Emerging Trends and Challenges in the Use of Artificial Intelligence for Teaching Abstract Concepts in Science Education

Abstract

Artificial Intelligence (AI) is increasingly shaping science education, particularly in areas requiring visualization and conceptual modeling. This experimental study examined emerging trends and challenges associated with using AI-driven teaching systems to teach abstract science concepts in Nigerian secondary schools. A total of 180 science teachers from 12 public secondary schools were randomly assigned to an AI-supported instruction group (n = 90) or a traditional instruction group (n = 90). Over eight weeks, teachers in the experimental group used an AI-powered adaptive teaching platform to deliver lessons on abstract topics such as atomic structure, energy transformation, and cell processes. Student conceptual understanding and teacher instructional efficacy were measured pre- and post-intervention. Results revealed significant improvements in student understanding in the AI group (p < .001) and enhanced teacher instructional confidence. However, challenges related to infrastructure, digital literacy, and ethical concerns were reported. Findings highlight both the pedagogical potential and implementation barriers of AI in science educationArtificial Intelligence (AI) is increasingly shaping science education, particularly in areas requiring visualization and conceptual modeling. This experimental study examined emerging trends and challenges associated with using AI-driven teaching systems to teach abstract science concepts in Nigerian secondary schools. A total of 180 science teachers from 12 public secondary schools were randomly assigned to an AI-supported instruction group (n = 90) or a traditional instruction group (n = 90). Over eight weeks, teachers in the experimental group used an AI-powered adaptive teaching platform to deliver lessons on abstract topics such as atomic structure, energy transformation, and cell processes. Student conceptual understanding and teacher instructional efficacy were measured pre- and post-intervention. Results revealed significant improvements in student understanding in the AI group (p < .001) and enhanced teacher instructional confidence. However, challenges related to infrastructure, digital literacy, and ethical concerns were reported. Findings highlight both the pedagogical potential and implementation barriers of AI in science education