A hormone was applied to cultured cells with or without a receptor antagonist, and a downstream response was measured. Prediction: The hormone response is mediated by the receptor blocked by the antagonist. Results: mean response units +/- SD, n=6: vehicle = 9 +/- 3; hormone = 88 +/- 9; antagonist alone = 8 +/- 3; hormone + antagonist = 14 +/- 4. How should the prediction be evaluated from the observed pattern and variation?