HCD.12 — Test Transfer to Unfamiliar Work or Learning from a New Source

Type: Human Capability Development practitioner method pattern Status: Stable

HCD.12:1 - Problem Frame

Use this when a person has demonstrated a contribution in familiar work and someone wants to rely on it after a meaningful change in task, data, tool, setting or coordination. Begin with the earlier performance and name what is unfamiliar to this particular person. A compatible HCD.11 result or equivalent qualified performance evidence is enough to enter.

The first useful result is a bounded transfer account: which contribution survived which change, what required adaptation or help, and what remains untested. It lets a practitioner preserve familiar success while addressing a specific transfer gap.

Also use §4.5 when the question is how the person learns with a newly supplied source. Begin with relevant prior preparation, familiarity, the source, access conditions and the task criteria. The result reports what the person could do after source access and with what help; prior successful execution of the new Method is not an entry requirement.

Keep these questions distinct. Do not call a renamed exercise unfamiliar, infer transfer from course completion, or use an unknown baseline as if prior capability had been demonstrated. When the intended claim concerns transfer of an earlier contribution and that contribution is uncertain, assess it through HCD.11 or leave the transfer claim open.

A delayed test of otherwise comparable work is primarily an HCD.13 question. Actual deployment, release, safety and employment reliance retain their own evidence and authority conditions.

HCD.12:2 - Problem

Familiar success can depend on the exercise’s wording, a named Method, a teacher’s cue or one repeated answer. Under new conditions, the person may fail to recognize the problem, fail to find an available response, apply it without adaptation, or reach an inadequate result. Those failures ask for different next actions.

A transfer test can also reward a new rigid habit. Someone who has learned to reject mismatching evidence may now reject complete evidence as well. A test containing only traps cannot show that the person can make a justified positive move when the conditions allow it.

HCD.12:3 - Forces

HCD.12:4 - Solution

HCD.12:4.1 - Recover the prior claim and the meaningful change

For a transfer claim, recover the same person’s earlier contribution, task, conditions, permitted support, criterion and observations. For a declared new-source-learning question, enter through §4.5 instead; use §§4.2–4.3 for task criteria and observations as applicable. Identify relevant prior exposure to the new task or response. If familiarity is uncertain, keep that uncertainty in the result.

Describe the material change and why the old answer might now fail. Examples include evidence from a different configuration, reduced capacity, a different tool with the same required result, or a new specialist interface. A change of names, numbers that leave the decisive reasoning untouched, or formatting alone is weak evidence of unfamiliarity.

Preserve the target requirement or explicitly identify its change. If the new task requires another professional Method or prerequisite, obtain that contribution or report the new requirement; do not diagnose failed transfer of the old contribution merely because the task now asks for something else.

HCD.12:4.2 - Design a discriminating unfamiliar task

Choose a task the person has not rehearsed in its decision-bearing form. State the changed fact, the response it should change, the criterion and the allowed means. Ask how a copied answer, an indiscriminate refusal and a lucky guess would fare.

Where the work requires both action and restraint, include sufficient-evidence and insufficient-evidence cases. A useful small comparison may be one condition that supports action and another that changes its limit. Broader transfer claims need coverage of the material variation they include; a single convenient task cannot supply it.

Keep irrelevant difficulty controlled where possible. If both the tool and the work problem change, either explain why that combined condition is the intended question or separate the changes enough to interpret failure.

Before use, have the domain criteria and expected consequences available. A task whose own correct answer is disputed cannot settle the person’s transfer.

HCD.12:4.3 - Observe the first attempt under its declared conditions

When independent recognition is part of the target, present the work problem before naming the Method, PatternID or decisive feature. The practitioner may use the ordinary sources and tools specified by the target. Independent recognition here means the required selection was not supplied; it does not mean working without every external resource.

Keep the following distinctions only as far as observations support them:

An incorrect final answer alone may not distinguish these contributions. Use the smallest explanatory observation or qualified differential that would change the next action. Do not force every attempt into five confident labels.

After recording the first attempt, feedback can support learning. Record what it supplied, and keep the corrected attempt separate. A Method supplied by the assessor can support application evidence; it cannot demonstrate the preceding independent selection.

