# Separate teacher key · use after the first independent response

Record the child's **exact** first words/point/AAC/sign/drawing, access route and adult prompt before coaching. A thoughtful “unknown”, “not tested” or “I need the source” can be the most accurate answer. “Pass” leads to a gentle later invitation, not an invented score. These samples guide teaching; they do not diagnose a child, certify a content description or replace wider work samples. The local object may behave differently from anyone's prediction.

## Day 15

**Source:** actual object with local date and direct access is DIRECT; an adult's exact description is ADULT REPORT; the two-arrow board is MODEL. For the model, today's use date is not an observation date. **Question:** any safe specific movement question about a named real object and flat lane is suitable. **Prediction:** stated before release; otherwise tag POST-HOC and teach sequence again rather than pretending it preceded a trial. **Actual option:** accept the object's real path/stop/stay/turn as observed, regardless of expected movement. Attributed adult description is valuable reasoning evidence but not child's own direct sensory observation. **Model option:** square at centre, one arrow left halfway, one right near catch; arrows are drawn possible routes, not proof that two objects moved. **Unknown:** material, cause, release, speed, time, local outcome all remain unshown by paper. Compare prediction SAME/DIFFERENT only with a genuine result; fallback is NOT TESTED.

**Look for:** source attribution; investigable question; prediction before action; accurate source-bounded description; one limit. **Next teaching move:** if source is confused, replay MODEL versus ACTUAL with labelled cards; if question is only a preference, practise “What happens when…?” with a specific approved source; if prediction is called result, rehearse two separate boxes; if actual movement is overclaimed, attach date/conditions and practise “in this try”. Keep a child who uses adult description in inquiry while labelling evidence correctly.

## Day 20

**Actual pair:** Only use if same verified material and roughly same shape, different verified size, and safe common lane/release are documented. No fixed result is in the key. Accept whatever each object actually does and a comparison confined to those two trials. Size is a *candidate influence*, but a single pair with possible differences in weight, finish or stiffness does not warrant “size caused it” as a general fact. Child should name shared/different conditions and at least one next control or repeat. **Fallback:** A says STOP AT CENTRE; B says STOP TO THE LEFT, but both are invented. Their ring/oval shapes differ, material is UNKNOWN and release/surface are absent; cause and local result are unknown. Do not credit an invented stop as an observed safe investigation or compare a child's prediction with it as if it were actual. A suitable later question is how two approved similarly sized/verified-material shapes move on one surface under one adult release, if school rules allow.

**Look for:** actual/model separation, A/B record reading, conditional comparison, missing-control reasoning and next check. **Next teaching move:** if “different stops” is confused with “shape caused it”, display two changed-factor cards and ask what must stay same; if material is guessed by shine, model verifying the label; if child handles a card accurately but no real trial happened, retain representation evidence and plan an accessible local observation later. A child may say “I cannot tell” with a clear reason.

**Recording grid:** date/source; actual / adult report / model; first question; prediction timing; real action and result or NOT TESTED; child's interpretation; prompts/access; follow-up. Keep child records in school-approved private systems, never in this public pack. Original key © 2026 NeuroForgeIO Pty Ltd, [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).