Practice Drill: Diagnose and Improve Learning Ability
A no-notes drill for diagnosing one learning bottleneck, selecting a strategy, and defining evidence of improvement.
Chinese version: /zh-hant/practice/drills/learning-ability-diagnosis/
Subskill
Diagnose a learning bottleneck, choose one condition-sensitive learning strategy, and specify performance evidence rather than relying on a general claim that a method works.
Performance condition
Without notes, complete the task in 10 minutes.
Case: A learner spends two hours rereading programming tutorials, feels familiar with the material, but cannot write a small loop independently the next day. They have not attempted retrieval, recorded errors, or tested themselves after a delay.
Required response
- Distinguish the external resources, the likely missing or weak world-model relation, and the required ability.
- Diagnose the earliest bottleneck: missing model, cognitive overload, insufficient practice, unsuitable strategy, or another stated reason.
- Choose one strategy from this list: premise check, plain-language teach-back, retrieval with feedback, narrow deliberate practice, metacognitive monitoring, or spacing plus sleep opportunity.
- State exactly how the learner will use the strategy in this case.
- Define one immediate and one delayed measure that would show whether the strategy helped.
- Name a result that would make the learner revise the diagnosis or strategy.
Success criteria
- Does not treat tutorial exposure or familiarity as knowledge, ability, or learning improvement by itself.
- Identifies one earliest bottleneck with a reason grounded in the case.
- Matches one strategy to that bottleneck and condition.
- Defines observable measures, including a delayed performance check.
- Names a plausible reason to revise the plan.
Common error
Listing every strategy at once. That prevents diagnosis: a useful first attempt tests one plausible cause and one targeted intervention.
Constraint
Use one strategy only. Do not look up a solution before writing the six responses. Repeat two days later with a different domain.
Next action
- If successful: Save the unedited response as an attempt and use the same structure on a real current learning task.
- If unsuccessful: Identify whether the missing step was classification, diagnosis, strategy selection, measurement, or revision condition; retry only that step with a shorter case.