Integrating Generative AI Tools for Personalized Science Instruction: A Mixed-Method Study in Vocational School

Abstract

Limited empirical evidence explains how generative artificial intelligence (AI) supports personalized science instruction in Philippine vocational education, particularly within competency-based learning environments characterized by diverse student readiness, digital inequities, and constrained instructional resources. This study examined the integration of generative AI tools for personalized science instruction among vocational students through a convergent parallel mixed-methods design grounded in Constructivist Learning Theory and the Technology Acceptance Model, contributing to Sustainable Development Goal 4 (Quality Education). Quantitative data were collected from 200 Senior High School students using a validated five-point Likert-scale questionnaire (Cronbach's α = 0.91), while qualitative data were obtained from 15 purposively selected students through focus group discussions and semi-structured interviews following an eight-week AI-supported intervention. Findings indicated positive exposure to AI-assisted instruction, with content delivery receiving the highest mean score (M = 3.81). Perceived effectiveness was strongest for conceptual understanding and motivation (M = 3.78); however, exposure was not significantly associated with perceived instructional effectiveness (ρ = .055, p = .443). Qualitative findings showed that generative AI simplified complex concepts, enhanced learning efficiency, and promoted independent learning, while raising concerns regarding information reliability, learner overdependence, and unequal digital access. The study recommends strengthening pre-service and in-service teacher training on AI-enhanced pedagogy, ethical AI use, and instructional design to support effective integration of generative AI in vocational science education.