This is an actual read-through of the authored instructions, source cards, keys, algorithm branches and rendered aids. It is not a classroom, family or funder field trial.
Teacher route
For Week 3 Day 12 I gathered clean scrap, one blank story card and a stable mat. I read the three invented material arrangements, asked for a prediction, and treated the physical trial as a new observation whose outcome could not be prefilled. The phases add to 4 + 6 + 12 + 10 + 3 = 35 minutes. A teacher should choose only two arrangements for the 12-minute trial, then collect remaining logs later. Day 15 uses the prior prototype so the 15-minute assignment period is plausible; it does not require a new build from nothing.
For Week 4 I traced the algorithm through three inputs: STORY gives the STORY output; NUMBER gives the NUMBER output; not clear returns to choice after a cue change, while clear ends. A paper trace is an honest alternative when no approved device exists. I checked each script's device language: actual system/components, creation, locating and guided digital sharing are recorded only after actual approved tool actions. No account or upload is needed.
Learner route
A simulated Week 3 learner says “Mina must have trouble seeing” after reading the brief. The correct next move is to return to Mina's exact request—three ordered visible cards—and label vision as unknown. If a stand tips during a real test, that is valuable evidence; the learner names the observed tip and revises one feature. Teacher key does not contain a fixed physical-test outcome.
A simulated Week 4 learner traces STORY correctly but sends NUMBER to the same output. The teacher uses two tokens to trace both choices; the learner moves the NUMBER output card and retests. If the learner builds a paper screen, algorithm reasoning can be observed, while digital-tool operation remains unobserved. The learner later chooses a real name for a mock slide; the teacher revisits the no-personal-data rule and asks what original symbol could communicate the station instead.
Home route
A caregiver can use paper rectangles and a pencil. Week 3: child directs folds/placements, gives a prediction, and marks what a gentle home test actually showed, with adult operation if appropriate. Week 4: child moves a token through two choices and a retry path; no device, account, photo, purchase or private family story. A home paper route does not become evidence of digital-tool use. Rich adult-read words are not called independent decodable English.
Funder route and limits
Open the two week guides and Week 4 guide, inspect the learner cards, separate teacher key, exact ACARA rows, source/rights ledger and manifest. Both original printable PDFs were rendered to page images and visually inspected: readable black text, generous blank drawing areas, intact margins, no clipping; the choice-path label initially intersected a connector and was moved inside its panel. The text/tactile route expresses both pages. The supported claim is a reproducible, original two-week draft with two larger assignments and explicit evidence limits. Student outcomes, classroom uptake, complete band coverage, translation, app privacy compliance and local state alignment are not demonstrated.
Corrections and next tests
- Kept fictional quote, learner prediction and observed class trial separate in every relevant step.
- Made algorithm branch/return paths traceable and tested them with all named input cases; paper and digital evidence are recorded separately.
- Ensured the Week 3 transfer changes a real constraint (two cards, no clip) and the Week 4 transfer changes mode (no screen/sound), rather than repeating the taught product unchanged.
- Retained the Years 1–2 band and later food/fibre, food-specialisation, account and longer-term design work as open scope. A local educator must check the school sequence, tool policy and users before live use.