# English formative checks, keys and later rechecks

These checks guide the next lesson; they do not grade a learner's identity, voice or willingness to speak publicly. All learners complete an exit, while the teacher observes a rotating 2–3 focus learners and uses a later quiet check for evidence hidden by pair work. Record exact wording, access support and `independent / prompted / not yet observed`. A defensible interpretation supported by the text may differ from the examples below.

| Day | Evidence to look for | Common gap and next move |
|---:|---|---|
| 1 | The narrative positions attention and accuracy in tension; a quoted line supports the judgement. | If only “Arin was wrong,” revisit his wish for readers to notice and ask how that affects the ethical choice. |
| 2 | Compares third-person scene/dialogue with the poem's questioning first person, using words from both. | If “poem is sad” has no text support, offer two line choices and ask what each makes the speaker do. |
| 3 | “Anyone who questions…” discourages scrutiny; 8.1/1.8 come from different net forces and no finish test. | If student repeats “proof,” find the missing comparison column and reframe the sentence. |
| 4 | B puts a command before limits; C gives scope/pattern and then explicit limits. | If response lists features only, ask what a reader is likely to notice first because of the order. |
| 5 | Check A below. | Separate technical-word accuracy from claim scope and paragraph/sequence feedback. |
| 6 | D supports four-month A price ($560 vs $680); E supports ten-month B ($980 vs $1,040), with stated limits. | If “one adviser lies,” align the two time horizons and recalculate. |
| 7 | Conditional sentence names horizon and valid $60 difference at ten months. | If “always cheaper,” test months 4, 8 and 10 to find the necessary condition. |
| 8 | Colon introduces a specific list or explanation; brackets add a visible qualification; questioning language welcomes scrutiny. | If punctuation merely decorates, read sentence without it and ask what relation it needs to show. |
| 9 | Two-paragraph claim for ten months cites B $980/A $1,040, treats A fairly for shorter need, and asks for a terms/access check. | If a real purchase is urged, restore fictional scope and missing conditions. |
| 10 | Check B below. | Identify whether the barrier is reading the figure, comparing horizons, implied value or making a scoped revision. |

## Check A — Day 5 evidence caption

**Prompt:** Write 3–4 sentences for a viewer who has not seen the fictional cart table. Include what was simulated, one pattern with at least one number and unit, and what the table **cannot** show. Do not copy Text C whole.

**Secure example:** “Eight acceleration readings were invented for a 0.50 kg model cart at net forces from 1 to 4 N. The two simulated accelerations at 4 N were 7.8 and 8.1 m/s²; the readings generally rise as net force rises. This is an approximately linear association in the constructed set. No wheel finish was compared, so the table cannot support a finish claim.” Equivalent concise wording is valid. More than one number is useful but not required; the example shows how to anchor a claim.

**Developing example:** “The cart got faster because of silver wheels. It was 8.1.” Feedback: “The table gives acceleration, not speed, and no wheel-finish comparison. Which column names what changed?” Ask the learner to revise with **net force**, **acceleration**, m/s² and a limit. If the learner correctly explains orally but cannot access the print, provide the data as text/tactile pairs before judging comprehension.

**Independent recheck (later):** Supply this new invented statement: “Six fictional routes had listed durations that generally rose with distance.” Ask the learner to add a precise scope and a limitation using [Table B](../mathematics/MATERIALS.md). Secure: “In the six invented routes, duration generally increased with distance; this cannot predict a real trip without traffic and stop information.” Do not reuse the same cart caption as a recheck.

## Check B — Day 10 fresh claim and response

**Original fictional statement:** “Plan A is the smart choice for everyone because its setup charge is $240 lower than B's. All households save with A, even after ten months. There is no need to read the terms.”

**Tasks:** Identify the claim; quote one figure that is valid; find an overstatement or missing condition; infer what priority the writer emphasises; revise in 3–4 sentences for the invented ten-month household using one alternative and one question to verify.

**Key:** A setup $240 vs B $480, so “$240 lower” is valid. The value emphasised is immediate upfront cost. “Everyone” and ten-month saving are unsupported; at ten months A=$1,040 and B=$980, so B is **$60 lower** within the simplified model and fits the fictional $1,000 cap. “No need to read the terms” is a poor decision premise because real terms/access/service were not supplied. A secure revision can recommend B **within the stated ten-month price model**, note A may be cheaper for shorter periods, and ask for actual contract terms or access/service equivalence. A justified “defer a real choice pending terms” response is also valid.

**Possible feedback:** If the student catches “everyone” but not the ten-month arithmetic, have them read A(10)/B(10) from the source sheet and explain the comparison. If they get the figures but omit values, ask “What is the writer asking a reader to care about?” If they use the qualifier “in this model” but still recommend a real purchase, ask what source is missing. If grammar is rough but reasoning clear through AAC/speech, preserve the content evidence and teach editing separately.

**Independent recheck (later):** Use the new fictional X/Y models from [maths source sheet](../mathematics/MATERIALS.md). Statement: “X starts $240 lower, therefore X costs less at every month.” Key: initial X $180, Y $420, so difference $240; they tie at month 8 ($900); at month 10 X $1,080 versus Y $1,020. The universal claim fails. The writer privileges setup cost; a fair repair names a horizon, arithmetic and model limits. This is separate from Day 10's original plan figures.

## Teacher record

`date | learner code (local only) | text/task | exact evidence | independent/prompted/not yet observed | access route | next move | recheck evidence`. Keep identifiable records under school policy and do not send them to SubjectNest.

**Reuse:** Original SubjectNest keys and fictional stimuli © NeuroForgeIO Pty Ltd 2026, [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Credit source/licence and mark changes.