HCD.5:4 - Solution
Recover the target action and its observed or proposed entry conditions, compare a small set of Methods that propose materially different change mechanisms, and select only the Method or combination whose task fit, learner fit, support, burden, risk, and evidence plan remain credible. Return external constraints before prescribing development.
Use observations of the target to decide whether to continue or change practice. Report the conditions and what changed; a before/after difference alone does not identify the intervention’s effect. Obtain additional causal evidence only when the receiving decision needs that stronger claim and the inquiry is feasible and worth its cost. If the target observation is unavailable, leave change unknown rather than treating it as a demonstrated lack of improvement.
HCD.5:4.1 - Bind the Selection Use
First distinguish selection for a known person from design for a future audience. Name:
- the person and participation basis, or the future audience and its assumptions; identify the receiving decision and decision owner in either case;
- the later-Work contribution, configuration, horizon, protected result and critical-error conditions;
- the qualified target for that person and any still-plausible rival explanation, or the target contribution and proposed entry conditions for each design branch;
- the person’s starting attempt, including task, permitted and received help, work product, result, errors, uncertainty and evidence window. For future design, state instead the entry task or existing evidence that would distinguish the branches, the allowed help and the response that would select or reopen each choice;
- current strengths, prior relevant practice, available time, access, tools, AI posture, provider contribution, environment, privacy, safety, and other protected conditions; and
- the observation that the receiving decision would use after the Method is applied.
For future design, keep unknown learner, support and resource values as assumptions with a return condition. Apply a personal choice only when the person’s entry evidence supports it. If rival explanations then change the human target or next action, return to HCD.3. If the target-domain Method, correctness criterion, access, permission, tool, specialist result, or safe participation condition is missing, issue a non-training return to its direct owner. Do not make the missing value into a motivation, practice, or knowledge deficit.
HCD.5:4.2 - Generate a Small Candidate Set by Mechanism
Use the observed action difference, or the stated difference between possible entry conditions, to generate serious candidates. The following cues are questions, not an exhaustive intervention classification:
| Observed difference | Candidate Method question | Boundary to preserve |
|---|---|---|
| the person lacks a usable representation or cannot follow expert reasoning | Would an explanation, demonstration, integrated worked example, or prompted self-explanation expose the decision-bearing relation? | Guidance should fit prior knowledge and later be reduced or changed; understanding the example is not independent performance. |
| the person cannot yet notice the difference named by an instruction | Would HCD.6.1 make that difference discernible through an accessible contrast and a response that uses it? | The encounter must be feasible from the learner’s existing capabilities; noticing the difference and performing the target action remain separate. |
| the person performs an operation when named but does not recognize when it applies | Would discrimination practice across contrasting situations improve recognition and Method selection? | Repeating execution with the Method name supplied cannot test this target. |
| a recurrent component is slow or error-prone and constrains the whole task | Would focused part practice with immediate correction stabilize that component? | The component must contribute to the later task; part success does not establish whole-task coordination. |
| separate contributions are adequate but their coordination fails | Would sufficiently whole-task practice with bounded support expose the interfaces and stopping conditions? | A true prerequisite may still need preparation; task realism does not excuse unsafe or uninterpretable complexity. |
| the next attempt repeats a known, actionable difference | What information-specific feedback, corrective activity, and retest would let the person change that action? | Praise, a score, or another attempt is not enough when it does not identify the difference to change. |
| the person has an endorsed goal and feasible action but misses a stable cue | Would a bounded if–then plan or environmental cue support enactment? | Goal attainment is not capability, and a plan cannot repair an unavailable or unsafe action. |
| competent participation in complex Work is needed | Would demonstration, coaching, guided participation, or supervised workplace practice provide the missing contribution? | The provider, task, access, supervision, and consequence boundary must be real; the label “apprenticeship” establishes none of them. |
| the observed limit disappears when data, permission, workload, tool, or support is restored | Is the supported next move an access or arrangement change rather than a development Method? | Return the condition to HCD.7, HCD.8, or its direct owner and preserve any remaining human target separately. |
Spacing, retrieval, interleaving, cue focus, support fading, and other practice features may refine a candidate when their mechanism matches the target. Their names do not select the Method. HCD.10 chooses the condition-sensitive practice change after the governing Method and observations are present.