HCD.12:4.4 - Return the transferred contribution and its boundary

For transfer, compare the new observations with the earlier claim under the named change. Preserve supported contributions even when another contribution fails. For new-source learning, return the source, access conditions, observations and limits specified in §4.5 instead of inventing a prior-performance claim.

State the tested change, independent and helped observations, failed criteria, uncertainty and untested variations. When a stronger receiving claim is pursued, state the observation it would require. Use “demonstrated in this unfamiliar case” when that is the scope. Do not use “transfers” without saying what transferred where.

If performance returns after a cue, report the cue-dependent result. A fresh uncued case is needed before strengthening an independent-recognition or adaptation claim. If the new task or access was defective, return that problem instead of treating the result as human failure.

Stop when the intended practice or support decision is supported. Stronger reliance may require several unfamiliar cases, independent assessment, a defensible coverage argument and checks of domain correctness and access. A failed transfer case does not erase the demonstrated familiar envelope.

Reopen only the contribution affected by a new variation, changed source or tool, corrected task, or stronger receiving use.

HCD.12:4.5 - Separate learning from a new source when that is the intention

Sometimes the useful question is how well a person learns a new Method during the task. State that intention before the test, together with relevant prior preparation and what is known about familiarity with the Method. Identify the newly supplied contribution, the exact source and the opportunity to learn from it. Distinguish finding the source from learning with equal access to it.

Observe prior familiarity, initial recognition, first application after source access, adaptation to a material change, help, time, material errors, corrections and unfinished work. Test the new Method’s own contribution; familiar arithmetic or an already known comparison is not evidence of acquiring a different whole Method.

A first application after reading the new source supports a performance claim under those learning conditions; a claim of improvement also needs a relevant comparison. Earlier knowledge may help the person learn from the source. If that contribution or the effect of an earlier programme is the question, specify the comparison needed to distinguish it. Application after source access alone does not establish that effect or retention of the new Method. A “faster learning at unchanged correctness” claim additionally needs comparable conditions, a suitable comparison and actual observations; shorter completion time alone cannot supply it.

HCD.12:5 - Archetypal Grounding

HCD.12:5.1 - Transfer requires a positive proposal and a changed resource answer

Learner-L17 and the following observations are fictional. They illustrate the Method and provide no data about Engineer-E27 or actual programme participants.

Earlier qualified observations show L17 checking configuration evidence and explaining a supported continuation in two fresh cases without a teacher’s configuration cue. In those cases, L17 rejected a report about FW7 when HW3/FW8 was required. Calculation and the 23-hour C continuation were also supported. Those familiar contributions are not themselves new transfer evidence. The earlier observations do not yet demonstrate adaptation to changed capacity.

The new case has a report that actually supports HW3/FW8/SCHED2 and confirms the required checks. Preparation takes one hour per homogeneous batch of up to four; required checks take two hours per change. Twelve changes arrive per week. At 24 available hours, ten changes require 23 hours, while twelve require 27. The release manager may authorize a bounded ten-change use and retains responsibility for the growing queue and release decision.

For this constructed comparison, L17’s task history stipulates that the standalone evidence-use question—a complete report offered without a preceding mismatch problem—has not been rehearsed. That changed question must now lead to a justified positive proposal; repeating the earlier FW7 refusal would be wrong. The task does not ask for a pattern name. Reports, calculator and ordinary source and tool access are available. The source and assistant traces place L17’s first recognition and proposal before any teacher or AI cue supplies the decisive condition.

The useful local return is complete: preserve the supported calculation and configuration evidence and identify capacity adaptation as unproven independently. If stronger reliance on independent capacity adaptation is pursued, arrange a fresh uncued variation before making that claim. HCD.14 can use this result immediately to propose a small practice revision; a retention test is not a prerequisite.

An assessor who accepts “the evidence is never sufficient, so I refuse” would miss the positive-action criterion in the first row. An assessor who scores only the corrected final answer would miss the independent adaptation failure in the second.

HCD.12:5.2 - The named Method supplies what the test meant to observe

In another constructed task, a person makes no relevant first move but produces a correct comparison immediately after the assessor says which Method to use. The result supports application after that cue. It leaves uncued recognition unresolved or failed according to the original task criterion; it does not establish global capability loss.

