Code as Anchor, Memory and Metaphor as Support: Learner Experiences with Multi-View Visualizations

Synthesis: Students spent ~47% of gaze time on code despite visual scaffolds. Three factors shape selective engagement with multi-representational tools: Agency (students want control over cognitive effort), Representational Fit (the same design feels helpful to some and overwhelming to others), and Legitimacy (metaphorical scaffolds are perceived as childish at university level).

Key Findings

• Problem: Students often ignore well-designed program visualizations; existing cognitive design principles don't explain learner engagement/disengagement.
• Method: Within-subjects study (N=19 undergraduates, post-CS1/CS2) using think-aloud, interviews, and webcam gaze tracking with a multi-representational probe.
• Gaze finding: ~47% of time on code despite visual scaffolds; students without prior experience anchored more in code and ignored metaphor views.
• Three engagement themes: Agency (control over cognitive effort), Representational Fit (wide individual variation), Legitimacy (metaphors seen as childish at university level).
• Implication: multi-representational tools need attention to affective and social factors, not just cognitive design.

What this means for practice

• Learners. Treat a visualization as a check on reasoning you have already committed to, not a walkthrough to follow: participants described productive use as predicting an expected state and then verifying it, and resisted automation that felt passive.
• Instructors. Position program visualizations as on-demand verification mechanisms rather than step-by-step guides, so that asking for help preserves rather than reduces the control students want over their own cognitive effort.
• Instructors. Offer toggleable abstraction levels so students can calibrate visual density to their current capacity — in this study identical designs were essential scaffolds for some students and overwhelming for others.
• Instructors. Model visualization use in your own reasoning and frame metaphor views as conceptual scaffolds rather than remedial notation, since students avoided the metaphor view partly because it read as childish at university level.
• Designers. Do not assume representational complementarity alone will produce engagement: students spent 47% of gaze time (median) on Code, 34% on Memory, and 9% on Metaphor.

Limitations

• Within-subjects study with 19 undergraduates (six women, thirteen men) who had recently completed CS1 and CS2 in Python, recruited from two campuses of one university, so the themes are analytically grounded characterizations of one institutional context rather than universal claims.
• The visualization tool was a research probe and the specific metaphors chosen inevitably shaped student reactions, so different visual designs might elicit different responses.
• Sessions ran about 60 minutes total across three topics, so assessment pressure, peer dynamics, and sustained use over a full term went unobserved.
• Task accuracy and completion rates were not analyzed as outcomes, so the engagement patterns cannot be linked to performance differences; gaze was tracked with webcam-based WebGazer and treated descriptively, which cannot support fine-grained attention claims.

Citation

Sibia, N., Wen, J., Richardson, A., Jain, Y., Malik, K., Simion, B., Nobre, C., Zavaleta Bernuy, A., Petersen, A., & Liut, M. (2026). Code as Anchor, Memory and Metaphor as Support: Learner Experiences with Multi-View Visualizations. ICER 2026.