# Teacher, learner, home and funder run-through · simulated

**Review date:** 29 September 2026. This is a desk simulation and file/render check, not a class or family field trial.

## Teacher: choose the sequence and run the clock

I started with the [Week 1–2 prior-code list](../weeks-01-02/README.md), then selected a school route for Day 16. With `ck` next, I prepared a joined `ck` tile, modelled `back` as /b/ /a/ /k/ in three boxes, used a Day 16 A context, and kept the B card out of the learner print. With a different school order, I used the complete B `ship/chip` contrast and B sentence cards, with no `ck` exposure. Each daily script adds to **3 + 5 + 7 + 7 + 3 = 25 minutes**. The three contexts are alternatives, not three required activities in one slot. I would observe a focus group at exit and review collected cards later; interviewing a full class in three minutes is not plausible.

## Learner: meaning and safe error correction

In a simulated Day 15 check, a learner reads `chum` as if it began with /sh/. I record the exact attempt, teach the contrast **after scoring** with `ship/chip`, and later test a different `ch` word. I do not silently read the held-out word for the learner or label the learner. The adult reads the microstory; a child can retell its problem with picture cards even if decoding is still developing. That retell is **listening comprehension**, separately recorded.

For a simulated Route A Day 20, a learner reads `peck` correctly but writes a final /k/ with one letter in the dictated word. I credit decoding, mark encoding as needing teaching, revisit the joined `ck` tile after scoring, and schedule a new word later. A second learner repairs the name capital and full stop but cannot explain the actor/action; the next lesson uses meaning tiles, not another punctuation drill. These decisions use the [teacher key](teacher-key.md), which is kept from the [student checks](student-checks.md).

## Home: no purchase and no invented family requirement

An adult reads a selected context and the original story aloud. The child reads only one controlled line using the sounds **actually taught**, then points, signs, speaks or arranges objects to show meaning. A pencil and scrap paper suffice; no app, printer, recording or family disclosure is needed. If the adult supplies a sound or whole word, that help is welcomed and logged, rather than counted as independent print reading.

## Funder and rights: inspect the evidence without overclaiming

Open [README](README.md), inspect the two route rules, one [student practice card](learner-practice.md), the separate [teacher key](teacher-key.md), [exact ACARA crosswalk](CURRICULUM-CROSSWALK.md) and [source/rights ledger](SOURCES-AND-REVIEW.md). The original A4 [PDF](print/sound-to-sentence.pdf) was rendered to an image and visually checked for readable black text, open writing areas, margins and clipping; the [selectable text/tactile route](print/TEXT-ALTERNATIVE.md) carries every prompt. The [hash manifest](MANIFEST.sha256) and [read-only verifier](verify_pack.py) protect the handoff. The actual claim supported is a reproducible, original **two-week draft** with controlled child-read text and a route choice. There is no evidence yet of student outcomes, full-year coverage, translation accuracy, state compliance or classroom uptake.

## Corrections made during review

- Separated adult-read prose from child-read controlled lines, and teacher keys from both learner packets.
- Kept Day 15 and Day 20 word items absent from prior modelling and practice, so transfer is not simple repetition.
- Gated `ck` by the school's actual sequence and supplied five full cumulative days when it is not next.
- Treated sentence marks and meaning as related but distinct checks; a neat full stop alone does not prove sentence understanding.
- Avoided equating an access response or adult-scribed rich sentence with independent phonics evidence.