HCD.10 - Choose Whether and How to Vary, Space, Interleave, or Progress Practice
Type: Human Capability Development practitioner practice-progression pattern Status: Stable
HCD.10:0 - Use This When
Use this pattern to decide whether to keep or change an existing practice arrangement, or to establish how its episodes were carried out. Start with the learner’s target action, evidence from the relevant attempts and any later working conditions that matter to the question. The question may concern which changes should alter or preserve the action, performance after an interval, selection among similar methods, or a change of challenge or help. Identify what an observation under those conditions could establish.
A condition-sensitive practice progression identifies the next episodes or an adequate arrangement to retain. For each selected change, state its purpose, task conditions, what the learner must retrieve or can consult, permitted help, the expected observation, when to continue or reverse the change, when to stop safely and what would warrant reopening the question. Specify separated times for spacing. For a completed episode, identify the help received and what the learner did. A proposal can combine compatible changes or retain stable supported practice.
Spacing distributes repeated practice across separated times. Choose retrieval, source access, AI, interleaving, difficulty and other support from the contribution being developed. The same legitimate references and tools can remain available on both occasions.
Retain an arrangement that serves the required action. Choose a change when it would help expose or develop that action. Explain why the learner should keep or change a response, compare tasks in a particular order or revisit one after an interval. State when to retain, change or restore support. A failed attempt can reveal a faulty design assumption.
Use HCD.6 for a first representative task, HCD.7 to arrange suitable available support, HCD.8 when a missing contribution must be obtained or built, and HCD.11–HCD.13 for an assessment of performance, transfer or retention. If the basic action remains infeasible, address the limiting condition. Before attributing a failure to capability loss, establish whether applicability, access, expression, adaptation, enactment or support explains it. An applicable E.23.CAE result can help where the distinction matters.
HCD.10:0.1 - Working Terms
| Name in this pattern | What it denotes |
|---|---|
| decision-bearing condition | A task, context, resource, support, configuration, timing, risk, or evidence value whose change can alter the warranted human action or Method. |
| variation | A designed change that tests what should alter or preserve the target action. Change a deciding condition, or change a potentially misleading cue while holding the deciding relation fixed. |
| spacing | Repeated practice across separated times. Identify the interval and relevant intervening activity. Choose retrieval and support separately. |
| retrieval condition | What the learner must recover without reopening a named source or answer, what remains supplied, and why that distinction matters to the target contribution. |
| source condition | Which references, data, patterns, tables, calculators, tools, or other resources are available and what the learner must do with them. Source use can itself be part of expert performance. |
| support condition | The human, tool or task support permitted for an episode and what is available when it takes place. This can include AI, a cue, a worked example or specialist help. Identify what the learner received when interpreting a performance. |
| interleaving | Mix cases when the learner needs to distinguish similar situations and decide which response or method fits. In a block, cases with the same learning target stay together. |
| progression | A conditional change in challenge, complexity, independence, speed, variability, consequence, or support based on prior attempts and the intended later use. |
| condition-sensitive practice progression | The next episodes or stable arrangement selected for the intended use, with its purpose, relevant conditions, expected observation, response rule and limits. |
An episode can combine variation, spacing, retrieval, source use, interleaving and changes in support or challenge. Explain what each selected feature contributes and which observations could call its design assumption into question.
HCD.10:1 - Problem Frame
Repeated success on an unchanged task provides evidence of performance under its observed timing, sources and help. A changed configuration, competing method, missing cue or later return calls for evidence under the conditions of that question.
Delaying a task, removing sources and guidance, mixing cases and increasing complexity at the same time can obscure why a learner failed. The needed data may have been inaccessible, or the combined demands excessive.
Use the target action and available evidence to decide what to retain or change. For each proposed change, identify a response that would support keeping, revising or reversing it. Stable practice can serve an established action or help a learner acquire one that variation would obscure.
HCD.10:2 - Problem
A calendar locates planned opportunities. To establish spacing, identify the repeated practice at separated times. To interpret a later response, identify intervening use, sources and support. A schedule on its own gives no evidence that the learner retrieved, transferred or retained the action.
