General Manager
If someone immediately replays a conversation and gets a better result, I want to know what changed. They may have understood the decision more clearly. They may also remember exactly which sentence the system rewarded.
Both are reasons to look beyond the second score. In Practice-Native Learning, Altaius's methodology for organising learning around consequential practice, repetition needs a purpose and an evaluation design.
Cepeda and colleagues' 2006 review examined 839 assessments of distributed practice across 317 experiments in 184 articles. It focused on verbal recall. The interval associated with stronger retention depended on how long the material needed to be retained. It does not prescribe a single ideal interval for leadership simulations. [1]
The design implication I draw is modest: timing deserves deliberate attention. A return to practice after an interval can answer a different question from an immediate retry. We still need to test the appropriate timing for the task, learner and intended use.
I distinguish three purposes when planning an illustrative programme.
Those are planning categories, not validated stages that every learner must follow in a fixed order. A novice may need several supported attempts. An experienced learner may need a more demanding variation. A learner who has misunderstood a safety-critical rule needs correction before further independent practice.
Suppose the target is checking authority before making a commitment. In the first case the learner negotiates a delivery date. In the next, a colleague requests an exception to a process. The details differ, but the underlying question is whether the learner identifies who may approve the commitment.
If the second scenario instead measures a different policy rule or requires unfamiliar technical knowledge, a score change becomes harder to interpret. The designer should record what stays consistent, what changes and why the variation is relevant.
I would also avoid making every successful path obvious through identical wording. The learner should need to recognise the decision structure, not a familiar phrase.
A team could begin with an initial attempt and debrief, return for a supported variation later that week, and arrange an independent case after a longer interval. The exact days should reflect the work cycle, the skill and the evaluation question. This is a scheduling example, not a research-backed dosage or a promise of retention.
For each attempt I would record the scenario version, elapsed time, coaching used, relevant prior exposure and the decision evidence. A learner who has already seen the assessment case should not be treated as equivalent to one seeing it for the first time.
Workload matters as well. A calendar full of short exercises can still become unmanageable if every attempt needs a substantial manager debrief. Plan the support capacity alongside the participant time.
An immediate improvement is useful feedback about that attempt. More convincing evidence of learning would include performance on an appropriate new case after time has passed. Evidence of workplace application requires observation outside the simulation as well.
I would report these results separately. Combining all attempts into one upward line can hide changes in difficulty, assistance or familiarity. If learners improve only on the repeated case, the programme has discovered an important limitation rather than a result to conceal.
The aim of replay is to help someone become more capable when the situation changes. The practice plan should make it possible to tell whether that is happening.
Read the Altaius definition of Practice-Native Learning
Previous in this series: Feedback and Debriefing in Practice-Native Learning
