Let a user reveal what the project actually does, make one evidence-led revision, then explain the finished outcome.
Ākonga present what they made and, more importantly, the thinking that produced it and the limits they still see in it.
This is the learn-by-doing summative. A fresh user attempts the project without coaching. The maker logs evidence, completes one justified revision, retests and gives a two-minute explanation beside the working outcome.
For each project specification, write one action a user can perform and one observable pass result. Example: “A new user starts the quiz within 20 seconds without verbal help.”
Reset the project, hand over controls, stay silent and record actions or hesitations.
Think aloud. Attempt the purpose honestly. Say what you expect before each input.
Record one function result, one accessibility/usability result and one question.
Cosmetic changes do not outrank a broken control, unreadable instruction or unreliable result.
Use stations or a half-class gallery. Each maker gives a two-minute explanation:
Visitors leave one specific acknowledgement and one question. Do not publish projects or learner identities beyond the class without consent.
“My outcome now serves its audience better because ___ evidence led me to change ___, and the retest showed ___.”
Applying principles of form, function, accessibility, and usability during development and evaluating outcomes against specifications
The working outcome also evidences statement 51630. See the lesson-to-evidence map.
Resources already provided: project record and rubric, devices, saved projects, headphones where required and a visible timer.
Paearu Angitu | Success Criteria: The project runs; another user can operate it; the log links evidence to a revision and retest; the maker explains core code and one design decision. Use the rubric in the assessment brief, with code, log and explanation as triangulated evidence.
Entry: Permit a teacher or peer to operate while the learner points, narrates and explains decisions. On-level: Require all four evidence components. Extension: Compare two user groups or justify a trade-off between accessibility and complexity. Offer quiet showcase stations, recorded explanations, written scripts and reduced audience size for neurodiverse learners.
Protect at least 18 minutes for revision; a showcase without improvement misses the assessment spine. Pair users across different project types. Assess the evidence and reasoning, not polish or prior coding experience. Manaakitanga applies to feedback, privacy and cultural permissions.
Scratch for Educators. Constructionism makes the working artefact the site of learning. Ako, whanaungatanga and manaakitanga shape reciprocal testing and respectful public explanation.