The mathematics of beliefs
Among the web resources mentioned above, I find particularly
enlighting Jon Butterworth's blob on the Guardian[13].
Let us go back to the expression he used to explain
the statistical meaning of the result, and compare it
with the last paragraph of the article,
split here into three pieces (1-3):
- (0)
- ``the paper quotes a one-in-ten-thousand (0.0001)
chance that this bump is a fluke.''
- (1)
- ``My money is on the false alarm at the moment,...''
- (2)
- ``...but I would be very happy to lose it.''
- (3)
- ``And I reserve the right to change my
mind rapidly as more data come in!''
We have already seen that proposition (0)
is just a misleading misinterpretation
of p-values, about which there is little to discuss.
Instead, the last paragraph is a masterpiece of correct good reasoning
(I would almost say Good's reasoning[22]),
that deserves some comments.
Subsections
Giulio D'Agostini
2012-01-02