Asian Journal of Advanced Research and Reports

The research compares AI-driven personalized learning algorithms and traditional learning in Science, Technology, Engineering and Mathematics curriculum design optimization, with emphasis on secondary-level biology class in West Wendover High School, Nevada, United States. Using a two-shot quasi-experimental design, the research investigates the academic performance of Grade 10 students, a sample size of (n=60) who are divided into an experimental group (AI-driven learning) and a control group (traditional learning). Pre and post-tests assessed student performance in two biology units studied—Energy in the Ecosystem and Heredity and Variation— while surveys tracked the influence of student engagement, prior knowledge, and learning preferences on AI-driven learning effectiveness. Paired and independent sample t-tests were used to compare the two groups. Results show that the AI group participants' post-tests were considerably higher, with high levels of mastery (83.0% for Unit 3 and 90.9% for Unit 4), in contrast with the moderate mastery of the traditional group (69.2% for Unit 3 and 70.9% for Unit 4). Paired and independent samples t-tests statistical analysis proved to the superiority of AI-driven learning compared to traditional learning in terms of improving academic performance, particularly in complex subjects like genetics, heredity, and variation. Besides, AI-driven learning improved learners' engagement by facilitating real-time feedback, adaptive learning paths, and interactive digital tools, while the study highlights the importance of consolidating foundational knowledge alongside AI integration. The paper concludes that AI-driven personalized learning considerably improves STEM learning because it meets personal learning requirements, stimulates interest, and makes complex ideas simpler. For optimum AI-driven teaching, though, STEM curricula should integrate personalized AI-driven with interactive, peer-to-peer learning exercises to promote greater conceptual comprehension.

Based on these findings, this study recommends the use of AI-driven tools in STEM courses, the development of advanced adaptive learning features, and the implementation of hybrid approaches to AI-driven and traditional learning. It also recommends district leaders, curriculum designers, and instructors to explore AI-driven personalized learning to enhance curriculum optimization and continue researching to design and test these methods for large-scale implementation.