HCD.5:4.3 - Compare Method Fit and Evidence
For each serious candidate, record:
| Selection position | Required content |
|---|---|
| target contribution | The exact human action or judgement intended to change and its role in later Work. |
| proposed mechanism | How the Method is expected to change that target, what source or prior observation supports the expectation, and which rival explanation remains live. |
| task and learner fit | Starting performance, prior knowledge, task complexity, similarity, support dependence, and the reason this Method fits better than the serious alternative. |
| enacted form | The minimally sufficient explanation, demonstration, practice, feedback, correction, plan, coaching, or support change; identify a qualified professional description when local design would be unsafe or under-specified. |
| provider and access need | Competence, time, data, tools, AI, environment, supervision, permission, privacy, and peak-load conditions required for the Method. |
| burden and risk | Learner and provider effort, displaced Work, error exposure, overload, autonomy, fatigue, clinical or safety risk, and the cost of another attempt or probe. |
| evidence plan | Starting observation, observation of target change, independent or changed-condition attempt where relevant, later-Work evidence needed for the stronger claim, and contamination or measurement limits. |
| alternatives and stop | Surviving alternative, rejected candidate with reason, condition for stopping or returning, and smallest change that reopens selection. |
Compare candidates at comparable enactment effort. A richly specified Method does not defeat an unnamed alternative; make only the smallest alternative description needed to expose a decision-bearing difference. Preserve source limits instead of transporting one average effect, hours threshold, cue, sequence, or professional arrangement to a different task and population.
HCD.5:4.4 - Select a Coherent Method Set
Select the smallest set whose members make distinct contributions to the target. For a combination, state:
- which target or task difference each member addresses;
- whether it prepares a prerequisite, enables the first attempt, supplies feedback, supports correction, or tests changed performance;
- which order is required by dependence and which order is merely convenient;
- which help is allowed during each observation and what the person must do without an answer-bearing cue; and
- when the set stops, changes, or returns because comparable observations show insufficient target change, another limitation appears, or a provider condition fails.
Do not accumulate techniques as insurance. If two candidates propose the same contribution at higher burden, retain the lower-burden adequate candidate unless a material evidence, safety, participation, or future-use value justifies the difference. If methods address different supported targets, keep those targets and evidence plans distinct even when one programme coordinates their enactment.
HCD.5:4.5 - Record the Method-Selection Account
| Result position | Required content |
|---|---|
| selection boundary | Person and participation basis, or future audience assumptions; later-Work contribution, configuration, horizon, evidence window, receiving decision, decision owner and outside owners. Mark which conditions are observed and which are proposed. |
| target and start | For personal selection: qualified target and starting attempt with its conditions, strengths, errors, help, uncertainty and still-plausible rivals. For future design: target contribution, assumed starting conditions and the entry evidence that would select or reopen each branch. |
| selected Method set | Each Method’s action, proposed mechanism, intended target contribution, enacted form, required support, and stop condition. |
| comparison | Serious alternatives, comparable effort, task and learner fit, evidence, burden, risk, and reason for selection or rejection. |
| evidence plan | Observation of target change, permitted help, independent or changed-condition observation where material, stronger later-Work evidence, and inference limits. |
| disposition | selected Method set, conditional design hypothesis for stated entry evidence, non-training return, or Method description or provider result blocked. |
| continuation | The practice-task design, provider/access result, first enacted attempt, missing professional description, outside-owner question, or smallest condition that reopens selection. |
For a future audience, return only a conditional design hypothesis tied to the stated entry evidence and planned diagnostic branches. Do not turn an audience description into an individual prescription. When a competent professional Method or provider result is required and unavailable, name that result rather than writing a generic local substitute.
HCD.5:4.6 - Recognition and Assurance
Recognition. Use HCD.5 when the question is “What kind of development activity could change this person’s specific action?” and a worked example, whole task, focused component practice, feedback loop, coaching arrangement, self-regulation aid, or support change would lead to different next Work. For a future audience, ask the same question conditionally for each proposed entry branch.
Assurance. Stronger reliance needs a qualified target, condition-compatible starting evidence, a task- and learner-matched mechanism claim, source limits, credible provider and access conditions, material burden and risk, and an observation that can detect target change without conflating helped performance with learning. Attribution of that change to the intervention requires a further basis; it is not implied by this observation. Clinical, safety, employment, legal, provider-competence, and target-domain correctness decisions require their direct owners.
HCD.5:4.7 - What Changes in Practice
A practitioner stops asking, “Which learning technique should we use?” and asks, “Which exact action must change, what Method could change it under these conditions, what serious alternative remains, and what observation would show that the target—not merely the activity or score—changed?” When the answer is access, permission, data, workload, or specialist input, the work leaves Method selection with a named return.