Goal: compare a genuinely prior prediction with later actual observation and separate visible difference from inferred growth. Proposed codes: AC9SFI01, AC9SFI03, AC9SFI04. Prepare: actual Day 64 prediction plus Days 62/66 records if complete, Notebook G, Card 67.
- Pair gate · 3 min. Confirm same actual plant/feature/view and chronological record. If any missing, say CANNOT COMPARE; G is FICTIONAL.
- Read · 4 min. Put actual dated observations beside the exact earlier prediction. A predicted change may not appear; that is useful evidence.
- Model · 5 min. Teacher compares G's two story labels only, then shows why changed FRONT/SIDE makes a biological change claim unsupported.
- Child try · 7 min. Card 67 routes choose SAME / DIFFERENT / UNCLEAR / CANNOT COMPARE, cite source/date/feature and say whether prediction matched actual later view if available.
- Check · 4 min. “My drawing got bigger, so the plant grew” is corrected to a scale/view unknown unless matched actual observations support a bounded visible difference.
- Exit · 2 min. Keep one cautious sentence or reason for insufficient evidence.