Randomly mixing cases or raising difficulty can leave the target discrimination unspecified. If a learner still lacks the basic operation, mixing tasks while removing help and blocking a necessary source can make the next failure uninterpretable.
Removing a tool or reference required in later work changes the task. A change of support can be useful for examining a particular human contribution. Distinguish performance with ordinary equipment, independent action, performance after a change of support, transfer and retention.
HCD.10:3 - Forces
- Stable supported practice can establish the basic action, while unchanged conditions can produce cue-bound success.
- Variation can expose decision rules, while cosmetic or excessive variation adds noise without a useful discrimination.
- Separated episodes can support later access, while one universal interval cannot fit different tasks, histories, and horizons.
- Retrieval can exercise unaided recovery, while expert later Work may legitimately depend on references, tools, or current data.
- Interleaving can expose differences between similar alternatives, while a block can reveal what diverse instances share or help establish an unfamiliar operation.
- Less guidance can expose a human contribution, while premature withdrawal can increase avoidable load or unsafe error.
- More challenge can expand the practice envelope, while a failed attempt may indicate access, applicability, support, or task-design failure instead of capability loss.
- A planned later episode provides an opportunity to observe. To assess transfer or retention, use evidence of what happened, including the task, delay, support and intervening practice relevant to the claim.
HCD.10:4 - Solution
When designing a progression, use the target action, prior evidence and later working conditions to decide what to retain or change. Select timing, task variation, retrieval, source access, help, order and challenge for their own purposes. Return the proposed episodes, the intended observations and the rule for continuing or revising. If the question concerns completed practice, establish which episodes took place, with what resources and help, and what the responses showed. Use that evidence to apply the progression rule when a further practice decision is needed.
HCD.10:4.1 - Bind the Target, History, and Later Conditions
Establish the following as they bear on the task.
- the human holder, target action, selected Method, representative task family, domain criteria, protected errors, and intended later Work;
- relevant previous attempts, including HCD.9 results where applicable, first-response conditions, help received and, where relevant, how it was used, corrections or grounds for retaining the action, any retries, open difficulties and intervening practice;
- the later conditions that matter: task variation, configuration, source and tool access, AI use, provider help, time pressure, consequence, and horizon;
- the basic action the learner already performs, any established difficulty and what remains unknown about their performance; and
- the receiving use of the next observation and the stronger HCD.11–HCD.13 inference, if any, that remains separate.
Use evidence of the target action and support to justify greater challenge. If that evidence is inadequate, recover a suitable record, arrange a bounded HCD.9 episode if possible, or state the limit on the proposed progression.
HCD.10:4.2 - Select the Smallest Change that Can Answer the Question
When the current arrangement already serves the practice question, retain it. Otherwise use the table to select a change that can answer the question.
| Question about the next practice | Candidate change | What the design needs to distinguish |
|---|---|---|
| Does the learner preserve or change the action for the right reason? | variation | Which deciding condition or incidental cue changes, which action should be preserved or changed, and which rival response is exposed? |
| Can the learner perform the action again after a material interval? | spacing | Identify the repeated action, separated times, reason for the interval, relevant intervening activity and observation sought on return. |
| Must a contribution be recovered without reopening a source? | retrieval | What is retrieved, what remains supplied, why unaided recovery matters, and what support is still allowed? |
| Must the learner recognize a shared basis across varied instances? | blocked comparison | Which instances belong together, what common relation must be noticed, and what action reveals it? |
| Must the learner distinguish which response or method fits a case? | interleaving | Which alternatives are mixed, which similarities cause confusion, and what response will reveal the needed discrimination? |
| Is stable performance ready for a changed level of challenge or help? | progression | Which feature changes, what prior evidence supports the change, and what condition pauses or reverses it? |
Several changes can belong in one episode when their contributions and the meaning of a failure remain distinguishable. Otherwise change one relevant feature at a time or use comparable branches.
