# Teacher-only fresh mathematics checks, keys and next moves

Keep these exact items unseen until Days 125 and 130. They are short **formative** samples, not a claim of full AC9MFST01/AC9MFSP02 attainment. Adult reads labels if needed; record mathematical sorting/location separately from print. Use large, tactile or digital-equivalent cards, a private response route and a record of the child's first independent answer before feedback. An inaccessible setup is **not observed**, not wrong.

## Day 125 · new picture-data question

**Question to state before collection:** “Which paper-story place did these **nine imaginary storytellers** choose?” Print/cut the [teacher-only original nine-card sheet](heldout-place-cards.svg): **ART TABLE A1–A4 (4 cards), READING MAT R1–R3 (3 cards), PAPER STAGE S1–S2 (2 cards)**. Its intact layout exposes counts, so keep it away from learners and mix all nine in a plain envelope. One card means one invented chooser's recorded selection; no actual child preference is collected. Show three empty labelled group spaces, without counts. For nonvisual access, use four rough ART TABLE, three smooth READING MAT and two ridged PAPER STAGE large cards, all named aloud; preserve one card per choice.

1. Child **collects** the mixed cards, sorts by the pictured place and gives a group rule.
2. Child represents each recorded choice with one object/image card on the display; asks/answers which setting has most and which has fewest, and compares reading mat with paper stage by matching or counting.
3. Ask what the cards **cannot** tell about the real class and how the child checked a comparison.

**Exact key:** art table **4**, reading mat **3**, paper stage **2**, total **9**. Art table has most; paper stage has fewest; reading mat has **one more card** than paper stage, or more without requiring the difference number. Category rule is pictured/labelled story setting, not card colour/serial suffix. One card represents one fictional recorded choice. The cards say **nothing** about actual classmates, their reasons or whether anyone else would choose the same thing. Credit exact large-card placement, one-to-one matching, counting, AAC/sign/voice or child-directed adult movement; record any adult reading or sorting cue.

**Next move:** cards mixed by colour/suffix → model one category image feature and re-sort a *different* three-setting set; double-placed/omitted card → one-to-one check with each item moved once; count okay but “most” reversed → align cards in rows and compare matched pairs, then new counts; claim about real class → distinguish fictional source from an actual anonymous sample; correct result without reason → ask “show me the pair/group you checked” with a fresh display. If 9 cards overwhelm, reduce to a new 2-category 4–5-card diagnostic and recheck later; do not silently alter the held-out key.

## Day 130 · new paper map and reference

Give the [teacher-only original new map](heldout-route-map.svg), or a tactile two-row, three-column equivalent, with a clear PAGE TOP mark. Top row left→right: **HAT, BOOK, LEAF**. Bottom row left→right: **CUP, STAR, BOX**. These labels differ from the practice map. Orient together without showing the route. Read cell labels as needed. This is not a real room or evacuation map.

**Adult-read direction:** “Start at CUP. Move to HAT. Stop there. Then move to BOOK.” Ask child to follow each connected leg with a visitor marker or direct an adult. Ask final object, where it is relative to HAT, and where it is relative to STAR **on the agreed page orientation**. Ask whether rotating the page silently would preserve the words “right” and “above” for the same viewer.

**Exact key:** first stop **HAT**, next/final **BOOK**. BOOK is **right of HAT** and **above STAR** when PAGE TOP is as given. If the page is turned, viewer-relative left/right/above must be restated from the new agreed viewpoint; rotating without saying so makes directions ambiguous. A child may use “beside HAT” as a true but less precise relation; request the explicit right relation for this check. Preserve first endpoint and all prompt/repeat detail.

**Next move:** wrong first stop → give one-leg two-symbol route with a new layout and name start; right route but relation missing → model “BOOK is right **of HAT**,” then test another object/reference; correct endpoint via guessing → ask child to show each leg and listen to a listener follow their words; orientation confusion → label PAGE TOP tactilely and retest a different map; inability to move marker → child-directed adult placement and re-observe. A paper-only success samples object-relative language. Do **not** mark self/other-person position in a familiar real space as observed unless a safe locally approved transfer actually occurs.

**Private record:** learner code/date ___; task version ___; first sort/count/compare or route/relations ___; explanation ___; exact adult read/repeat/model ___; mode ___; revised response ___; next lesson ___; real-space transfer observed separately yes/no ___. No named classmates, personal locations or class photos are needed.

Original assessment sets and keys © 2026 NeuroForgeIO Pty Ltd, SubjectNest, [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Credit, link and indicate changes. Curriculum/rights scope in [README.md](README.md).