Can the attempt–worked-example cycle, shown to benefit individual and small-group training, be generalized to mixed-format workplace programs where learners alternate between solo AI-supported quizzes and small-group sessions, such that synchronizing the cycle across formats (e.g., solo unguided attempts feeding into group worked examples on similar problems) improves transfer of skills more than running independent cycles in each format?

Answer

It is plausible that synchronizing the attempt–worked‑example cycle across solo AI‑supported quizzes and small-group sessions yields modest improvements in transfer of skills over running independent cycles in each format, but those gains are contingent on careful alignment of tasks and strong safeguards for individual retrieval and equitable participation. Poorly implemented synchronization (e.g., misaligned problems, weak role structures, dominant-solver patterns) is likely to produce little or no benefit beyond well-designed independent cycles, and can sometimes worsen equity or reliance on AI/peers without improving transfer.