Read the safe-source and complete model protocol first. Every lesson uses 3 + 4 + 5 + 7 + 4 + 2 = 25 minutes, a flexible same-target route, and an honest source/date/access record. An actual claim about a mark needs an approved teacher-made practice surface and adult-controlled trial. If unavailable, write REAL OBSERVATION NOT AVAILABLE and use the model for questioning/representation only. A real school safety sign is never altered; this Science sequence does not substitute for the Week 7 safeguarding lesson. The held checks and key stay out of learner view until the first independent response.
Week 7 · Day 31 — Is a printed claim the same as evidence?
Goal: turn an invented “always stays” claim into a specific safe investigable question and identify what the poster cannot prove. Proposed codes: AC9SFH01, AC9SFI01, AC9SFU03. Prepare: Model T poster, source-claim mat, Card 31.
- Boundary · 3 min. Say PRACTICE SIGN and distinguish it from real school safety signs; no one marks or wipes a real sign.
- Read source · 4 min. Present Model T's rectangle, thick line and exact claim. Child says what is printed, not what happened.
- Model · 5 min. Teacher turns “always stays” into “What happens to this dry mark on this approved broad surface after one adult gentle dry wipe?” Put ACTUAL RESULT at NOT TESTED.
- Child try · 7 min. Card 31 routes invite a new source-specific question and one prediction. Adult labels question/prediction, not observation; material label remains UNKNOWN until verified.
- Check · 4 min. Ask whether a drawn line tells us a material's property. Child names missing surface, mark, action, date and result.
- Exit · 2 min. Save question for adult preflight; no wiping occurs by surprise.
Week 7 · Day 32 — Which method would be safe and fair?
Goal: select a possible approved dry-wipe method, separate unsafe/model options, and name controls. Proposed codes: AC9SFI01, AC9SFI02. Prepare: Model U, safe-mark routine, blank source-claim mat, Card 32.
- Preflight · 3 min. Adult confirms whether an approved broad practice surface, dry non-toxic mark and cloth may be used; if not, plan only.
- Predict · 4 min. Child predicts for a specific approved surface before any actual action; record material verification source or UNKNOWN.
- Model · 5 min. Read U1/U2/U3 fully. U1 is possible after site approval; U2 alters a real exit sign/uses spray; U3 only draws a result.
- Child try · 7 min. Card 32 asks child to arrange source, one dry mark, same adult/cloth, one gentle pass and observation. Adult sets the stable tabletop boundary and no-touch route.
- Check · 4 min. Ask what happens to comparison quality if one mark is wet and another dry, or pressure changes. Child says the condition changed; leave cause open.
- Exit · 2 min. Keep prediction and NOT TESTED fields separate for Day 33.
Week 7 · Day 33 — What happened to this one mark?
Goal: observe or attribute one actual approved dry-wipe result and compare it with a prior prediction. Proposed codes: AC9SFU03, AC9SFI02, AC9SFI03, AC9SFI04. Prepare: preflighted practice surface A and dried large mark, broad dry cloth, wipe-log, Card 33.
- Source gate · 3 min. Adult checks dry mark, surface, cloth, access, local date and approval immediately before action; stop if unsuitable.
- Before · 4 min. Child states prediction before the adult wipes and describes mark before if directly seen or attributes adult report.
- Model · 5 min. Use fictional Model W language cards to separate “I predict”, “I saw”, “the adult reported” and “the drawing shows”. W gives no local result.
- Child try · 7 min. Adult makes one gentle dry pass on A. Child directly views or hears neutral adult description; record exact after-mark (unchanged/lighter/gone/smeared/other/cannot tell), source/date/action/access. No actual try: REAL OBSERVATION NOT AVAILABLE.
- Compare · 4 min. Child compares prediction with actual only if result exists. “This exact mark looked lighter in this try” is bounded; “all paper marks erase” is unsupported.
- Exit · 2 min. Save the first actual record even if surprising, or retain NOT TESTED for a later safe try.
Week 7 · Day 34 — Is another surface the same?
