# Ten small investigations · Days 151–160

Read [the actual setup and complete fictional notebook](MATERIALS.md) first. Use actual safe materials when accessible; otherwise use the labelled fictional route. Each session totals 25 minutes, with breaks and pacing adapted by the educator. Every child can choose a useful question or decline a sensory action. Evidence refers to the exact task accessed. These are Science sessions, not a requirement to teach every subject daily.

## Week 31 · Day 151 · Build a window question

**Goal:** distinguish an object from its parts and notice an accessible material property. Proposed opportunities: AC9SFU03, AC9SFH01, AC9SFI02, AC9SFI05.

**Prepare:** adult-prepared frame/panel object; puppet; two broad sample pieces; [window parts aid](print/window-parts.pdf). If actual materials are unavailable, label the whole route FICTIONAL MODEL.

1. **Meet Pip · 3 min.** “Pip wants a little window. What could we find out?” A child chooses see-through, feel, bend gently or another approved property question.
2. **Meet the object · 4 min.** Show the whole window. Point to the frame and panel separately. Name material identities from reliable information, or mark unknown; do not ask children to guess a plastic type from looks.
3. **Explore a part · 7 min.** Offer one desired safe sensory action or child-directed adult action. Ask “What did you notice about this part?” Keep the child's words. Seeing a part and an adult's material report are different sources.
4. **Make a question · 5 min.** Choose a practical question: “Can we see Pip's circle through this panel?” or “Does this card bend with a gentle press?” The adult checks suitability before trying anything.
5. **Tell a helper · 4 min.** Child indicates object, part, source and question through their chosen mode. A two-word AAC answer plus pointing can be enough.
6. **Keep and next · 2 min.** Save one property/question with access type. With no actual access, mark REAL OBSERVATION NOT AVAILABLE. Tomorrow we keep one target steady.

## Week 31 · Day 152 · Predict before peeking

**Goal:** make a prior prediction grounded in experience. Proposed opportunities: AC9SFH01, AC9SFI01, AC9SFI02, AC9SFI04.

**Prepare:** approved sample not yet viewed against the puppet; no-cover baseline; [setup](print/view-setup.pdf) and [prediction log](print/prediction-view-log.pdf). Model option: keep one supplied notebook result covered until after prediction.

1. **Choose the job · 3 min.** “Should this window show Pip's circle or hide it?” Child chooses. Check the uncovered target is available from the viewing place. If not, pause/adjust, change to an accessible property or use an attributed report; do not attribute nonvisibility to the cover without an available baseline.
2. **Predict · 4 min.** Ask “What do you think you will see through this sample, and what makes you think that?” Record exact words **before** the first view. “Not sure” is welcome.
3. **Try one view · 7 min.** Put the sample at the marked position. Child observes or receives a named adult report. In the model route reveal the specified row; do not present it as an experiment.
4. **Compare · 5 min.** Read the prior words beside the later observation. Ask “Same as you expected, different, or unclear?” Avoid turning a later guess into an earlier prediction.
5. **Use the result · 4 min.** Would this sample help the chosen show/hide job in this setup? Accept changing the choice after seeing the result.
6. **Keep and next · 2 min.** Save both records and any interruption. Tomorrow we compare samples from the same viewing place.

## Week 31 · Day 153 · Three covers, one target

**Goal:** describe observed material properties and represent a guided comparison. Proposed opportunities: AC9SFU03, AC9SFI02, AC9SFI03, AC9SFI05.

**Prepare:** approved paper/card/clear sheet; puppet; fixed viewing marks; log. Keep the local identities distinct from the fictional P/C/T IDs.

1. **Check the target · 3 min.** Child confirms the uncovered circle is available, or chooses a nonvisual property question. Keep NO / UNCLEAR as baseline data, then pause/adjust, change property or use an attributed report. Do not say a cover hid a target that was unavailable before covering it.
2. **Plan together · 4 min.** Choose which sample to try first. Keep puppet, sample place, viewer place and room lighting steady; explain one thing to hold still in child-sized words.
3. **Observe three samples · 7 min.** Adult positions each in turn. Record each actual response without replacing it with the model's expected result. Pause after one sample if needed and finish later.
4. **Make a record · 5 min.** Child places YES / NO / UNCLEAR beside each sample ID and adds a word/drawing/AAC selection describing the view. The educator attributes scribed words.
5. **Compare a pair · 4 min.** Ask “Which two views were different here?” Do not equate clearer with better for every use; hiding can be the chosen purpose.
6. **Keep and next · 2 min.** Keep raw rows and accessible property evidence. Tomorrow: one sheet and two sheets of the same paper.

