For adult learners on online platforms, how does alternating between unguided attempts and worked examples across multiple spaced sessions compare to massing the same sequence in a single session in terms of long‑term retention and transfer of skills, and are there identifiable thresholds of task difficulty where spacing this attempt–example cycle stops being beneficial?

Answer

Alternating unguided attempts and worked examples across multiple spaced sessions generally produces better long‑term retention and transfer for adult online learners than massing the same attempt–example cycles into one long session, as long as the tasks stay in a productive struggle range. When tasks are too easy or too hard, the added benefit of spacing the attempt–example cycle shrinks and can effectively disappear at the extremes.

More specifically:

• Long‑term retention: Spaced attempt→example cycles repeatedly trigger retrieval and correction after some forgetting, so learners remember procedures and concepts better days or weeks later than if they did the same total number of problems and examples in one sitting.
• Transfer: Spacing gives multiple opportunities to re‑encode and re‑compare their own attempts with worked examples in slightly different contexts, which improves abstraction of underlying principles and thus transfer, compared with a massed block where attempts and examples blur together.
• Difficulty thresholds:
  • When tasks are very easy (little or no productive struggle), spacing the attempt–example cycle adds little beyond generic spacing benefits; both massed and spaced schedules lead to ceiling performance.
  • When tasks are in a moderate productive‑struggle band, spacing cycles is clearly beneficial: each session’s unguided attempts are challenging but doable, and worked examples can connect to existing knowledge without overload.
  • When tasks are very difficult for the learner’s current level (high intrinsic and extraneous cognitive load, frequent failure on attempts), spacing the same hard attempt→example cycles does not reliably improve outcomes over massing. Learners forget almost everything between sessions, repeatedly re‑experience confusion, and cannot effectively use the examples as contrasts. In this regime, it is often better to lower difficulty or increase support (e.g., more scaffolded examples) than to rely on spacing alone.

There is no single universal numeric cutoff for when spacing stops helping, but a practical threshold is crossed when unguided attempts rarely reach partial solutions or interpretable reasoning and learners report sustained confusion across sessions; beyond this point, spacing the attempt–example cycle ceases to be reliably beneficial and can be neutral or even demotivating.