HCD.28:4.6 - Exercise the whole route and separate what has been established

First use qualified materials and representative enactments to expose disconnected instructions, missing inputs, impossible timing or contradictory help conditions. When a reader-work question is live, use HCD.18/.19 with the relevant audience preparation and support. A design walk-through can find a defect; it cannot supply an observation that a real learner performed the task.

For a delivered use, preserve the entry, first attempt, actual help, feedback, correction and subsequent observation needed for the intended decision. HCD.11 addresses bounded performance, HCD.12 a qualified unfamiliar or new-source question, and HCD.13 delay and support dependence. A material diagnosis, operational provider check and personal capability judgement remain separate results.

Revise the connection supported by the evidence. Correct calculation with repeated unsupported causal choice reopens model selection and its teaching; late feedback reopens supply; incompatible versions reopen content distribution. Preserve useful contributions unaffected by the diagnosis. HCD.14 supplies development revision where its evidence fits; HCD.16 supplies a changed method-claim question.

A causal claim that the product improves learning requires a suitable research contribution. Identify the unit receiving the intervention, interactions within and between groups, shared providers, comparison conditions, attrition, observation horizon and alternative explanations. A successful demonstration or a before/after gain alone does not settle that question.