## Week 31 · Day 154 · Add one more sheet

**Goal:** pose a question about a change, then compare prior prediction and actual observation. Proposed opportunities: AC9SFU03, AC9SFI01–AC9SFI04.

**Prepare:** two comparable sheets of the same paper, source recorded; identical view arrangement; [layer pair aid](print/layer-pair.pdf). Do not give the child's predicted result in advance.

1. **Recall yesterday · 3 min.** Read the child's paper row. “That was one sheet in yesterday's setup.” It does not settle today's two-sheet view.
2. **Predict the change · 4 min.** Child says what adding a second sheet might do and why. Record first; permit “I don't know.”
3. **View one and two · 7 min.** Try one sheet, then two stacked sheets, at the marked place. Record both views, including unclear results or sample movement. No compulsory touch.
4. **Compare · 5 min.** Put prediction, one-sheet result and two-sheet result side by side. Ask what changed in the arrangement and what the child actually noticed.
5. **Choose a smaller claim · 4 min.** “With these sheets, from here, I saw ___.” If the result was unchanged, report that. Do not require the fictional notebook's answer.
6. **Keep and next · 2 min.** Keep sample origin and both rows. The fresh check tomorrow uses a new supplied notebook, not today's paper.

## Week 31 · Day 155 · A new notebook

**Goal:** interpret a supplied comparison and separate recorded result from a wider guess. Proposed opportunities: AC9SFU03, AC9SFI03, AC9SFI04, AC9SFI05, limited to reasoning about the record.

**Prepare:** keep [teacher-held Case AH](ASSESSMENT.md) and [educator key](TEACHER-KEY.md) unseen until first independent response. Offer an uncoached access mode.

1. **Settle · 3 min.** “You can point, tell, sign, use AAC or direct a scribe. This is an invented notebook.” Explain that it is a fresh case.
2. **Meet the source · 4 min.** Read the case exactly, without selecting answers or adding a prediction. Give the source ID and both rows.
3. **First independent response · 7 min.** Ask the three case prompts. Record words/choices and assistance. Allow more time or another short sitting; the clock is not a Science criterion.
4. **Read it back · 5 min.** Read the response back for the child to confirm; keep first response and later change distinct. Use the key only after it has been recorded.
5. **Teaching afterwards · 4 min.** Return to an already practised P/C/T model to teach the specific next step. Do not rehearse tomorrow's or Day 160's held case.
6. **Keep and next · 2 min.** Save interpretation evidence; mark REAL OBSERVATION NOT AVAILABLE for direct investigation claims based only on this case.

## Week 32 · Day 156 · The helper has a question

**Goal:** explore how observation and a question help someone choose a material. Proposed opportunities: AC9SFH01, AC9SFI01, AC9SFI05.

**Prepare:** complete fictional notebook and helper message: “I want visitors to see Pip's circle. I can use P, C or T. What should I check?” [Share aid](print/helper-share.pdf).

1. **Give the helper a job · 3 min.** Read the message. This helper is invented; no real designer interview is claimed.
2. **Choose a useful question · 4 min.** Child asks about seeing the circle through a particular sample, rather than price, strongest material or tomorrow's weather, which the notebook cannot answer.
3. **Find a relevant row · 7 min.** Use P/C/T records. Child locates the view that helps the show-circle job and tells where the information came from.
4. **Say what is missing · 5 min.** Could we say this will work outdoors or with every kind of plastic? Identify one missing check. The notebook reports a particular view only.
5. **Make a practical message · 4 min.** “Try T for this model. The helper could see the circle's edge through T.” Choose spoken, signed, drawn or scribed delivery.
6. **Keep and next · 2 min.** Save question and record-based explanation. Tomorrow we use the actual frame/panel object if accessible.

## Week 32 · Day 157 · Pick a window or a hiding cover

**Goal:** recognise materials in an object and describe properties relevant to a child-chosen purpose. Proposed opportunities: AC9SFU03, AC9SFI02, AC9SFI03, AC9SFI05.

**Prepare:** approved adult-made frame and secured panel; puppet; alternative broad hiding cover; actual prior sample log. Adult handles assembly and sample changes.

