Year 12 · Student investigation
When is an early confident answer premature?
Keep an individual record even when you investigate in pairs. Record the exact settings, source and data split.
Will more samples fix a decision made before the decisive late clue arrives?
What did you change? What stayed the same? What did you notice?
Use the late-reversal case with threshold 0.75. Compare additional rollouts with delaying the exit. Compare seeds 11, 23 and 47 with the same settings. Report every result and the range. Lock your final settings before opening seed 89; changing settings afterwards turns that check into development data.
A cost–accuracy analysis separating sampling noise, late evidence and correctness of the final answer.
This activity can show _ . It cannot tell us _ . A new case I would test is ____ .
Use fictional data. Download or print your work before leaving.