HCD.10:4.3 - Design Variation around the Needed Discrimination
To practise adaptation, change a condition that can change the warranted action. A release can require FW8 instead of FW7. Twenty-four test hours can support work that 22 hours cannot. Eighty records representing only 60 requests, readings near 448 versus 400 kPa, missing source information or a different authority can each alter the appropriate response in its own task.
To practise preserving an action, keep the deciding relation fixed and change a cue on which the learner may wrongly rely. HCD.6.1:4.4 describes how to construct this contrast. The challenge is to retain the warranted action under a misleading change.
Before the attempt establish the intended discrimination and the domain grounds for judging it. The relevant response could be a justified refusal, continuation, calculation, request, method change or retention of the earlier action. Ask for the action that reveals the distinction. Ask the learner to explain it if explanation is itself part of the target or resolves a question that matters to interpretation or the next step.
HCD.10:4.4 - Space Repeated Episodes and Select Other Changes Separately
Identify the repeated action, later-use horizon, feasible practice windows and time for first attempts, feedback, correction and any needed retries. Choose a provisional gap and what response would justify changing it. Where the learner cannot yet perform under the intended source and support conditions, first make an accessible attempt possible. HCD.9 can help supply practice and feedback. A delay alone supplies no missing instruction.
A prior comparable delayed attempt can guide the initial gap. Check intervening use and conditions when interpreting what remained accessible. Without suitable prior evidence, choose a feasible separated return early enough to detect difficulty and permit repair before later use. Spread the remaining opportunities over the available period, retaining a later return if the horizon matters. An immediate successful retry shows performance on that retry. Estimating a personal forgetting curve requires evidence across delays.
When using a return to judge access after the interval, preserve the learner’s first response before new feedback or answer-bearing help. Keep the sources and tools permitted for the target task. Record relevant omissions and errors, response time if the criterion requires it, intervening practice, help received and its use where that can be established. Use the evidence to plan another opportunity when that decision is needed.
| Observed response under comparable conditions | Next practice decision |
|---|---|
| A required contribution is missing. | Establish whether the learner could enter the task and reach or recover the contribution. Restore instruction, access or help as needed and obtain a usable retry before choosing another delay. If the response was accessible before the interval but is unavailable under comparable conditions now, try a shorter feasible gap after relearning when planning the next opportunity. |
| The response is inaccurate. | Address the observed misconception, discrimination or execution error with feedback and a comparable retry. Try a nearer return when the delay has exposed unreliable access to an earlier response. Repeated error directly after feedback points to a change in learning method or support. |
| The answer is accurate but too slow for the criterion. | Identify which contribution took too long, including any legitimate source search. Practise it with feedback while preserving correctness. Keep or shorten the gap while speed remains the unresolved requirement. |
| The required response is accessible with effort and meets the working criterion. | Keep a comparable gap while that challenge is useful. A further comparable response can support trying a longer gap. |
| The response is readily accessible and meets the criterion. | Try a longer feasible gap when it serves the later-use horizon and still leaves a repair opportunity. Change the gap without simultaneously withdrawing useful support. |
| A change in access, help, task or intervening use prevents interpretation. | Restore the relevant condition or use a comparable observation where possible. Otherwise identify the limit and leave the interval provisional. |
Choose the next gap among available windows, measured from the previous relevant practice. Rescheduling uses a remaining opportunity. If relearning consumes the budget, report the shortage and which later-use question the remaining opportunities cannot answer. For consequential work retain necessary support or arrange another way to supply the result. A claim of readiness requires performance evidence.
For each episode identify the conditions that bear on the intended action or observation.
- whether the learner retrieves a contribution or receives it;
- which sources, patterns, data, tables, calculators, tools, and AI are available;
- which human, tool, cue or worked-example support is permitted, and which help the learner received if reporting a completed attempt;
- whether the task, configuration, time, consequence, or Method choice also changes; and
- what the planned episode could reveal and, once performed, what its result supports.
Two performances of the repeated practice at separated times establish spacing even when the references and tools stay available. Closing a source or removing AI changes the support conditions. A delayed response following intervening work is evidence under that practice history. To claim retention without intervening use, establish that condition too. These response rules support revisable scheduling. An optimal personal interval would require other evidence.
