Comparative Analysis of AI Tools in Computer Education Outcomes Across Regions

Thesis Abstract

The rapid integration of artificial intelligence (AI) tools into computer education has amplified regional disparities in learner outcomes, access to resources, and instructional quality, necessitating a systematic comparison of how AI-enabled practices influence student achievement and engagement across diverse educational contexts. This study addresses the problem of uneven effectiveness of AI tools in computer education by examining variations in learning outcomes, motivation, and self-regulated learning among students from contrasting regional settings, and by probing how teacher practices and institutional support mediate these effects. The aim is to evaluate whether AI-assisted pedagogies produce comparable gains in computer education outcomes across regions and to identify contextual factors that enhance or hinder their effectiveness. The specific objectives are (1) to compare student achievement in computer science concepts between AI-augmented and traditional instruction across three distinct regions; (2) to examine differences in student engagement, perceived usefulness, and self-efficacy related to AI tools; (3) to analyze teacher instructional practices, technological readiness, and institutional support as mediators of AI effectiveness; (4) to identify contextual moderators such as internet access, device availability, and curriculum alignment that influence outcomes; and (5) to develop a framework of best practices for regionally adaptive AI-enhanced computer education. The study employs a mixed-methods cross-sectional design, integrating quantitative and qualitative data to provide a comprehensive understanding of AI tool impact. The population comprises secondary and tertiary level learners enrolled in computer education courses and teachers delivering AI-enhanced instruction across three regions with distinct socioeconomic and educational profiles. A stratified random sample of 1,200 students (400 per region) and 120 computer education teachers (40 per region) will be drawn. Quantitative data will be collected via standardized assessments of computer science knowledge, validated engagement scales, motivation inventories, self-efficacy measures, and usage analytics from AI tools. Qualitative data will be gathered through semi-structured interviews and focus group discussions with a subset of 60 students and 24 teachers, purposively selected to represent diverse experiences with AI-integration. Instrument validity and reliability will be established through expert review, pilot testing, and Cronbach’s alpha thresholds above 0.70 for internal consistency. Data collection will occur over a 12-week instructional cycle. Data analysis will proceed in two strands. Quantitative analyses will include descriptive statistics, multivariate analysis of covariance (MANCOVA) to compare learning outcomes and engagement across regions while controlling for prior achievement, and hierarchical linear modeling (HLM) to account for nested data structures (students within classes). Mediation analysis will test whether teacher practices and institutional support mediate the relationship between AI use and outcomes. Subgroup analyses will explore regional moderators such as device availability and curriculum alignment. Qualitative data will be analyzed using thematic analysis, guided by the Technology Acceptance Model (TAM) and the Expectancy-Value Theory, to elucidate user experiences, perceived usefulness, ease of use, and motivational factors. Triangulation will integrate quantitative and qualitative findings to enhance explanatory power and validity. The theoretical framing will anchor the study in constructivist learning theory, situated cognition, and the diffusion of innovations theory to interpret regional adoption patterns and knowledge construction processes. Expected findings anticipate heterogeneity in AI-enabled educational gains, with regions characterized by robust digital infrastructure, aligned curricula, and teacher proficiency exhibiting larger improvements in achievement, engagement, and self-efficacy. The study is expected to reveal that teacher professional development and institutional readiness significantly mediate AI effectiveness, while access disparities moderate the strength of observed outcomes. The contribution to knowledge includes a comparative, evidence-based understanding of how regional context shapes AI-assisted computer education, a validated analytical framework for assessing cross-regional AI impact, and practical recommendations for policymakers and schools to implement regionally adaptive AI strategies. The main conclusion will emphasize the necessity of synchronous investments in infrastructure, curriculum alignment, and teacher capacity-building to achieve equitable AI-enhanced learning gains. Recommendations will address scalable professional development models, governance of AI tool selection, monitoring and evaluation protocols, and guidelines for ensuring equitable access to AI-enabled computer education across regions.