# Teacher worked answers, feedback and next moves (public resource)

This teacher-targeted key is publicly accessible in the library. Use the checks as formative prompts and change a case locally if prior access matters. Record **independent / prompted / not yet observed** for each response, plus the exact source detail, access route and later revision. The rubric's 0/1/2 descriptions are a planning aid, not a pass/fail score or Year 8 achievement decision. A learner can analyse an exact read-aloud source without demonstrating independent decoding. Mark an AAC/text-to-speech presentation as that route; do not infer vocal control from a synthetic voice. A calculator can support arithmetic but cannot create a citation, qualify a promise or choose a literary interpretation. Keep all learner evidence in school-approved private storage.

## Daily feedback map

| Day | Evidence to recognise | If missing, teach next |
| ---: | --- | --- |
| 11 | `12 + 18n + 12` and `24 + 18n` both include 24 cm of paper margins and the same `18n` panel widths; the drawn rail is not a measured real wall. | Make 12/18n/12 cards, combine only the margins and read the model-scope sentence aloud. |
| 12 | Caption 1 leads with an inviting universal and hides conditions in a small line; Caption 2 begins with scoped model evidence and gives physical limits. | Put HEADLINE, SOURCE and LIMIT on separate strips; ask how a new visitor reads them in order. |
| 13 | Four panels give `24 + 18 × 4 = 96 cm` **in Text A's drawing**; an embedded margin clause adds needed information; clips/gaps remain untested. | Reopen the `n=4` row, then ask which *which* clause adds evidence and which line adds a limit. |
| 14 | Literal margin is 12 cm each end; “somewhere to arrive” and the line as “invitation” are layered readings. Ren's reader question and Theo's rereading support a shift. | Sort measurement words from story images, then ask which event changed Theo's view. |
| 15 | See fresh Check A below. First response must use its source, not reused Text A numbers. | Select A1–A4 next move; re-sample on another neutral source if needed. |
| 16 | `5 + 15d ≤ 50` allows at most three complete demos in the invented timing; it does not prove everyone gets a turn or account for questions/changeover. | Separate *demonstration count* from *participant count* on two cards; ask which is actually in the source. |
| 17 | *Allocation* hides an actor; the coordinator note plans it but names no allocator, and Kit's reading is one perspective, not a population result. | Ask “Who is named? What tense tells us the action is future?” Repair without inventing a role. |
| 18 | Dash suspends the first thought; semicolon links two complete ideas; gate/sky images and personified ink/ruler can shape tone. A different reading is valid with words from the poem. | Read two adjacent lines aloud/quietly, compare literal with nonliteral, then revise one punctuation mark and state its effect. |
| 19 | Fair reply retains three-demo timing, challenges “everyone”, cites the unmeasured condition and asks one checkable question. Actual delivery mode and listener response are documented. | Give four sequence cards: acknowledge, source, limit, request; listen for which step a reader missed. |
| 20 | See fresh Check B below. The planned distribution is not an observed finished event. | Select B1–B4 next move and a fresh context for transfer. |

## Check A Day 15 worked key

**Source truth:** `10 + 16c + 10` combines the left/right 10 cm margins into `20 + 16c`; the 16 cm card width is unchanged. For `c=4`, `20 + 64 = 84 cm`, exactly the **drawn** line. For `c=5`, `20 + 80 = 100 cm`, above 84. The model leaves physical wall dimensions, clips, gaps, mounting and safety untested. Thus “every wall” and “proves the real installation” outrun the source. A useful next check is measuring an actual permitted site and components under school process; a paper-only next step could explicitly add a proposed clip/gap allowance and retest the drawing, still marked MODEL.

**Worked visitor caption:** “This paper plan puts 16 cm cards between two 10 cm margins. The design note shows that four cards total 84 cm, the length of its drawn line. That result does not prove the arrangement fits every real wall or that clips and gaps are safe. The team would need approved real measurements and a tested mounting method before making an installation claim.” Equivalent source-bound prose can earn the same evidence.

| Criterion | 0 · absent/incorrect | 1 · developing | 2 · secure in this sample | Immediate next move for 0/1 |
| --- | --- | --- | --- | --- |
| A1 · read model/equivalence | No accurate dimensions or treats combined margins as extra panels. | Gives 84 cm or two forms without explaining the combined margins. | Names `10+10=20` cm of margins, unchanged `16c`, and valid `c=4` result **for the drawing**. | Move left/card/right cards and say each part before computing; check arithmetic separately. |
| A2 · source reach | Accepts “every wall” or “real installation proved”. | Spots one overclaim but does not say why the paper result cannot prove it. | Cites drawn line/model versus unmeasured physical mounting or wall as reason for the boundary. | Put DRAWN and REAL labels side by side; ask which one was measured in the source. |
| A3 · reader caption | Fragment/unsupported slogan without a source. | Some evidence and caveat but unclear order, quotation or referent. | Three/four coherent sentences giving model, exact result and limitation for a new reader. | Order source → worked result → “however” limit; reader-test the referent of “it”. |
| A4 · next check | No check or asserts a real result without evidence. | “Test it” with no object or criterion. | Names permitted wall/clip/gap measurement or a clearly labelled new paper-model allowance and what to compare. | Ask “What precisely was missing from the note?” then plan one observable next step. |