HCD.10:4.5 - Choose Blocks or Interleaving from the Needed Comparison
Choose order from the distinction the learner still needs. HCD.6.1:4.2–HCD.6.1:4.4 can help construct the examples and the response that exposes it. Use an existing adequate comparison where available.
If the learner misses a shared relation across instances of one category, group a few together. Vary incidental features and ask the learner to use the relation that stays constant. This can help even when they already perform the individual operation.
If the learner confuses similar categories or methods, place contrasting instances close enough to compare the deciding difference. Then mix further instances without announcing the required route. Ask for the discriminating choice or action. If explanation is part of the target or would settle what the choice means, ask which condition makes one route applicable and how it bears on a serious alternative. Avoid a predictable alternation that gives the answer away.
A within-category block, a close comparison across categories and a mixed set answer different difficulties. Use subsequent responses to choose the next arrangement. If the common basis remains hidden, another within-category comparison may help. If the learner still confuses alternatives, a cross-category comparison may help. When the basic operation is unavailable, a worked example or stable supported practice may be needed before adding selection demands.
Keep cases together or close in time when the learner needs to perceive their relation. Choose the gap before revisiting the practice separately under §4.4. Long gaps between the members of one comparison can obscure what they share or change. Preserve legitimate tools and support. Once the learner performs the work, use their response to judge the order.
HCD.10:4.6 - Progress Challenge and Support Conditionally
When progression is justified, change a material feature. Possibilities include combining contributions, changing speed or consequences, varying cases, reducing an answer-bearing cue, changing AI or tool support, asking for a more independent first move or approaching later working conditions. Retain the safety, privacy, authorization and domain requirements for the task.
Choose the next practice from the observed response and the intended work. After an independently successful variant, a later case can omit the detailed cue while preserving reference access. An applicability error may call for the working entry and one discriminating question. A calculation error may call for calculation feedback and a changed value. Repeated dependence on help may warrant comparing performance with and without that particular aid.
Pause, reverse the change or redesign the practice if an error becomes unsafe, the basic operation is infeasible, removing support changes the intended task, the learner cannot use the feedback or the observation no longer distinguishes the design assumption.
If HCD.4’s local qualitative model guides an episode, identify the relation in the model and the observations supporting this learner’s current position. Choose whether the opportunity should test a proposed transition or consolidate an established contribution. Greater complexity, less guidance and a different strategy are distinct changes.
Use the HCD.6 pathway as a revisable plan. A harder task can need renewed help while the learner’s earlier independent performance remains established for its original conditions. N24 succeeded in an unfamiliar context and failed in a familiar one (HCD.6:5.6). An outcome crossing the proposed ordering warrants examining the model and the cues. Use HCD.11 for the needed performance judgement, HCD.4 for a material revision of the proficiency model and HCD.14 if the development arrangement needs to change.
HCD.10:4.7 - Differentiate Apparent Failure before Assigning More Practice
When competing explanations matter to the required judgement or next work, examine recognition of applicability, access to the response or source, expression, adaptation, enactment and conditions of support. Use applicable evidence, including an E.23.CAE differential where helpful. Choose the HCD response from what is established and what remains open.
An unavailable report or AI, inaccessible interface, invalid task, excessive load or changed configuration can account for a failed response. Where the cause is known, choose the relevant remedy or report its effect on what can be inferred. Options include restoring the condition, retaining supported practice, changing the variant or choosing another method. Intensify challenge only when it addresses the established need.
HCD.10:4.8 - Return the Progression and Observation Rule
For a practice decision, return the selected arrangement and its grounds. For a report on completed practice, return the episodes, conditions, learner responses and limits that answer the question.
