Opportunistic Retrieval

Seeing a heart-disease patient on Tuesday already retrieves what you studied about heart disease on Monday — the workday ran a spaced retrieval session that nobody scheduled. Retrieval happens any time knowledge is recalled and used, so every work task that draws on recent study is a real rep: spaced by the job’s own calendar, interleaved by whatever the day throws at you, and costing zero extra study time. In a purely academic setting nearly all retrieval must be planned; once the knowledge is in active professional use, the work changes from scheduling sessions to auditing the reps the job already supplied and topping up only what they missed.

When transfer is close — the material gets used at work in exactly the form it was learned for — planned retrieval can drop to zero. For a knowledge worker whose study feeds daily output, that covers most material, and the scheduling question shifts from “when do I review this” to “what should I study before the work that will test it.”

Retro-label each rep

Labelling is what converts an accidental rep into practice. After work uses studied knowledge, record what just happened:

• Order check. Did the task demand higher-order use — building a strategy, evaluating a plan, synthesising a big picture — or lower-order use, recalling facts and following a known procedure? An interaction handled purely at the factual-recall level leaves higher-order memory untested.
• Type check. Was the knowledge exercised declaratively (stating, explaining) or procedurally (doing), and does that match how it ultimately needs to perform?
• Gap inference. Whatever the day’s tasks did not exercise is the open gap — located without running a single test question.

Close the gap or upgrade the task

A labelled gap has two repairs and one honest exit:

• Targeted planned session. Run a small deliberate session aimed only at the slice work skipped, usually the higher-order layer. Method choice lives in Spaced Interleaved Retrieval and the Interleaving Table.
• Convert the work. Reshape a future task so it demands the missing order: draft the strategy yourself instead of executing someone else’s, critique the plan you produced, force a big-picture view of the case before acting.
• Move on. Time pressure and job responsibilities often forbid conversion — the task gets done the one way it can be, and the next one starts. The reps were still free; only the top-up was lost.

Reorder study to meet the work

• Control the study queue. Work assigns its own tasks; what you study the weekend before is yours to choose. Sequence study by the retrieval opportunities coming up rather than by curriculum order.
• Flip the curriculum. Working in cardiology now with a renal rotation two months out, learn the heart material first and the kidney material later even if the course teaches them in the opposite order — daily cases then supply reps no schedule could match.
• Forecast the reps. Before sequencing anything, ask what the next few weeks will force you to recall and which knowledge type they will demand; study that first.

Manufacture reps the job doesn’t supply

• Add an activity for the missing order. When routine tasks only exercise one level, insert an activity into the workweek that tests another — a review, a write-up, a design critique.
• Teach a junior. A short training session for a peer or junior is high-order retrieval — structuring the whole topic, fielding live questions, moving between big picture and detail — and it produces team value at the same time. One session can plug the hierarchy gap that task-level work leaves behind; WPW is the structured form of the move.

Stop optimizing at the goal

• Optimization is goal-relative. Pushing memory far past what the goal demands wastes the time the whole workflow exists to save, and holding a standard well above the actual requirement turns counterproductive.
• Good enough has a shape. For a low-stakes topic — an isolated problem to solve, a single client presentation, an elective that only needs a pass — the bar is: questions answerable, memory not too leaky, retention depth present. Natural daily retrieval alone may clear it.
• The expertise exception. Knowledge you will need genuine expertise in later deserves long-term retention and depth now, however little the current assessment asks for.

Links into the system

Runs on the same principles as Spaced Interleaved Retrieval, with work supplying the sessions; the order and type labels come from Higher Order Learning; teaching a junior is WPW moved into the workplace; and Encoding and Retrieval decides how many reps the knowledge needs at all.