Goal: compare a genuinely earlier prediction with actual observation if both are recorded, or state why comparison is not available. Proposed codes: AC9SFI01, AC9SFI04. Prepare: one actual dated prior prediction/result pair if present, Notebook N for model-only reading, Card 46.
- Source gate · 3 min. Confirm prediction timestamp/order and actual action/result. If prediction was not recorded before action, mark NOT AVAILABLE rather than fabricate it.
- Read · 4 min. Adult reads exact earlier words and actual result or model N with FICTIONAL label.
- Model · 5 min. Teacher compares N's prediction/result inside the story, then says it is not evidence that a classroom child observed movement.
- Child try · 7 min. Card 46 routes mark SAME / DIFFERENT / CANNOT COMPARE for an actual pair, with reason and source/date. A changed idea is recorded as today's idea, not an invented historical prediction.
- Check · 4 min. Ask “Was a different prediction wrong?” Child explains predictions are checked against results, not judged as effort or intelligence.
- Exit · 2 min. Write one question the actual record still leaves open.