Mira had practised very hard.

She had completed three weeks of beautifully organised Mathematics revision. Every evening she chose one chapter, reviewed the worked examples, completed a clean block of similar questions and checked the answers.

Then she sat a mixed paper.

The mathematics itself had not become harder. The training demand had changed.

No heading told her which chapter the question belonged to. No worked example sat immediately above it. A familiar method appeared in unfamiliar wording. Two questions required ideas from different chapters to be combined. One easy item arrived after a difficult one, when her attention was still carrying the previous failure.

Mira knew a great deal.

She had practised the wrong performance.

Practice Has to Resemble the Capability We Want

Practice specificity is the principle that training should increasingly match the information, decisions, representations and constraints of the performance the learner eventually needs.

This does not mean every lesson should look like an examination.

Novices need explanation, modelling, worked examples, scaffolding and controlled practice. But as a capability develops, the learner must gradually assume the conditions that the final performance will require.

Train the component simply. Then train the capability honestly.

The Final Performance Contains Hidden Demands

An examination question does not test only subject knowledge.

It may also require the learner to:

• retrieve without notes;
• identify which topic is relevant;
• ignore irrelevant information;
• translate between words, diagrams, equations or graphs;
• select among competing methods;
• work under a time constraint;
• maintain accuracy after several previous questions;
• recover after an error;
• check efficiently;
• express the answer in an accepted form.

If practice repeatedly removes these demands, success during practice can overestimate readiness.

Blocked Practice Is Useful, but It Gives Away the Routing Answer

Blocked practice means working on one type of problem repeatedly.

It is often excellent during early acquisition.

If a student is learning factorisation for the first time, several similar examples allow attention to remain on the new structure rather than on deciding among many methods.

But the block itself contains a cue.

The page says: use factorisation.

When the examination removes that cue, method selection becomes a new demand.

This is why later practice must become mixed.

Practice Specificity in Mathematics

Suppose Jonas can differentiate ten functions correctly when the worksheet is labelled “Chain Rule.”

That demonstrates execution under strong cueing.

A more specific performance sequence gradually adds the actual examination demands:

1. differentiate standard chain-rule examples;
2. mix chain, product and quotient cases;
3. remove method labels;
4. embed the differentiation inside a larger problem;
5. change representation;
6. add realistic timing;
7. require independent checking.

The learner is not merely doing harder mathematics. The practice is becoming more specific to the performance environment.

Practice Specificity in English Reading

A reader can become excellent at answering one question type while failing on a full passage because real reading requires switching.

Practice should therefore progress from isolated skills into integrated reading.

• First, identify reference words cleanly.
• Then distinguish reference from cause, contrast and inference.
• Then answer mixed questions in a complete passage.
• Then work under realistic time while preserving evidence quality.

The final target is not “can answer reference questions.”

It is “can read a passage, recognise what each question demands and produce evidence-based answers without external routing.”

Practice Specificity in Writing

Writing practice often becomes unspecific when students memorise polished model paragraphs.

The model can teach structure and language. But if the final performance requires generating ideas from a new prompt, selecting material, planning, drafting and revising under time, the student must practise those decisions too.

Specificity means moving from imitation to generation.

Study a model.

Reconstruct its architecture.

Use the architecture on a different prompt.

Then remove the scaffold.

Practice Specificity in Science

Science performance often requires several layers at once: concept recall, representation, variable analysis, evidence interpretation and explanation.

A student may know the concept in notes yet fail when it appears inside an unfamiliar experiment.

So practice should eventually move:

definition → concept explanation → familiar application → changed experiment → data interpretation → justified conclusion.

The learner trains the route Science assessment actually demands.

Specificity Does Not Mean Copying the Exam

There is a danger of taking this principle too literally.

If every practice task is a full paper, students receive poor opportunities to isolate and repair small weaknesses.

Full performance contains too many interacting variables.

High-performance training therefore alternates resolution levels.

• Decompose: isolate the weak component.
• Repair: stabilise it under controlled conditions.
• Recombine: put it back into a realistic task.
• Verify: check whether it survives under the final performance demands.

Specificity and Transfer Are Not Opposites

At first glance, practice specificity sounds as though students should practise only tasks identical to the final one.

