Instructor Guide and Review Status

Status

This is an authored candidate with executable behavioral calibration. Independent technical/design review, accessibility review and learner pilots remain pending. Neither tests nor page count establish a professional percentile outcome. See authoring/ood-foundations-review-2026-09-25.md in the source repository for scope, corrections, verification and remaining work.

Review process

Preserve P and A first attempts before opening keys. Score each gate separately. Ask the learner to defend a changed case and a decision to keep existing code. Check public contracts and failure behavior before naming a pattern. Accept composition, functions, values or classes where the evidence supports them.

Run the extracted kit in a fresh directory. The reference must pass and the untouched starter must fail. Check deliberately faulty implementations for aliasing, state changes on rejected withdrawal, overly strong preconditions, wrong currency, swallowed provider errors and unsafe output. Inspect the claimed representation difference manually; a public behavior checker cannot enforce it.

Design review is not reducible to runtime assertions. Record ownership rationale, alternatives, cost assumptions and what tests leave unresolved. Mark assisted, self-assessed and independently reviewed work distinctly.

Pilot plan

Observe learners with differing language backgrounds. Record time per phase, confusing snippets, hint use, false passes, reviewer disagreement and delayed transfer. Have another reviewer solve the quiz independently and challenge the solution criteria. Revise workload from observations; do not mark pending reviews complete or silently waive the semester accessibility strand.