Developing an AI-powered Adaptive Learning System for Computer Science Education

Thesis Abstract

The rapid advancement of computer science education necessitates innovative pedagogical approaches capable of addressing learners’ diverse needs and enhancing engagement and comprehension. Despite the proliferation of traditional e-learning platforms, there remains a significant gap in personalized learning experiences tailored to individual student’s cognitive abilities, prior knowledge, and learning pace. This study aims to develop and evaluate an Artificial Intelligence (AI)-powered adaptive learning system specifically designed for computer science education, with the objective of improving learner engagement, knowledge retention, and academic performance. The research seeks to answer whether an AI-driven adaptive learning environment can effectively personalize instruction and whether this personalization leads to statistically significant improvements in learning outcomes. The study employs a mixed-methods research design, combining quantitative experimental methods with qualitative usability assessments. The target population encompasses 200 undergraduate computer science students enrolled at a comprehensive university, selected through stratified random sampling to ensure diversity in academic performance and backgrounds. A control group of 100 students experiences traditional computer science instruction, while the experimental group of 100 students interacts exclusively with the AI-powered adaptive learning system over a semester. Data collection methods include pre- and post-intervention assessments of knowledge using standardized tests, system usage logs, and structured questionnaires measuring learner motivation, self-efficacy, and system satisfaction. Additional qualitative data are gathered through semi-structured interviews with participants to explore perceived system effectiveness and usability. The quantitative data are analyzed using descriptive statistics, paired t-tests to assess learning gains, and multiple regression analysis to identify predictors of student success within the adaptive system. The qualitative data are subjected to thematic analysis to identify recurring patterns related to user experience and system usability. The study also applies the Technology Acceptance Model (TAM) and Cognitive Load Theory as theoretical frameworks to interpret user acceptance and learning processes involved in adaptive environments. Expected findings of the research include statistically significant improvements in post-test scores for students using the AI-powered adaptive system compared to the control group, alongside increased motivation and self-efficacy scores as measured through questionnaire responses. It is anticipated that the adaptive system’s ability to personalize content based on real-time performance data will contribute to better engagement and reduced cognitive overload, aligning with Cognitive Load Theory principles. Furthermore, qualitative insights are expected to reveal high levels of user satisfaction and perceived relevance of the adaptive content, although some challenges related to system complexity and technical literacy may emerge. The intended contribution to knowledge lies in the empirical validation of AI-driven adaptive learning in the context of computer science education, providing evidence of its efficacy in enhancing learning outcomes. The research also offers a conceptual model integrating AI, pedagogical strategies, and learner-centered design, which can serve as a blueprint for developing similar systems in other disciplines. Additionally, it advances understanding of the factors influencing student acceptance and effective implementation of intelligent tutoring environments. The main conclusion underscores the potential of AI-powered adaptive systems to transform computer science pedagogy by delivering more personalized, engaging, and effective learning experiences. Based on findings, recommendations include adopting AI-enabled adaptive platforms at institutional levels, designing user-friendly interfaces to accommodate diverse learners, and fostering faculty development for effective integration. Future research directions suggested involve longitudinal studies to assess sustained impacts, expansion to include diverse educational contexts, and exploration of advanced machine learning techniques to further enhance system personalization and scalability.