OF PROSPECTIVE PRIMARY SCHOOL TEACHERS IN COMPLETING SCIENCE ASSIGNMENTS

Artificial Intelligence (AI) holds significant potential to transform education, particularly in teaching methodologies and task completion. This study aims to identify the factors influencing the perceptions and behaviors of elementary education students in utilizing ChatGPT and Gemini to complete science-related assignments. The research design employs a quantitative approach with both descriptive and causal methodologies. Data testing and analysis are conducted using Structural Equation Modeling (SEM), P-value, and Prediction-Oriented Segmentation (POS). Path analysis results reveal that perceived benefits significantly impact perception (.403) and behavior (.406). AI effectiveness significantly affects perception (.303) but minimally influences behavior (.018). Preference for AI usage positively influences behavior (.305), whereas dependence on AI negatively impacts perception (-.050). Restrictions on AI usage reduce perception (-.077) but increase behavior (.115). The p-value analysis indicates that the perceived benefits of AI use significantly influence behavior (.000) and perception (.000), supporting the hypothesis that perceived benefits play a crucial role in enhancing AI adoption and fostering positive attitudes toward its use. Conversely, AI effectiveness significantly affects perception (.000) but not behavior (.862). Dependence, restrictions, and the impact of AI show no significant effects on either behavior or perception, except for AI usage preferences, which significantly influence behavior (.033). Segment analysis reveals that perceived benefits influence behavior in Segment 1 (.510) and perception in Segment 2 (.493). AI effectiveness negatively impacts behavior in Segment 2 (-.633) but shows moderate effects in Segment 1 (.214). Preferences for AI usage exert a more substantial influence on behavior in Segment 2 (.614), while the effects of dependence and restrictions vary across segments. The perceived benefits of AI encourage technology adoption among students, while dependence and restrictions introduce complexities in formulating AI based educational policies.