| Part of the result | What establishes it for this use |
|---|---|
| progression boundary | Holder, target action, Method, task family, later conditions, criteria, risks, and receiving use |
| prior evidence | Relevant attempts, help received, corrected or sound actions, any unresolved difficulty and what practice intervened |
| practice decision | Which conditions to retain or change and why, including any selected variation, spacing, retrieval, sources, interleaving, support or challenge |
| episode design | The task, materials, available sources and help, order, workload and feasible times. For spacing, identify separated episodes and their interval. |
| expected observation | The action or other performance to observe, any competing response it should distinguish and what would support changing or retaining the design |
| response rule | Which findings warrant continuing, stabilizing, pausing, restoring support, redesigning or using another method, including an E.23.CAE differential if needed |
| qualification | What the design could reveal, what a completed episode established and which stronger claims still need evidence, as applicable |
| continuation | Any next episode or evidence need, a safe stop or precise gap, and what could justify reopening a question that remains material |
A calendar can communicate the selected opportunities. A learning product may implement the sequence. Explain how the task serves the learner’s target, which evidence it could supply and when its design might need revision.
HCD.10:4.9 - Recognition and Assurance
Recognition. Use HCD.10 when a proposal to make practice harder, space it, mix topics, test retrieval or remove help leaves its purpose or response rule unclear. Establish which distinction, timing, sources or support the task calls for. A bundled schedule can make the cause of failure difficult to interpret.
Assurance. For stronger reliance, check prior performance, the target contribution, representativeness of the tasks and any applicable risk conditions with their competent source. For a selected change, establish why it serves the target and what evidence would justify keeping or revising it. A claim that spaced practice occurred needs evidence of the repeated action at separated times. For interleaving, show which similar cases the learner must distinguish. A response may narrow the performance claim even when the design remains suitable. Retrieval, source access, assistance and challenge each have their own grounds. For a claim about transfer or retention, use evidence suited to that claim.
HCD.10:4.10 - What Changes in Practice
For completed practice, establish which conditions the learner encountered and what the observed responses show. When choosing future episodes, identify the action to change or preserve, the order, timing, retrieval and support that would reveal it, and when the responses would warrant retaining or revising the plan.
HCD.10:5 - Archetypal Grounding
HCD.10:5.1 - Choose and Revise Scheduled Returns
In this constructed inventory case a learner needs to apply eight brief checking rules at work on day 18. The rules sheet remains available. Each check should be correct within 30 seconds. Today’s corrected attempts and retries met the criteria. No delayed attempt has been observed. Four ten-minute windows can be booked or moved, starting on day 2.
The practitioner provisionally books days 2, 6, 11 and 16. Day 2 is the first available separated return. Day 16 leaves time for correction before day 18. In each ten-minute window, four minutes cover first responses on all eight checks, three cover feedback and correction and three cover targeted retries. The four bookings reserve 40 minutes. These dates fit this learner’s horizon and resources. No optimal interval or forgetting rate has been estimated.
On day 2 the learner omits one check and misapplies another despite access to the sheet. The practitioner preserves the first responses, addresses the two difficulties and obtains usable retries. They move the day-6 booking to day 3, the earliest available return, to try one day after relearning, following the unsuccessful two-day gap. Days 11 and 16 remain the two later provisional bookings. If the day-3 response is accurate but too slow, practise the slow contribution and keep or shorten the gap. An effortful response meeting the criterion supports a comparable gap. A ready, accurate response can justify a longer one. If early repair uses a later booking, report the lost delayed observation and reconsider the remaining budget or workplace support.
The day-18 work observation needs its own task and assistance conditions. Account for any intervening practice or consultation. The corrected retries give evidence about those attempts. Even completing all 40 planned minutes would leave the later response open. To report which rules the learner could apply on day 18, use evidence from that occasion.
FoundationsProgramme revisits six selected patterns at weeks 8, 17, 25, and 34. Week 8 returns to F.1 and F.0.1 in a new quantitative source question after about six weeks. Week 17 revisits C.16 and B.5.2 in a manager or engineering case outside the main project. Week 25 combines F.1, C.16, and A.10.1 around sources, measurement, and revision for a configuration question. Week 34 uses F.1, C.16, and C.11 in an integrated case.
The FoundationsProgramme schedules repeated practice at separated times. Each two-week topic also includes an independent variation within the time set aside for earlier learning. The programme records which methods were used between sessions. Interpret later observations with that delay and intervening work. A claim about retention without intervening use requires evidence that there was no such use.
