# Desk run-through and critique · Year 6 English Weeks 3–4

**Status:** simulated internal review of files on 29 September 2026. One person read/used the pack in teacher, learner and potential funder roles. This was **not** a classroom pilot, child usability study, impact evaluation or independent curriculum review.

| Role and task simulated | What worked on the desk | Issue noticed and action |
| --- | --- | --- |
| Teacher, prepare Day 11 | Source A poster, Source B table, lesson timing and D11 choices are linked from the start page. Source-origin question makes it possible to separate poster/tally emphasis from independent evidence. | A new teacher might say “two sources confirmed 11.” The packet and teacher key now explicitly say A/B/D share the same fictional event. Teacher models the distinction before choices. |
| Learner, inspect Day 12 poster without seeing it | The full alternative gives exact words and order, including large **ELEVEN** and the small lower caveat. A strip/tactile route can reproduce prominence. | Small caveat would otherwise be easy to miss. The alternative states it fully and the print poster labels “Read the closing detail too”. Still needs real screen-reader/low-vision review. |
| Learner, choose D13-B or D14-B | Garden and sports mini-cases ask for source-bounded claims instead of a repeat of Dockside. A learner can use speaking, text, cards or AAC to express the same inference. | Weatherproof and unlogged-return language risk invented outcomes. Source packet/key now require indoor one-day and not-yet-logged boundaries. |
| Teacher, Day 15 check | The unseen Borrow Box packet has a different domain and numbers from preceding models. Four items separate context, design effect, supported claim and next question. Staff key is outside learner file. | D15 practice could coach the result. Choices and lesson mark those routes **later**; the ten-minute check replaces practice/exit. Accessibility support and content hints are logged separately. |
| Learner, build Days 16–19 argument | Blank four-stage planner gives a decision reader, source evidence, limit and next step without pre-filling a response. D16–19 offer multiple domains and a private pitch option. | A cue card alone can be mistaken for delivered speaking evidence. Day 19 and staff key explicitly distinguish planning from delivered presentation. No accent, loudness or public confidence score. |
| Teacher, Day 18 grammar | The opening `Although ... , ...` example and restrictive `that` clause are grammatically distinct; `records` is a narrower verb than `proves` for a tally. | ACARA rows cover more than these examples. Crosswalk names the taught slice and keeps list/dialogue commas, full tense/adverb range and more clause work for later lessons. |
| Learner, Day 20 check | Fictional reading-cart sources are unseen before check day and have new figures (`19 + 11 = 30`). The response asks for one X fact, one attributed Y view, a limited recommendation and an unknown/next step. | The source and four-sentence prompt are substantial for ten minutes. Spoken/AAC/dictated response is allowed and first response is saved. A school may split the check for access/time, but must record the change before treating it as comparable evidence. |
| Potential funder, assess reuse and honesty | Original source texts, three vector/print pairs, clean learner material, separate keys, exact source rows, rights ledger, alt/tactile routes and hash manifest are present. | There is no evidence of classroom effect, uptake, state approval, formal accessibility testing or independently checked learner outcomes. The README and source ledger state this openly; no ROI or impact figure is offered. |

## Arithmetic, language and print desk audit

- Closing counts checked: Dockside `11 + 4 + 3 = 18`; zines `18 + 12 = 30`; sports `16 + 4 = 20`; labels `32 + 8 = 40`; fresh checks `9 + 3 = 12` and `19 + 11 = 30`. Counts are **fictional**, not empirical findings.
- Day 12's ten-word example in the staff key counts as: Eighteen(1) of2 thirty3 zines4 were5 taken6 at7 the8 trial9 shelf10.
- Day 18: a comma follows the fronted dependent clause in “Although eleven items worked at closing, the log does not show whether they lasted.” The identifying `that` clause in “The four discs that had no logged return at closing need checking” has no enclosing comma. These are examples of the relevant rule, not exhaustive grammar instruction.
- Three print PDF pages checked with `pdfinfo`, selectable text with `pdftotext`, font subsets with `pdffonts`, and raster render at 1000px inspected for clipping/contrast. The lower caveat is deliberately smaller to make design analysis possible, and the full text route supplies it accessibly.
- Local links, exact ACARA workbook source rows, lesson/choice/check counts, PDF/SVG attributes and pack hashes are checked by [the read-only verifier](verify_pack.py). It does not replace professional editorial or accessibility review.

**Next iteration after field feedback:** revise any confusing phrasing or time pressure with teacher/learner evidence, add a verified jurisdiction adaptation layer, and extend later weeks/other genres. Preserve the clean fresh-check boundary when revising examples.