A+, a Framework for Adaptive Computer Education Pedagogy

Thesis Abstract

This study addresses a persistent gap in computer education where heterogeneous learner profiles and dynamic classroom conditions undermine uniform attainment of computational competencies. Despite rapid advancements in educational technologies, existing pedagogical frameworks often fail to adapt instruction to individual differences in prior knowledge, motivation, and cognitive load within programming and software development curricula. The aim is to develop and validate A+, a framework for adaptive computer education pedagogy that systematically aligns instructional strategies,Assessment, and learning analytics with learner profiles to optimize achievement, engagement, and transfer of computational skills. Specific objectives are (1) to synthesize existing adaptive learning theories and computer education practices into a cohesive framework; (2) to operationalize the A+ framework into a set of adaptive pedagogical modules and an accompanying instructional design model; (3) to implement the framework in university-level introductory and intermediate programming courses across three departments; (4) to evaluate the framework’s impact on learning outcomes, motivation, and cognitive load using mixed methods; and (5) to refine the framework based on empirical evidence and stakeholder feedback for broader applicability. A sequential explanatory mixed-methods design is employed. The population comprises 1,200 undergraduate and graduate computer science and information technology students enrolled in programming-related courses at three public universities over an two-semester period. A stratified random sample of 600 students is selected for quantitative analysis, with 200 students from each university, ensuring representation across gender, prior programming experience, and academic level. Qualitative data are gathered from purposively selected participants (n=60) representing high, medium, and low performers, as well as instructors and teaching assistants (n=12). Data collection instruments include standardized baseline and end-of-course assessments aligned with Bloom’s taxonomy (knowledge, application, analysis, and creation), instrumented learning analytics from the learning management system (LMS) capturing time-on-task, resource usage, and assessment attempts, validated motivation scales (intrinsic and extrinsic motivation), cognitive load scales (NASA-TLX adapted for programming tasks), and semi-structured interview protocols for students and instructors. Instrument validity and reliability are established through content validation by a panel of six computer education experts and pilot testing (Cronbach’s alpha > .80 for all scales). Quantitative data are analyzed using hierarchical linear modeling to assess the impact of adaptive interventions on learning gains, controlling for prior knowledge and demographic covariates. Regression analyses identify predictors of success, while repeated-measures ANOVA examines changes in motivation and cognitive load over time. Learning analytics data inform the operationalization of adaptive components such as personalized hint delivery, problem-solving scaffolds, and task sequencing. The qualitative data undergo thematic analysis following Braun and Clarke’s methodology, with expert coding to triangulate quantitative findings and illuminate mechanisms of action within the A+ framework. A convergent mixed-methods integration will compare and synthesize results to provide a comprehensive evaluation of the framework’s effectiveness and feasibility. Expected findings include (i) statistically significant improvements in post-test scores for adaptive groups compared to non-adaptive controls, with effect sizes in the small-to-moderate range (Cohen’s d ? 0.30–0.50); (ii) higher intrinsic motivation and reduced perceived cognitive load among learners exposed to adaptive pedagogy; (iii) positive correlations between LMS-derived engagement metrics and achievement; and (iv) qualitative evidence of enhanced learner autonomy, satisfaction with feedback, and perceived relevance of learning activities. The study anticipates identifying critical moderating factors such as prior programming experience, perceived usefulness of adaptive prompts, and instructor fidelity to the framework. The contribution to knowledge includes (a) a rigorously developed and empirically validated adaptive pedagogy framework for computer education, (b) an operational model and guideline set for implementing adaptive modules and instructional design in programming curricula, (c) empirical evidence on the relationship between adaptive instructional strategies, learner motivation, cognitive load, and learning outcomes, and (d) a transferable methodology for future cross-institutional validation of adaptive educational technologies in STEM disciplines. The final conclusion is that A+ offers a viable, scalable approach to personalize computer education, balancing learner needs with curricular goals. Recommendations include refining adaptive rule sets with domain-specific programming tasks, expanding to collaborative programming contexts, integrating real-time affective computing to further tailor instruction, and conducting longitudinal studies to assess long-term retention and transfer of programming competencies.