Predictions are not definitions.

Notes from the experiment: what was observed, what was inferred, and what would prove the thesis wrong.