The Impact of AI-Based Learning on Academic Performance

This study compellingly demonstrates the effectiveness of AI-driven personalised learning algorithms in boosting academic performance among secondary school students in Portugal. Using a rigorous quasi-experimental, non-randomised two-shot pre-test and post-test design, we engaged sixty 10th-grade students divided into two distinct groups. The experimental group experienced AI-assisted instruction through innovative platforms, including Brisk Teaching, Khanmigo, ChatGPT 4.0 Turbo, and Quizizz AI, while the control group adhered to traditional teaching methods. Both groups participated in identical pre-tests and post-tests for two essential units: Energy in the Ecosystem and Heredity and Variation.Robust statistical analyses, including paired and independent samples t-tests, revealed significantly greater learning gains in the AI-driven group compared to the control group. Moreover, we assessed the influence of key factors, including student engagement, prior knowledge, and learning preferences, using validated Likert-scale questionnaires. The results clearly indicated a strong positive correlation between AI-driven learning and enhanced student motivation and comprehension. These findings strongly support the use of AI-based personalised instruction as an effective strategy for enhancing learning outcomes in STEM education, particularly in diverse classroom settings.