Teaching-method notes and evidence boundaries

Reviewed 2026-09-22. The prior compressed note cited author names and years without exact primary-paper links, samples or outcome definitions. Its effect sizes, ideal spacing intervals, MOOC completion percentages and categorical claims about competing methods are withdrawn pending traceable primary-source review. See the claim ledger; no historical study is counted as recent interview evidence.

The current teaching standard is explicit and testable: begin with a specific user operation and expected values; explain a repeatable approach; show how the representation changes under a new requirement; provide a runnable starting point where code exists; separate candidate and assessor material; and assess an unfamiliar variation. A learner who can recite the supplied answer but cannot transfer it has not demonstrated the intended skill.

Worked examples, delayed reattempts and gradually reduced hints are curriculum choices to evaluate through independent attempts. We do not promise a quantified learning gain, an ungameable test, a universal optimum spacing interval or a hiring outcome. Keep teaching results and research claims in separate records.