**Feedback decision:** A1 absent → read the representation with a mathematics colleague if necessary; avoid diagnosing English from a calculation slip. A1 secure but A2 partial → model *source versus claim reach*. A2 secure but A3 partial → paragraph cohesion. A4 partial → write one inspectable next question. Keep the first response and recheck with different numbers later.

## Check B Day 20 worked key

**Source truth:** `7 + 11b ≤ 40`, with `b` a non-negative whole number of complete paper-copy runs. Three use `7 + 33 = 40` minutes; four use `7 + 44 = 51`, above the slot. This only bounds **model runs**. It cannot establish a copy for every visitor, queue size, actual machine performance or access to instructions. “Distribution will be arranged” names the planned action but **not who** will arrange it; future tense does not mean it has happened. A truthful sentence is “The notice says distribution is planned but does not name the person responsible.” Do not invent a coordinator. A next check asks how many people are expected, whether one run yields one copy, who is responsible, and how time and access routes would be tested; one specific question is sufficient.

**Worked reader note:** “In this invented 40-minute session, the timing model allows three complete paper-copy runs after seven minutes of setup. It does not show that every visitor will receive a copy, because the number of visitors and machine results are unknown. The notice says distribution will be arranged but does not name the person who will do it. Before promising copies, the team should check how many readers may arrive and how many usable copies one run produces.” Other well-supported questions and clear structures are valid.

| Criterion | 0 · absent/incorrect | 1 · developing | 2 · secure in this sample | Immediate next move for 0/1 |
| --- | --- | --- | --- | --- |
| B1 · valid source result | No correct time result or confuses minutes with people. | Says “three” but not the setup/runs/40-minute condition. | Cites `7+11×3=40` minutes (or equivalent words), and four needing 51 as optional contrast. | Put SETUP, RUN and TOTAL cards in order; check arithmetic with a calculator if needed. |
| B2 · unsupported promise | Treats “everyone” as proved by three runs. | Recognises headline overreach but gives only vague “more information”. | Explains why run capacity cannot establish people served or actual performance, citing the source limit. | Ask “Three of what? Everyone how many?” before changing a headline. |
| B3 · hidden actor/tense | Names an invented allocator or treats distribution as completed. | Identifies missing person but not the planned/future status. | Says *distribution* hides who arranges it, and “will be” marks a plan, not finished action. | Replace noun with verb but leave actor unknown: “Someone will arrange…” and ask what source could name them. |
| B4 · reader note and next check | No coherent note or no source. | Some model detail but no linked limit, actor uncertainty or specific question. | Three/four ordered sentences preserve valid result, narrow the promise, identify missing actor and ask a checkable question. | Order model → limit → responsibility → question, then have another reader paraphrase it. |

**Feedback decision:** Treat B1 arithmetic support separately from English analysis. If B2 is partial, teach source reach. If B3 is partial, contrast *will be arranged* with *was arranged* and demand a named source before assigning responsibility. If B4 is partial, rebuild a reader route rather than merely asking for more words. No sum of eight criteria is a complete Year 8 standard.

## Optional practice quick key

| Day | Expected answer boundary |
| ---: | --- |
| 11 | `3+2n+3 = 6+2n`; `4+10n+4 = 8+10n` cm. Joining margins changes form, not the model width. |
| 12 | “Every team wins” is unsupported by four improved attempts of six; Caption 1 can invite without universal fit. |
| 13 | For `n=2`, Text A totals 60 cm; an embedded clause should name two 12 cm margins, not a new measured wall. |
| 14 | Theo tries the plan as an unfamiliar visitor after Ren's question; “frame” can be a literal edge and a reader-guiding image. |
| 15 | “Perfectly” needs the drawn-model condition; extras refer only to earlier texts after the held-out response is collected. |
| 16 | `5 + 10r ≤ 45` allows at most four complete fictional rounds; neither rule counts players. |
| 17 | “Completion was announced” leaves builder/announcer unclear and does not itself establish a usable display; Kit is one reader. |
| 18 | A changed gate/sky image changes tone; Text F's colon leads to an explanation of “not nothing”. Different effects can be justified. |
| 19 | Retain three-demo model timing, ask how participation and question time would be checked. |
| 20 | After collection, Text D's model omits question time; Text F's semicolon links two complete line/page thoughts. |

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