This is a document and simulated-use review, not a classroom trial, student outcome, accessibility certification or funder endorsement. Re-run after changes; locally test with teachers and learners before relying on it in class.
| Perspective | Run through | Fault found and correction | What remains |
|---|---|---|---|
| New teacher, Day 11 | Follow 25-minute script using Text A and the source-scope sheet. | A six-row table invites use of a blank as zero. Text, visual aid and script now mark five complete pairs and a separate missing B endpoint. | A teacher still checks class pacing and syllabus fit. |
| New teacher, Day 17 | Use Text E with two reports and ask which status remains known when route is unreadable. | “Unknown” risked sounding like all fields or people are unknown. Every version now says route unreadable, while status 1 same-day/5 later remains countable. | One constructed table cannot establish a cause or teach all categorical-data ethics. |
| Learner, Day 14 | Choose one of three routes for Text D without being asked to disclose a personal moral story. | One interpretation could be marked as the author's fixed intent. Script/key now accept distinct defensible readings with two exact details and a fair alternative. | Literature breadth, including authorised First Nations voices, is not supplied by one original scene. |
| Learner, Days 15/20 | Receive unfamiliar source values and same-target route options; teacher retains first work and supported revision. | A public library cannot promise secure assessment. Both checks and guide now openly say they are publicly accessible and require a new local prompt if prior access matters. | Freshness relative to lessons is not proof of unseen access for a specific learner. |
| Accessibility | Read all three prints without colour; follow exact text/tactile equivalents, AAC, transcript and enlarged routes. | Each visual panel now has a number/word label and complete text alternative; missing-data tokens are distinct from zero. | Validate with the learner's own Braille/screen-reader workflow; no device certification is claimed. |
| Potential funder | Check whether an exact curriculum row, original rights and credible evaluation support claims of reach or impact. | Crosswalk pins 9 exact descriptions but marks selective opportunities. Ledger separates original materials from ACARA and font rights. Fictional counts and desk review are clearly marked. | No efficacy, uptake, state approval, full-year coverage or translated edition claim. |
Two actual paper replays
Learner route 17 B, simulated response: “For all 36 fictional tags, 22 are marked same day, including one of the six tags whose route is unreadable. The 70% report uses 21 of 30 tags with readable routes. These figures answer different group questions; the register cannot say that route caused the finish status.” Teacher check: the answer retains the unknown-route tags, distinguishes 22/36 from 21/30, and avoids people/causality. It would still need the learner's own response and access record in a real class.
Learner route 19 C, simulated captions read without the drawing: “Frame 1: All 36 invented tags, including six with unreadable route labels; 22 are same-day. Frame 2: Of 30 with legible drawer/web route labels, 21 are same-day. A status count does not show why a tag was finished or how a real studio works.” Teacher check: the captions carry the numbers, source scope and limit even if the drawing is unseen. This caught an early risk that a visual storyboard could rely on colour or an unspoken axis; the learner route now asks for complete text captions.
Replay receipts: python3 verify_pack.py; sha256sum -c MANIFEST.sha256; one-page A4/text/font checks for each PDF; three print visual inspections for clipping, hierarchy and contrast. Human teaching and accessibility judgement still matter.
Original resource rights: © NeuroForgeIO Pty Ltd 2026, SubjectNest, CC BY 4.0. Credit author, source, licence and changes.