Goal: compare two approved surface results cautiously and name differences besides verified material. Proposed codes: AC9SFU03, AC9SFI01, AC9SFI02, AC9SFI03. Prepare: dated A log, new broad safe practice surface B with dried mark, same adult/cloth action, Card 34.
- Match gate · 3 min. Adult checks B's identity, verified material or UNKNOWN, dry state and no stain risk; keep mark size, desk, cloth/action similar where possible.
- Predict · 4 min. Child predicts B result before adult wipe and names how B differs from A. Material may not be the only difference.
- Model · 5 min. Teacher places A's dated log beside a blank B line. A drawn model never fills B's actual result.
- Child try · 7 min. Adult safely wipes B once; child records exact observed/attributed result or NOT TESTED. Compare A/B only if both actual records exist.
- Reason · 4 min. Child names colour/coating/drying time/pressure as possible differing conditions and repairs “material caused it” to a conditional statement about these trials.
- Exit · 2 min. Mark one feature to hold the same in a future repeat; keep a separate log per source.
Week 7 · Day 35 — Fresh claim and one-trial check
Goal: independently distinguish a fresh printed claim from an unseen actual source or model and give one bounded result/unknown. Proposed codes: AC9SFU03, AC9SFI01, AC9SFI02, AC9SFI03. Prepare: unseen Day 35 check, blank claim/log mats, Card 35 only after first response.
- Welcome · 3 min. Explain that this small check chooses next teaching; offer direct view, adult report, model, private route or pass.
- Retrieve · 4 min. Revisit known Model T's “claim is not result” idea only, without showing the held wording or mark.
- Fresh source · 5 min. Present a new approved practice sample and its claim or the distinct invented fallback; record source/date/access without cueing an answer.
- First independent response · 7 min. Child asks a specific question, predicts before any real wipe, reports exact actual result or paper marks, and says whether the claim is supported here. Save first words/point/AAC/sign/prompt before feedback.
- Limit · 4 min. Child labels DIRECT / ADULT REPORT / MODEL and names one unknown. Paper-only path has REAL OBSERVATION NOT AVAILABLE.
- Close · 2 min. Use teacher key for next teaching; release Card 35 practice afterward.
Week 8 · Day 36 — A drawn repeat or a pattern in evidence?
Goal: distinguish a repeating decoration from a pattern across comparable actual Science observations. Proposed codes: AC9SFH01, AC9SFI03. Prepare: invented Model V, repeat-record mat, actual dated W7 logs if any, Card 36.
- Source gate · 3 min. Name Model V as three fictional records. Retrieve real W7 logs separately; never combine model and actual as one data set.
- Read · 4 min. Read V1/V2/V3 fully. Child says which drawn after marks look alike and which differ; no real wipe claim.
- Model · 5 min. Teacher contrasts “line, circle, line, circle” decorative AB sequence with a Science record pattern: results from repeated comparable actual tries. A repeated picture alone is not repeated observation.
- Child try · 7 min. Card 36 routes ask sorting V model results and planning three separate actual marked areas on one approved surface; source/date/action/result fields stay blank before real tries.
- Check · 4 min. Ask whether two fictional thin lines prove real marks lighten. Child says no; a real local comparable set is needed.
- Exit · 2 min. Child predicts whether three future real trials might look alike; mark prediction, not result.
Week 8 · Day 37 — Try the same safe action three times
Goal: gather up to three separate comparable actual observations and log any changed condition. Proposed codes: AC9SFU03, AC9SFI01, AC9SFI02, AC9SFI03. Prepare: one approved broad practice surface with three separate large dry marks, clean broad cloth areas, same adult, repeat-record, Card 37.
- Preflight · 3 min. Adult verifies three safe separated marked areas, material label or UNKNOWN, dry state, table/access and cleaning plan. Stop if unsuitable.
- Predict · 4 min. Child predicts likely after-mark for each or one shared prediction before any wiping; adult records timing.
