Intelligent Tutoring System for Enhancing Programming Skills among Computer Science Students

Thesis Abstract

The persistent challenge of effectively developing programming skills among computer science students necessitates innovative instructional approaches that transcend traditional classroom methods. This study investigates the design, implementation, and evaluation of an intelligent tutoring system (ITS) tailored to enhance programming competency in undergraduate computer science cohorts. The primary aim is to develop a scalable and adaptive digital learning environment that provides personalized feedback, real-time problem solving guidance, and interactive coding exercises to foster deeper understanding and skill acquisition. Specific objectives include examining the efficacy of the ITS in improving programming performance, identifying student engagement levels, exploring user perceptions of the system’s usability and pedagogical effectiveness, and determining the pedagogical impact of adaptive feedback mechanisms. The methodology adopts a mixed-methods research design combining quantitative and qualitative data collection techniques. The study population comprises 200 second-year undergraduate computer science students enrolled at a comprehensive university, with a stratified random sampling technique employed to select 120 participants for the experimental group and 80 for the control group. The experimental group interacts with the developed ITS over a 12-week semester, while the control group follows traditional instruction without digital augmentation. Data collection instruments include validated coding skill assessments, user satisfaction questionnaires based on the Technology Acceptance Model (TAM), and semi-structured interviews. The coding assessments are administered pre- and post-intervention to measure skill improvement, whereas questionnaires and interviews capture perceptions and attitudes toward the ITS. Data analysis involves descriptive statistics to summarize demographic and baseline data, paired t-tests and ANOVA to evaluate the significance of skills improvement within and between groups, and regression analysis to examine predictors of performance enhancement. Thematic analysis is employed for qualitative data to identify key themes related to system usability, engagement, and pedagogical impact. Additionally, the study adopts Vygotsky’s Zone of Proximal Development (ZPD) theory and constructivist learning principles to underpin the system’s adaptive mechanisms, ensuring scaffolding aligns with individual learner needs. Expected findings indicate that students utilizing the ITS will demonstrate statistically significant gains in programming proficiency compared to the control group, with increased engagement levels and positive perceptions of system usability. It is anticipated that personalized feedback and adaptive exercises will positively influence learners’ motivation and reduce frustration related to complex programming tasks. The research is also expected to reveal critical system features that contribute most substantially to skill development and learner satisfaction, providing essential insights for future ITS enhancements. This study contributes to the existing body of knowledge by empirically validating the effectiveness of AI-driven tutoring environments in computer science education, specifically in programming skill acquisition. It advances understanding of how adaptive instructional strategies impact learner outcomes in technical disciplines and provides a theoretical framework for integrating intelligent systems into conventional pedagogy. The findings will inform educators, developers, and policymakers about the potential of incorporating intelligent tutoring solutions at scale to address persistent pedagogical challenges. The main conclusion emphasizes that well-designed ITS can serve as a supplementary pedagogical tool that significantly boosts programming skills, fosters self-directed learning, and enhances overall educational quality. Recommendations include integrating the ITS into mainstream curricula, investing in ongoing system refinement based on user feedback, and conducting longitudinal studies to assess long-term impacts. The research advocates for broader adoption of intelligent tutoring systems in computer science education and suggests avenues for future research, such as exploring the integration of machine learning algorithms for even more personalized learning experiences and scalability in diverse educational settings.