In adult training systems that currently treat illusions of learning mainly as an individual metacognitive problem (addressed via dashboards, meta‑nudges, and unified hint‑gating), do session-level “collective calibration” summaries that compare a cohort’s unguided performance and hint reliance to their own prior sessions (e.g., “as a group, your transfer to new problems barely improved despite higher confidence and hint use”) reduce illusions of learning and social loafing more effectively than further individual-level tuning—or do such group summaries backfire by diffusing responsibility and encouraging defensive comparisons, especially in low-psychological-safety environments?

Answer

Collective, session-level calibration summaries are plausibly a modest net improvement over more individual-level tuning for reducing illusions of learning and (to a smaller degree) social loafing—but only when (a) psychological safety is at least moderate, (b) summaries are clearly framed as informational and non‑evaluative, and (c) they include some lightweight mechanism that links group patterns back to each individual’s own data or commitments. In low-psychological-safety environments, or when summaries feel like public scorecards without individual linkage or guidance, they are likely to backfire by diffusing responsibility and triggering defensiveness.

In practice, they should be treated as a small, conditional upgrade, not a default replacement for individual dashboards/meta‑nudges.