- Model · 5 min. Teacher shows three blank trial rows and how to note if adult pressure or cloth area differs; Model V remains FICTIONAL.
- Child try · 7 min. Adult makes one gentle pass on each of three separate marks, one at a time, if safe. Child views or hears attributed report; record actual after-mark for Trial 1/2/3 individually, including NOT TESTED where a try stops.
- Check · 4 min. Child says whether three actual results are alike/different in these tries. If fewer than three comparable actual records, mark INSUFFICIENT COMPARABLE OBSERVATIONS.
- Exit · 2 min. Save dated unaltered rows, including unexpected outcomes.
Week 8 · Day 38 — How strong is our small pattern?
Goal: identify a cautious local pattern from three comparable actual records, or explain why available records are insufficient. Proposed codes: AC9SFI03, AC9SFI04. Prepare: actual Day 37 log if available, invented Model V for paper fallback, Card 38.
- Source check · 3 min. Count only comparable real dated Trial 1/2/3 records for actual pattern; missing or changed-condition rows are flagged, not repaired by fiction.
- Compare · 4 min. Read each actual before/action/after or V's fictional drawing records with explicit MODEL label.
- Model · 5 min. Teacher says “In these three actual tries, [teacher fills from log]” only if data exist; otherwise demonstrates “the invented drawings have two thin after-lines and one blank”.
- Child try · 7 min. Card 38 asks child to group same/different actual results with guidance, compare prediction if available, and name one changed condition. Paper route groups V marks only.
- Check · 4 min. Offer “three similar marks prove every surface cleans this way.” Child repairs to exact source/ink/action and the small sample, or says insufficient evidence.
- Exit · 2 min. Ask which one new safe observation could challenge or support the local pattern.
Week 8 · Day 39 — Share a useful evidence report
Goal: communicate a source-bounded question, prediction, result/pattern or insufficient evidence, and a next safe check. Proposed codes: AC9SFH01, AC9SFI03, AC9SFI04, AC9SFI05. Prepare: actual Week 7/8 logs or Model V labelled fiction, evidence-summary, Card 39.
- Choose source · 3 min. Use dated actual records if present; otherwise MODEL ONLY. Do not turn a drawn after-line into a real property.
- Read · 4 min. Review question, prediction timing, adult action and each actual result/NOT TESTED; mark access as DIRECT or ADULT REPORT.
- Model · 5 min. Teacher gives a model-only report on V: “Two invented cards draw thin lines; that does not tell us what happened here.”
- Child try · 7 min. Card 39 offers private speech, AAC/point, tactile order or drawing; child names source/date, what records show, one limit and next question. Adult captures exact actual communication.
- Listener · 4 min. If a real consenting listener asks something, note exact question/response; otherwise NO ACTUAL LISTENER RESPONSE. No fictional feedback.
- Exit · 2 min. Keep one new question for later, without promising a result.
Week 8 · Day 40 — Fresh repeated-result check
Goal: independently read or gather a new three-try source, describe any limited local pattern, and bound the claim. Proposed codes: AC9SFU03, AC9SFI01, AC9SFI02, AC9SFI03, AC9SFI04. Prepare: unseen Day 40 held check, blank repeat mat, Card 40 only after check.
- Welcome · 3 min. State that this is a small formative sample, not a year judgment; offer no-touch, attributed adult report, model and pass.
- Retrieve · 4 min. Revisit V's known distinction between fictional three cards and actual records, without showing held cases.
- Fresh source · 5 min. Present a new locally approved three-area surface or the separate invented record set. Log source/date/material label/access.
- First independent response · 7 min. Child predicts before real action, records each actual result or reads model marks, identifies a same/different pattern or insufficiency, and states a limit. Capture first response and prompt before teaching.
- Bound · 4 min. Child compares prediction with actual only if trials occurred, and names one changed condition or next check. Model results stay representation-only.
- Close · 2 min. Use separate key to choose follow-up. One three-try check cannot certify full Science achievement.
Original scripts © 2026 NeuroForgeIO Pty Ltd, CC BY 4.0.