29 September 2026 · document and print simulation, no real classroom trial. I followed the start guide, opening daily map, later pack guides, learner packets, scripts, answer keys and original printed-page exports as a new teacher, a learner using text, and a potential funder checking claims. This establishes internal usability and arithmetic; it does not establish learning outcomes for real students.
| View | Action taken | Finding and response |
|---|---|
| New teacher, Day 1 preparation | Followed links from the start guide; laid out learner reading/data, separate teacher keys and the A4 graph; counted the three core blocks. | 25 minutes English + 25 mathematics + 35 integrated = 85 minutes. Each has six numbered phases and accessible alternatives. This is a useful set of core periods, not a full school day. Economics Week 2 needs local subject selection. |
| New teacher, Day 4 English | Tried to hand pairs the “four sentence strips” named in the script. | The original draft named a card without supplying it. STUDENT-TASKS.md now has four exact scrambled route sentences; the teacher master gives a reasoned ordering. |
| New teacher, Day 5 and Day 10 assessment | Compared learner handouts with teacher readings, answers and later rechecks. | The original English reading master mixed literary interpretation with student texts. A separate learner reading set and learner check cards now keep prompt and key apart. All three Day 5/10 blocks gather work before feedback. Review of a full class's work occurs after the session; the daily 2–3 minute close is not enough for individual conferences with everyone. |
| Learner, graph route | Read the eight cart pairs, then plotted or described them using the printed frame and text alternative. | The first draft teacher direction said 0–9 m/s² while the page and student card used 0–10. All now use 0–10. Repeated x-values remain distinct; text coordinates and tactile axes allow the same association explanation without requiring tiny visual placement. |
| Learner, media claim | Read the actual fictional post, identified 8.1 m/s² at 4 N and 1.8 m/s² at 1 N, then checked the mathematical and scientific claims. | The first draft Day 4 mathematics prompt paraphrased the headline incorrectly. It now quotes the source's “four times faster” wording. 8.1 ÷ 1.8 = 4.5, the data measure acceleration, force changes, and no wheel finish is compared. A robust student response needs those distinctions. |
| Learner, device decision | Recalculated A/B at months 4, 8 and 10; solved the whole-month cap; then attempted new X/Y Day 10 data without the teacher key. | A/B tie at 8 months, $880. At 10 months A is $1,040 and B $980; only B fits the fictional $1,000 cap. New X/Y tie at 8 months, $900, but at 10 months X $1,080 and Y $1,020, so neither fits. Price does not verify service, access or real terms. |
| Learner, alternative access | Followed the graph and cost board through selectable text without opening a PDF. | Exact coordinates, units, table row labels and claim-check prompts are available in text. The student text alternative no longer reveals the crossing or an association model answer. The teacher should preserve access support in the evidence note. |
| Teacher, transfer beyond one context | Worked the eight new learner cards against their separate key. | Library, audio, path and seed-tray cards test claim scope and association; poster, music, exhibition and event hire test model crossing and assumptions. These are optional later practice; they are not silently squeezed into the scripted 25 minutes. |
| Print reviewer | Rendered all three generated PDFs to images and inspected page layout; checked labels, blank writing space and copyright line. | Three single-page A4 black-and-white aids are readable. Two SVG formulas initially ran together because SVG collapses spaces; the cost board now positions A and B separately. The claim-check boxes now have two writing lines. Selectable text is supplied because the PDFs are visual print exports. |
| Funder or public reader | Read the source crosswalk, rights ledger, manifest and year map. | Original fiction/data/graphics and CC BY 4.0 reuse are explicit, with separate ACARA and DejaVu rights. The exact pinned ACARA code check and file hashes are reproducible. Weeks 1–4 are scripted; Weeks 5–40 are plan-only, no state overlay or verified translation is released, and no real pupil outcome or teacher trial is claimed. |
New Weeks 3–4 desk check
- Teacher: From the English, mathematics and integrated guides, the same-numbered Day 11–20 scripts make three separate 25/25/35-minute blocks. The five-area menu is optional, not a missing part of that clock. The public worked keys are teacher-targeted but accessible to anyone; a school must use a changed case for consequential independent evidence when prior access matters.
- Learner, mathematics: I independently recalculated Check A's invented proposed line
L(4)=13against listed 17, giving a +4 listed-minus-model residual with no known cause. For Check B, Ocean8+4=12, Space9+9=18, No-theme2+3=5, and the grand total is 35;8/12≈66.7%is Ocean-only, while19/35≈54.3%is all foil designs. Dropping the No-theme row changes the question to17/30. - Learner, English: In Check A only four R/S instruction pairs are complete: R totals 16 panels, S totals 12, means 4 and 3, with one pair tied and a fifth pair missing S. A one-panel mean difference says nothing about build speed or every reader. In Check B, named trays give
30/40=75%while all invented cards give32/50=64%; the unreadable-tray category has known statuses and must not disappear. I checked that the prompt and answer pages remain separate even though both are public. - Reviewer: The later pack crosswalks pin actual ACARA rows and scope, their original A4 assets have text/tactile alternatives, and the agent desk reports identify what was only planned rather than performed. This is still no classroom, assistive-technology, local-syllabus or learning-impact validation.
Findings that still need real people
- Have an Australian Year 10 teacher trial the timing, source difficulty, assessment decisions and local syllabus fit with a class. Ask learners using text, speech, AAC, tactile/large-print and quiet routes whether the prompts genuinely preserve agency and comprehension.
- Ask a qualified science educator to review any proposed real cart activity and its risk/access plan before adding apparatus. The current sequence uses paper, simulation and design critique only.
- Extend beyond four weeks with fully written learning and assessment in the remaining curriculum areas. Add Australian literature and authorised First Nations texts through appropriate local relationships, rights and cultural review. Commission qualified bilingual review before issuing a translation.
Run python3 verify_pack.py inside this folder after any edit. It checks code level and exact source text against the pinned import, complete day counts and minutes, arithmetic, local links, student/teacher separation, A4 SVG/PDF structure, font provenance and the hash manifest. This is an internal integrity check, not evidence of classroom effectiveness.
Reuse: Original SubjectNest review notes © NeuroForgeIO Pty Ltd 2026, CC BY 4.0. Credit source/licence and mark changes.