If the practical question is application with a routinely supplied Method, that result may already be useful. If the work requires the person to select the Method, the next task must preserve that selection problem.

HCD.12:6 - Bias-Annotation

Unfamiliarity is person-relative. A task familiar to the assessor may be novel to the learner, and a nominally new task may have been rehearsed elsewhere. Retain relevant exposure instead of treating cohort membership as proof.

Do not make “caution” a universal success criterion. A person who refuses every case may avoid one error while failing the required useful action. Also check that apparent adaptation is not merely the assessor naming the changed condition.

HCD.12:7 - Conformance Checklist

• A transfer claim has prior qualified performance for the same person and contribution. A new-source-learning question instead names relevant prior preparation, familiarity, the new source and its access conditions.
• The material change and person-relative unfamiliarity are recoverable.
• The task preserves the target or explicitly declares the new requirement.
• Positive action and refusal are distinguished where both matter.
• Required selection is observed before the Method or cue supplies it.
• Allowed sources, tools, help and first versus corrected attempts remain visible.
• The conclusion answers the declared transfer or new-source-learning question, with the supported contribution, failures, help and untested conditions.
• New-source learning, delayed retention and deployment authority are not inferred from the transfer result.

HCD.12:8 - Common Anti-Patterns and How to Avoid Them

HCD.12:9 - Consequences

The practitioner gets a precise next target rather than a general judgement that someone “cannot transfer.” A well-chosen small case can distinguish rigid repetition, cue dependence and useful adaptation without a large new examination.

The cost is explicit novelty and criterion design. Broader variation needs more evidence, and some failures remain ambiguous. That ambiguity is preferable to prescribing the wrong intervention. More refusals or a higher final score can be worse when they hide loss of justified action or increased dependence on feedback.

HCD.12:10 - Rationale

Transfer is a relation between demonstrated contributions and changed conditions, not a property conferred by finishing a course. Preserving the prior claim makes both success and failure informative. Keeping recognition, access, adaptation and result separable shows which new observation could change the conclusion without pretending that every failed answer reveals its cause.

HCD.12:11 - SoTA-Echoing

Working question: how can an unfamiliar task distinguish useful transfer from repeated execution, while allowing learning resources when the intended work uses them?

Selected line — adopt and adapt. The workplace-instruction review by Kraiger and Ford distinguishes instructional activity, learning and later transfer. HCD.12 turns that distinction into the prior-claim and changed-condition comparison in §4.1–§4.4, with the action-versus-refusal test in §5.1.

Serious alternatives: repeat the same exercise under new labels, or remove all resources to avoid assistance. Repetition is useful for practising execution but leaves consequential adaptation untested. A resource-free test is useful for an independently required core but can omit the ability to learn with a source. At comparable task time, a decision-bearing variation gives a more relevant answer than cosmetic novelty; source access remains a declared condition rather than a hidden advantage.

The historical Schwartz and Martin preparation-for-future-learning study showed why assessment with a later learning resource can distinguish outcomes missed by a resource-free test in ninth-grade statistics. HCD.12 adapts that design distinction in §4.5. It does not use that study as evidence that a particular adult programme accelerates learning.

This selected line requires a task whose criteria discriminate the contribution, not an automatic claim of transfer. Reopen the comparison if prior exposure defeats the novelty claim, the variant changes an unacknowledged prerequisite, or source access and feedback cannot be separated from the claimed independent contribution.

HCD.12:12 - Relations

HCD.11 supplies prior performance evidence; equivalent qualified evidence permits direct entry. HCD.13 concerns delay and support dependence rather than novelty alone. HCD.14 can use a transfer limitation to revise an affected development assumption without waiting for every other assessment. HCD.3 owns diagnosis of a limiting human target when the cause remains decision-changing.

E.23.CAE can supply a qualified differential among recognition, access, adaptation, enactment and capability change. A recovery after a cue or restored condition has only that differential meaning; it is not proof of transfer beyond the tested case.

Target-domain Methods govern the actual response and correctness criteria. A learning-product designer owns the concrete tasks, timing and instrument; this Method supplies the transfer and new-source-learning distinctions that the design must preserve.

HCD.12:End