HCD.16:2 - Problem

Human-development Methods are often named at the wrong grain. “Coaching,” “deliberate practice,” “apprenticeship,” “retrieval,” or “AI tutoring” can hide different operations, learner conditions, supports, providers, tasks, and evidence claims. A new source, tool version, or local success is then used either to replace the whole repertoire or to preserve it unchanged.

Both reactions lose decision value. A source may contradict only one mechanism claim, one population branch, or one assisted-performance inference. A local failure may concern provider capability, access, assessment, or target-domain correctness rather than reusable Method semantics. Conversely, a reassuring average can conceal a critical error, failed transfer, weak retention, or dependence on support.

HCD.16 keeps the evidence question human-specific while returning Method identity and generic repertoire maintenance to Method Engineering. It changes only the exact HCD claim and bounded repertoire use that the evidence can defeat.