Week 8 keeps references and ordinary tools available and allows five minutes for independent recognition before search. Week 17 permits pattern texts and AI and includes a short separate internal-core probe. Week 25 retains the same means and records intervening use. Week 34 provides normal resources for the main task and a separate fragment with changed support. These are choices about access within particular episodes.
HCD.10:5.2 - Keep the Thirty-Minute Core Probe and Its Fifteen-Minute Fragment Distinct
The final core probe lasts 30 minutes. During one 15-minute fragment AI is unavailable, while the supplied tables and an ordinary calculator remain available. The learner must recover the specified time or configuration constraint, reject an unsupported release and propose a supported next move. The bounded question can be answered without a large calculation or full written solution.
Normal pattern and AI access remains permitted for the programme’s main task. For the 15-minute fragment, record whether the person made the specified contribution with the remaining resources and what other help was needed. Use the observation for that change of support. HCD.13 addresses a further question about delay, intervening practice and retention.
HCD.10:5.3 - Vary the Release Case instead of Rewarding Refusal
One episode supplies a report about FW7 when the release requires FW8. Refuse reliance on that report for this release. A later case supplies compatible HW3/FW8/SCHED2 evidence and 24 test hours. Ten changes in three homogeneous batches need 23 hours. The available testing time supports proposing ten changes in this release. Demand for twelve leaves a growing queue. Releasing the ten changes requires the release manager’s authorization. With only 22 hours the promise of ten changes becomes unsupported again.
The learner needs to recognize the configuration covered by the evidence and apply the capacity bound separately. Under the current conditions, either refusal, continuation or revision of the constraint can be warranted.
Variation can also preserve the answer. A learner had chosen a measure by the visual template of a report. Two reports now use different templates for the same measurement question and relevant data. Selecting the same warranted measure and explaining its basis tests reliance on the template. Changing only the client name would add little unless it exposed a consequential cue.
HCD.10:5.4 - Choose the Order for the Missing Contribution
A learner who cannot yet compare energy per volume receives a worked example, integrated data and stable supported practice. The next attempt changes the values and keeps the operation and source access. Once the learner can perform the operation, a mix of pump explanations, ticket-load comparisons and release-configuration cases can address a separate need to choose the method.
A block can help a learner who already follows every arrow in a diagram and can find what stops when an input is removed. In this constructed case, the learner recognizes a shared prerequisite only in a familiar layout. The teacher presents three layouts together, each showing one result required by both later actions. The learner points to that result and explains what its loss would prevent. The block enables comparison of the common relation.
Another difficulty calls for a different order. The learner confuses a shared prerequisite with two independent inputs when drawings look alike. The teacher places one example of each side by side using similar visual conventions. The learner chooses whether to check one common input or two distinct inputs before the later actions and explains the choice. Further cases are mixed without category labels or predictable order. HCD.6.1 helps construct either contrast. Observing these attempts can establish what the learner did. Attributing a difference to order needs separate grounds. A return on another day is a separate scheduling decision.
Following an independently successful variant, the next design may omit a detailed cue while retaining ordinary references. If the learner then cannot identify the task, restore the working entry and one discriminating question. Choose the support from the observed difficulty and intended work.
HCD.10:5.5 - Do Not Diagnose Loss from an Inaccessible Source
The later task requires a current configuration report. It is inaccessible, and the cached report concerns different firmware. This episode establishes an access problem. It supplies no new observation of the learner applying the current rule with the required report. Restore the relevant source if possible or return the access limit. If a remaining distinction matters to a development decision, use the applicable E.23.CAE differential.
HCD.10:6 - Bias-Annotation
Lenses: Gov, Arch, Onto/Epist, Prag, Did. Scope: selecting variation, timing, task mixing, retrieval, source, support, and challenge changes across bounded human practice episodes.
Visible effort, novelty and calendar completion can draw attention away from the target contribution. Judge the learner’s response under the conditions in which it occurred. Section 8 gives examples of how the interpretation changes a practice decision.
HCD.10:7 - Conformance Checklist
- Identify the learner, target action, selected method, representative task family, relevant prior attempts and intended later conditions.