1. **Choose the purpose · 3 min.** “Would your puppet like a peek window or a hiding cover?” No particular choice is the right personality or preferred learner route.
2. **Name the parts · 4 min.** Show frame and panel. Use known material source information or unknown; child indicates which part gives the view/cover in the setup.
3. **Try the chosen object · 7 min.** Test from the marked place. Child observes or directs an adult; an attributed report does not become direct visual evidence. Allow an actual accessible tactile property question instead.
4. **Record and change · 5 min.** Child says whether the object helps this purpose, based on the actual property encountered. Adult swaps an approved cover if requested. Record a changed result separately.
5. **Give it a story · 4 min.** Child names the paper puppet or invents a tiny scene. Keep Science property evidence separate from creative storytelling and craft success.
6. **Keep and next · 2 min.** Save part/material/property/source and the practical choice. If no actual access happened, use the full fictional route and label it.

## Week 32 · Day 158 · Check the view again

**Goal:** use another observation to check a bounded statement. Proposed opportunities: AC9SFH01, AC9SFI01–AC9SFI04.

**Prepare:** same actual sample and target, log, marked view; educator-approved repeat. Choose one question arising from earlier raw rows, not an imposed expected outcome.

1. **Read the first claim · 3 min.** “I saw ___ through this sample from here.” Ask whether a later view might match, differ or be unclear.
2. **Predict a repeat · 4 min.** Record the child's prior words. State what will be kept in the same place. A different actual light condition must be recorded, not hidden.
3. **Make the second view · 7 min.** Recheck the uncovered baseline, then repeat safely. Keep the second raw observation even if it disagrees; pause/change property or use an attributed report when access is unsuitable. No unavailable baseline is evidence that the cover hid the target.
4. **Compare records · 5 min.** Child indicates same/different/unclear and one possible setup change. A suggested explanation is a question to check, not proved cause.
5. **Revise the helper message · 4 min.** Include both dates/views when needed. “Both views showed ___” or “The views differed; we need another check” are useful outcomes.
6. **Keep and next · 2 min.** Keep prior words, two rows and interruptions. Model-only access gets record-reasoning evidence, not direct experiment evidence.

## Week 32 · Day 159 · Leave a useful note

**Goal:** share a material question, observation and uncertainty with a willing listener. Proposed opportunities: AC9SFU03, AC9SFH01, AC9SFI05.

**Prepare:** child's own record and window/cover or the labelled fictional notebook; share aid; optional willing adult/peer. No public photo or recording needed.

1. **Choose the message · 3 min.** Child picks one useful discovery or unanswered question about their window. A private explanation to the educator is enough.
2. **Choose the source · 4 min.** Point to the actual row or model row. Include a short source phrase: “In my view today…” or “In Pip's pretend notebook…”.
3. **Make the note · 7 min.** Use picture, words, sign, AAC or child-directed scribing. Include material/part, what was noticed, purpose and one unknown. No handwriting speed criterion.
4. **Try it with a listener · 5 min.** If someone is willing, ask what the note says. Keep actual words or NONE. A supplied listener example is not feedback from a person.
5. **Improve one thing · 4 min.** Child chooses a bigger circle picture, clearer sample label or smaller claim. Let the note remain private.
6. **Keep and next · 2 min.** Save first note/change and optional response. Keep tomorrow's fresh two-material object case held.

## Week 32 · Day 160 · A different little object

**Goal:** connect two material parts to recorded properties and identify a sensible next observation. Proposed opportunities: AC9SFU03, AC9SFH01, AC9SFI03, AC9SFI05, restricted to the source accessed.

**Prepare:** [teacher-held Case AI](ASSESSMENT.md); key unseen before response; accessible words or pointing without answer cues.

1. **Settle · 3 min.** “Here's another invented object. You can choose how to answer.” Name the source as a paper case.
2. **Read the case · 4 min.** Read its exact part/material table and observation. Do not add outdoor or strength facts.
3. **First independent response · 7 min.** Ask its three prompts and save response/support. Extend or split the sitting if needed; speed is not the target.
4. **Confirm · 5 min.** Child checks the scribed response. Preserve changes as later responses. Review the key afterwards.
5. **Choose what comes next · 4 min.** Teach the identified next step using the earlier window object, or ask a new safe question for another day. Do not convert this paper case into actual sensory evidence.
6. **Keep and next · 2 min.** Save bounded evidence and next teaching decision. Mark REAL OBSERVATION NOT AVAILABLE for any direct-investigation claim supported only by the case.