But examinations vary. Real life varies. Therefore one of the things the final performance may require is transfer.

A specific training programme for adaptable performance must include variation.

This links directly to Transfer Distance.

If the final demand is “solve unfamiliar problems using familiar mathematics,” then unfamiliarity itself belongs in the training.

Specificity and Attentional Control

Practice conditions teach students what to attend to.

If every worksheet is heavily labelled, the learner can attend to the label instead of the discriminating features.

If every Science diagram contains arrows highlighting the relevant area, the learner may never practise locating the signal independently.

Specificity therefore includes gradually removing artificial attentional cues so the learner learns to select what matters.

Specificity and Performance Recovery

Most worksheets allow the learner to begin each question in a neutral state.

Real examinations do not.

A difficult Question 4 changes how Question 5 begins.

Therefore performance preparation should sometimes include the recovery skills described in Performance Recovery.

Train not only perfect runs but the ability to resume after disruption.

Specificity and Time

If the final task is timed, timing eventually belongs in practice.

But it should be introduced after enough accuracy exists.

Timing an unstable procedure can train rushing.

Timing a stable procedure tests whether fluency survives pressure.

Specificity has an order.

The Specificity Ladder

1. Modelled: the learner sees the route.
2. Guided: the learner executes with prompts.
3. Independent familiar: same structure, no prompts.
4. Mixed: the learner must select the route.
5. Transferred: representation or context changes.
6. Timed: decision quality must survive a clock.
7. Sustained: performance must remain reliable over longer duration.
8. Recovered: capability must survive a local disruption.
9. Exam-like: multiple demands operate together.

Not every skill needs every stage every day.

The ladder shows how support can gradually approach the real demand.

The “What Are We Actually Training?” Question

Before assigning practice, ask:

• Are we training understanding?
• retrieval?
• fluency?
• method selection?
• transfer?
• timing?
• endurance?
• checking?
• recovery?

The same worksheet cannot optimise all of these equally.

Training becomes more efficient when each task has a clear job.

The Home-Study Version

A student preparing independently can make practice more specific without constantly doing full papers.

• Close notes before retrieval.
• Mix topics after initial fluency develops.
• Occasionally change representation.
• Use short timed sections before full timed papers.
• Track whether errors increase late in sessions.
• Practise known checking routines.
• Return to difficult items after a deliberate reset.

This gradually converts study into performance preparation.

The Tutor Version

A tutor can change one variable at a time.

Keep the Mathematics constant and change the wording.

Keep the wording constant and remove the topic label.

Keep the task constant and add a mild time limit.

This helps identify exactly which performance demand causes failure.

Specificity becomes diagnostic, not merely difficult.

Mira’s Revision Changes

Mira did not abandon chapter practice.

She changed its role.

New and repaired skills were still practised in clean blocks.

Once stable, they entered mixed retrieval.

Once selection improved, they entered changed representations and unfamiliar contexts.

Later, short timed sections tested pacing.

Only then did full papers become the main integration environment.

She still practised hard.

Now the practice was training the performance she actually needed.

The Practice Specificity Test

1. What capability is this practice supposed to build?
2. Which demands of the final performance are currently present?
3. Which demands are intentionally removed because the learner is still acquiring the skill?
4. When will topic labels and prompts disappear?
5. When will method selection be trained?
6. When will representation change?
7. When will timing and endurance be introduced?
8. When will recovery and checking be tested?
9. Does the practice still isolate weak components when repair is needed?
10. Does the programme eventually resemble the performance honestly enough to reveal readiness?

Next: Learn to Choose, Not Just Repeat

Once practice begins resembling real performance, different methods and topics start appearing together.

That creates another high-performance demand: interleaving.

Next: How High Performance Learning Works | Interleaving — Learn to Choose, Not Just Repeat.

Research Note

This article synthesises established ideas from transfer-appropriate processing, retrieval practice, cognitive load, contextual interference and expertise research. The instructional principle is not that practice should be identical to final performance from the beginning. Early learning often benefits from simplified and supported conditions. Specificity increases as the learner becomes ready to assume more of the real task.

Series Note

“High performance learning” is used descriptively throughout this eduKatePunggol series and does not claim affiliation with any third-party branded framework using similar terminology.