- For each selected change, identify its purpose, the response or performance it could reveal, any consequential alternative that response would distinguish, and how to use the finding.
- Variation tests a needed discrimination, including preservation of the action when an incidental cue changes.
- For a spacing design, specify the repeated action and separated times. For an account of performance over an interval, establish which episodes and intervening activities occurred.
- Retrieval, source access, tools, AI, human help, task variation, and challenge are selected separately from spacing.
- Interleaving serves a distinction the learner needs to make, such as choosing between two methods or applying one method differently when the conditions change.
- If the basic operation remains difficult, assess whether stable supported practice can make it possible. Identify any required support that is unavailable.
- Check whether the effects of selected changes can be interpreted. Separate features or use comparable branches when needed, and identify conditions for pausing or restoring an earlier arrangement.
- Preserve resources required in later work, unless the task expressly asks what the person can do under a different support condition.
- Apparent failure is differentiated from access, applicability, expression, adaptation, enactment, support, and task-design failures when material.
- Keep a proposed opportunity distinct from observed performance. Use evidence suited to any claim about cause, an optimal interval, transfer or retention.
HCD.10:8 - Common Anti-Patterns and How to Avoid Them
| Problem in the proposed practice | Corrective decision |
|---|---|
| Several difficulties are added together, so a failure cannot identify the assumption that failed. | Select each change for a named purpose and keep its failure interpretation recoverable; separate changes or use comparable branches where necessary. |
| Dates alone are treated as a spacing effect, transfer, or retention. | Identify the repeated episodes, interval, intervening activity, task, and support. Use HCD.12 for a transfer claim or HCD.13 for a retention claim. |
| A later return automatically becomes closed-book recall. | Select retrieval, source access, tools, AI, and human help separately from timing. |
| Random mixing replaces a needed comparison. | Group cases that reveal a common basis within one category, or compare and mix instances to reveal a difference between similar alternatives. Establish which response would show the learner using the relation. |
| Help is removed on every return, including means needed in later Work. | Change support for the particular human contribution, preserve legitimate equipment, and restore help when the withdrawal prevents useful practice. |
| The next task becomes harder despite an unstable operation or an uninterpretable result. | Address the limiting condition or stabilize the operation. Increase challenge only when the evidence supports it and the next response can guide continuation or reversal. |
| Failure is called forgetting before its conditions are understood. | Distinguish applicability, access, expression, adaptation, enactment, and support before assigning repetition for presumed loss. |
| The final probe is reported as 15 minutes entirely without AI. | Report the 30-minute probe, its 15-minute no-AI fragment, the tables and calculator still available and the narrow action required. If reporting performance, state what the learner did under those conditions. Normal support remains available for the main task. |
HCD.10:9 - Consequences
The practitioner can select interpretable changes, retain adequate stable work and revise the arrangement after a revealing response. A practice sequence can revisit an action after an interval, expose a needed discrimination or adjust particular support.
Interpretable design takes time to establish relevant prior attempts, intervening practice, task similarity, source access and help. A combined change whose failure would be hard to interpret may warrant a narrower comparison.
HCD.10:10 - Rationale
The useful comparison relates the target action, the task conditions and the observation. Variation, timing, retrieval, source access, support, case order and challenge change different parts of that relation. Distinguishing their contributions makes the progression revisable and allows legitimate tools to remain available.
A later practice attempt gives evidence about the learner’s action under its observed conditions. HCD.11–HCD.13 use the task, delay, intervening practice and help as relevant to stronger claims about performance, transfer and retention.
HCD.10:11 - SoTA-Echoing
Which changes in variation, timing, order, retrieval, support or challenge would serve this learner’s target and the evidence needed about it?
Carpenter, Pan and Butler’s 2022 review (COG-01) supports spaced encounters and effortful retrieval as candidates for durable access under the conditions reviewed. It leaves the particular gap, repetition count, threshold and curriculum to be established for the intended use. Brunmair and Richter’s 2019 meta-analysis (COG-03) distinguishes discovering the common basis within a category from discriminating between similar categories. It also distinguishes close comparisons across categories from spacing later encounters. The heterogeneous results support choosing between blocking and interleaving for the discrimination at hand. Samples largely drawn from students and uneven coverage of materials limit transfer to a professional schedule. Sections 4.2–4.5 use these differences when selecting timing, retrieval and task order.
