From Visualization to Adaptive Learning in Algorithms and Data Structures

Computer science students are required to master data structures and algorithms, yet learning these topics often presents significant cognitive challenges. This study investigates whether a bundled package of adaptive personalization is associated with learning outcomes, perceived workload, and usability, using an existing educational visualization tool called VZOU (Visualization Zone of Understanding). The enhanced adaptive version of VZOU incorporates personalized learning paths structured around a visible learning progression map, enabling learners to monitor their progress, identify gaps, and take ownership of their learning path. Basic gamification features, Bloom-aligned progressive sequencing, and topic-specific formative quizzes are designed to support self-regulated learning through immediate feedback and motivational reinforcement. In a between-subjects study (N = 55), we compared control, non-adaptive, and adaptive groups. The adaptive group scored significantly higher on the post-test than the non-adaptive and control groups (mean 89.3 vs. 73.3 and 71.0 out of 100, respectively), and showed numerically but not statistically higher usability scores (SUS). Perceived mental workload showed a mixed pattern: the overall six-subscale NASA-TLX composite did not differ significantly by condition, though Mental Demand specifically was significantly higher in the adaptive group than in the control group, and a workload composite excluding self-rated success also differed significantly across conditions. The bundled adaptive VZOU package, combining a diagnostic-driven graph, gated progression, formative quizzes, and gamification, was associated with higher post-test performance, while raising certain aspects of perceived workload. As these components were not manipulated independently, the estimate reflects the package as a whole, not any single element.