# Weeks 5–6 formative check: count, compare, split

**Original SubjectNest draft · educator and accessibility review pending.** © NeuroForgeIO Pty Ltd 2026, [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). This checks selected evidence for AC9MFN01, AC9MFN02, AC9MFN03 and AC9MFN04, not an entire Foundation achievement standard, diagnosis or report grade. Exact curriculum source, attribution and separate ACARA terms are in [README.md](README.md).

## Run it over several days

Take close observations during the individual tries: A from Days 21–22, B from Days 23 and 25, C from Day 24, D from Days 26–30. Use a **fresh** arrangement for any unclear or partner-supported answer. Several children can work on familiar stations while the teacher checks two or three individually. Do not administer all 22 points in a single Day 30 block. Mark unavailable evidence `NA`, keep the assessed denominator, and schedule a later check. A response after modelling helps plan instruction but is not scored independent. Stop and return later if the child is tired or distressed.

Provide actual objects, clear mat boundaries, large or tactile whole numerals, spoken instructions and a response route the learner can use. The child may speak, sign, use AAC, point, draw or direct an adult's movements. Note the access format and whether it changes the evidence: a tactile or described small group may show quantity recognition but does not directly establish *visual* subitising. English pronunciation, speed and pencil control are not the score. Keep identifiable evidence under school policy, never in a public product.

## Items and scoring

| Part | Fresh teacher prompt | Independent score |
| --- | --- | ---: |
| **A. Count a loose set** | Scatter 17 large objects; later scatter 20 differently. For each: “How many? Show how you know.” | For each, 1 for one count action per object and 1 for the final whole total: **4 possible**. |
| **B. Compare** | Make 14 vs 18, 16 vs 16 in different layouts, and spread-out 19 vs close-together 20. “More, fewer or same? How do you know?” | Each pair: 1 correct relation and 1 count/pairing/path-order/leftover reason: **6 possible**. |
| **C. Numeral and order** | Show a checked set of 14, then 20; ask the child to select each whole numeral from mixed cards. Ask to order `0, 8, 14, 20` in mixed presentation, and fill `18, [empty], 20`. | 1 each for the 14/20 card matches, 1 for full four-card order, 1 for missing card: **4 possible**. |
| **D. Parts and whole** | Give a fresh set of 5, 7, then 10. For each: “Split into two parts. Show each part. Put them together. How many altogether?” Finally give 8 and request two *different* splits, recombining after each. | For 5/7/10, 1 for two valid disjoint parts accounting for all objects and 1 for the correct recombined whole. For 8, 1 for two distinct valid splits and 1 for confirming both recombine to eight: **8 possible**. |

**Maximum: 22 points if all items are independently accessible and assessed.** Record `A /4, B /6, C /4, D /8` and `total earned / total assessed`. There is no validated cutoff. Keep domain notes and an example of the child's reasoning; a single overall number cannot tell the teacher what to teach tomorrow. AC9MFN02 appears as a subitising strategy in lessons; this short check does **not** separately score visual subitising. Use the [Weeks 1–2 assessment](../weeks-01-02/assessment.md) and fresh small-group examples if that evidence is still needed.

### Answer key

| Item | Correct mathematics |
| --- | --- |
| A | First set 17; second set 20. Every object is counted once, and the final count is the cardinal total. |
| B | 18 > 14 by four; 16 = 16; 19 < 20 by one, regardless of the sets' spacing. Reversed wording is valid when the referenced set is clear. |
| C | Fourteen objects match `14`; twenty match `20`; order `0, 8, 14, 20`; missing `19`. |
| D | Five: e.g. 2 and 3; seven: e.g. 4 and 3; ten: e.g. 5 and 5; eight: e.g. 5 and 3, then 4 and 4. Any two non-overlapping parts that use **all** the given objects and total the whole are valid. Zero is a valid empty part. For eight, the two splits must use different combinations of part sizes; swapping 5/3 to 3/5 alone is the same combination. |

## Copyable private record

```text
Learner code: ______  Dates: ______  Access format/language: __________
Independent response? ______  Cues before answer? ______
A __/4 or NA: ____  B __/6 or NA: ____  C __/4 or NA: ____
D __/8 or NA: ____  TOTAL EARNED / ASSESSED: ____ / ____
Exact action or words: _____________________________________________
Last secure number range and split: _______________________________
Next teaching move: __________________  Fresh recheck date: ______
```

## Respond with one small next step

| Observed evidence | Teach next | Fresh check |
| --- | --- | --- |
| Skips or double-counts in a loose 17/20 set | Count by moving each object once to a second mat; return to the last secure range, then build up. | New loose set of 12, then 17 when ready. |
| Correct counts but wrong `14` or `20` card | Match whole numeral, spoken/verified language name and object set; hide one and rebuild. | Fresh 14/20 card selection beside new sets. |
| Size or spacing determines comparison | Pair objects across mats and identify leftovers; rearrange without changing members. | Fresh 14/18 and same-count 16/16. |
| A part is treated as the whole or an object appears in both parts | Start with five, empty the whole space as objects move into two bounded parts, then reunite the *same* objects. | Split five as 2/3 and 1/4, then seven. |
| Can show one split but not a different split of eight | Move exactly one object from one part to the other while the whole stays visible; recount parts and whole. | Eight as 5/3, then 4/4 without a model. |
| All sampled domains are secure in fresh contexts | Revisit across later weeks and move into the next planned Foundation number and other-strand lessons; offer a school-approved accessible virtual representation when available. | New count/compare/split scene after several days. |

Record useful teaching evidence, not a permanent ability label. [AERO monitoring guidance](https://www.edresearch.edu.au/guides-resources/practice-guides/monitor-progress) informs the observe–teach–recheck cycle; this local rubric itself has not been standardised or classroom piloted.