Mettler, Massey, Burke, and Kellman (2020) used accuracy and response time to adjust item presentation in country-fact experiments. The studies distinguish efficiency per presentation from raw accuracy and examine retiring mastered items. This supports using response evidence to adapt spacing. Applying the tested algorithm or its numerical intervals to a work calendar would require grounds for that transfer. Section 4.4 instead gives a qualitative procedure for a known horizon and budget. The provisional bookings in §5.1 depend on the initial gap, the next available window and the support the task needs.
One easily standardized approach combines a common timetable, closed-book retrieval, interleaving, reduced help and greater challenge. If a learner fails, that combination can obscure whether timing, access, task selection, help, load or capability contributed. Select each feature for the practice it serves and combine them only when the result remains interpretable.
Czyż, Wójcik and Solarská’s 2024 motor-learning meta-analysis (COG-04) reports heterogeneous transfer results comparing random and blocked practice. Its estimate for applied settings was not statistically significant. Adult laboratory results alone leave the professional schedule open. Paas and van Merriënboer’s 2020 review (COG-05) relates worked examples and guidance to prior knowledge and task complexity. Sections 4.5–4.7 use these conditions to choose stable practice, adjust help or progress the task. A lower load by itself does not establish better learning.
Tullis, Goldstone and Hanson’s 2015 laboratory motor experiment (COG-11) found that better assisted practice can be followed by poorer unassisted performance under the conditions studied. Bastani and colleagues’ 2025 school-mathematics trial (AI-04) likewise found that improved AI-assisted practice could coexist with unchanged or worse later unassisted performance, and that the design of safeguards mattered. The studies inform the questions about support in §§4.4–4.7 and 5.2. Adult workplace performance and a particular policy on AI use require evidence from conditions pertinent to those claims.
Designing an interpretable comparison costs more than copying a schedule. It can preserve the relevant discrimination and legitimate tools, identify a difficulty and support safe revision. Reconsider the choice when changes in the target, history, task similarity, horizon, interval evidence, sources, help, risk, a competing explanation or a feasible alternative practice affect the conclusion. A simpler design may support the same decision.
HCD.10:12 - Relations
HCD.5: supplies the selected development Method, proposed mechanism, fit, limits, and evidence need.HCD.6: supplies representative task families, decision-bearing variations, criteria, protected errors, permitted help, and domain-provider returns.HCD.6.1: supplies the construction of contrasts and the choice of the next comparison from an actual confusion; HCD.10 selects their order and separated returns for the present practice need.HCD.7andHCD.8help establish available providers, sources, equipment, environments, capacity and any required human checking. HCD.8 can help compare routes, obtain or build a missing contribution, or establish that existing provision serves the intended use.HCD.9supplies evidence from attempts, help received, feedback, corrections, grounds for retaining a sound action and any retries. Use a remaining question to decide whether the later practice needs changing.E.23.CAEcan distinguish applicability, access, expression, adaptation, enactment, support and capability explanations when the distinction bears on the required finding or practice decision. Use its result to inform that judgement.HCD.11can use a compatible observed attempt to assess performance within its conditions.HCD.12interprets performance in unfamiliar work or learning from a new source. Variation or mixing may supply a task for examining such a claim.HCD.13interprets delay, intervening practice, sources and support for a claim about retention. It needs evidence of the later attempt.HCD.14: may revise only the affected Method, task, schedule, support, or evidence assumption from compatible results.- Learning-product practices can implement the calendar, task sequence, prompts, resources and interfaces. HCD.10 explains their place in the learner’s progression and the evidence needed for the intended conclusion.
- The relevant domain, safety, privacy, legal, employment, provider, platform or assessment practice supplies the conditions for correctness, access, risk and capacity. Consequential decisions belong to their authorized holders.