Comparative Analysis of E-Learning Effectiveness in Computer Science Education Across Universities

Thesis Abstract

The rapid proliferation of digital technologies has transformed computer science education, prompting a critical need to evaluate the effectiveness of e-learning modalities across diverse university settings. This study addresses the gap in comprehensive comparative analyses of online learning outcomes in computer science by systematically evaluating the pedagogical impact, student engagement, and learning achievements in different institutional contexts. The primary aim is to identify the relative strengths and weaknesses of e-learning approaches employed across selected universities, thereby informing best practices for optimizing digital instruction in computer science programs. The specific objectives are to (1) measure student academic performance in e-learning environments; (2) assess students’ perceptions of e-learning effectiveness; (3) analyze pedagogical strategies and technological tools used; and (4) determine institutional factors influencing e-learning success. Drawing upon a mixed-methods research design, the study combines quantitative approaches—using a quasi-experimental comparison of student performance data, surveys, and questionnaires analyzed via Analysis of Variance (ANOVA) and multiple regression analysis—with qualitative analyses, including thematic analysis of students’ and instructors’ interview transcripts. The population comprised undergraduate and postgraduate computer science students (n=1,200) and instructors (n=50) across five universities with established e-learning platforms. A stratified random sampling technique was employed to derive a representative sample of 300 students per institution, ensuring diversity across academic levels and demographics. Data collection instruments included validated questionnaires measuring learning effectiveness, engagement, and satisfaction, alongside institutional pedagogical documentation and platform analytics. Validity and reliability of the instruments were established through pilot testing and Cronbach’s alpha coefficients exceeding 0.85. Data analysis began with descriptive statistics to summarize demographic variables and baseline performance metrics, followed by inferential statistical tests—ANOVA to compare performance means across institutions, and multiple regression models to identify predictors of e-learning success. The qualitative data obtained from interviews were subjected to thematic analysis, guided by constructivist grounded theory, to explore contextual factors influencing e-learning effectiveness. Expected findings indicate significant variations in student performance and perceptions of e-learning effectiveness across institutions, with technological resource availability, instructor training, and learner preparedness acting as critical moderating factors. It is hypothesized that universities with more advanced digital infrastructure and comprehensive faculty development programs will demonstrate higher student engagement and learning outcomes. The analysis is anticipated to reveal that pedagogical strategies integrating active learning, interactive assessments, and personalized feedback correlate positively with student success. The study contributes to existing knowledge by providing a comparative framework specific to the domain of computer science education, highlighting institutional and pedagogical determinants of e-learning effectiveness. It advances understanding of how technology, curriculum design, and institutional support influence digital learning success, offering empirical evidence for policymakers, educators, and technology developers. The research findings are expected to generate practical recommendations for optimizing e-learning implementation, including scalable best practices, faculty development initiatives, and technological investments. Conclusively, the study advocates for a tailored, evidence-based approach to digital education in computer science, emphasizing the importance of context-sensitive strategies that align technological capacities with pedagogical needs. Recommendations include the development of standardized e-learning quality metrics, targeted faculty training programs, and continuous performance monitoring to ensure sustainable improvement. It is suggested that future research extend this comparative analysis to incorporate longitudinal assessments and cross-disciplinary applications to further generalize the findings and refine e-learning models in higher education. The comprehensive contribution of this research lies in providing a robust, multidimensional understanding of e-learning effectiveness across diverse university contexts, thereby guiding stakeholders towards more informed, strategic decision-making in the evolving landscape of digital computer science education.