When Usefulness Outweighs Norms: Predictors of Generative Artificial Intelligence Use in a Mobile-First Teacher Education Context
Background. Generative artificial intelligence (GenAI) tools are increasingly integrated into higher education, yet the factors driving their adoption among pre-service teachers remain underexplored in Southeast Asian contexts. Objectives. This study investigated predictors of generative AI use among 118 pre-service science teachers at a state college in Masbate, Philippines. Drawing on the Technology Acceptance Model, the Unified Theory of Acceptance and Use of Technology, and the Theory of Planned Behavior, five constructs were examined: Perceived Usefulness, Attitude Toward AI, Effort Expectancy, Facilitating Conditions, and Subjective Norms. Materials and method. A quantitative descriptive-predictive design with stratified random sampling was used. Data were collected using a validated 32-item instrument and analyzed using four multiple linear regression models. Results. Generative AI use was rated high (M = 3.67, SD = 0.72). Perceived Usefulness (M = 3.79), Attitude Toward AI (M = 3.71), and Effort Expectancy (M = 3.75) were rated high, while Facilitating Conditions (M = 3.44) and Subjective Norms (M = 3.35) were rated moderate. Across all four models (F = 50.47 to 67.14, p < .001), Perceived Usefulness was the sole predictor consistently significant at p < .001 (β = 0.443 to 0.479), explaining 63.9 to 64.6% of the variance (R² = 0.639 to 0.646). Attitude Toward AI and Effort Expectancy reached significance in selected model configurations. Multicollinearity was within acceptable limits (VIF = 1.40 to 4.76). Conclusion. These correlational findings, drawn from a single institutional sample, point to the primacy of perceived usefulness in technology adoption. As a hypothesis for future confirmation, teacher education programs may benefit from demonstrating the practical value of generative AI in science instruction.