In interactive visual explanations that already use prediction-before-manipulation and manipulation budgets, does explicitly revealing a learner’s own interaction trace (e.g., a simplified history of tested values and outcomes) during or just before delayed, out-of-context retention checks improve durable conceptual understanding and far transfer relative to (a) no trace display and (b) generic textual recap, and for which learners does trace-visibility shift productive struggle into counterproductive over-analysis of past moves?

Answer

Trace-visibility during or just before delayed, out-of-context retention checks probably yields small gains in durable learning and far transfer over no-trace and over generic textual recap, but mainly when (a) the trace is highly compressed and aligned to target relations and (b) learners are at least modestly regulated. For fragile novices and ruminative, high-anxiety learners, visible traces can tip productive struggle into counterproductive over-analysis and self-blame unless they are tightly structured and framed as neutral evidence.

Relative effects

• vs no trace: modest benefit. A short, visual history of a few key moves (values tried, outcomes, brief labels) helps learners reconstruct their earlier hypotheses and errors, supporting spaced retrieval and error-based reflection.
• vs generic recap: trace wins when it is concrete and minimal (e.g., 3–6 annotated “pivotal moves”), because it anchors reflection in actual behavior instead of generic prose that many skim.
• Risk: long or fine-grained traces increase cognitive load, invite narrative overfitting (“why did I do that weird move?”), and can crowd out focus on current retention tasks.

Who benefits most

• Supported novices & intermediates with moderate self-regulation: likely to gain from brief, structured trace views plus 1–2 focused prompts (“Which earlier move best shows the main rule?”).
• High prior-knowledge / strategic learners: small extra benefit; some may use traces to optimize strategies for future tasks, but effects on core concept gains are modest.
• Fragile novices, high anxiety, or strong rumination: higher risk of over-analysis and self-critique; better with either no trace or a designer-selected micro-trace framed as “example paths,” not as a full personal log.

Design implications

• Use short, designer-filtered snapshots of each learner’s path (few moves, aligned to core contrasts), not scrollable logs.
• Time trace-visibility immediately before or interleaved with delayed retention items, always with explicit, forward-looking prompts ("What pattern do these moves suggest? How would you apply it here?").
• For at-risk profiles, either turn traces off or use heavily curated, anonymized patterns (“typical paths”) instead of full personal histories.