``The essential problem of the experimental method''

Human minds reason very naturally in terms of how believable (or `likely', or `probable') are different hypotheses in the light of everything we know about them (see e.g. [15]) and the mathematical theory of how beliefs are updated by new pieces of information was basically developed in a monumental work of Laplace exactly two hundreds years ago[16,17], although nowadays this way of reasoning goes under the name Bayesian. This approach considers valid sentences such as ``probability that the CDF bump is a fluke'', ``probability that the Higgs boson mass is below 130 GeV,''9 and similar, all expressions that refer to ``a problem in the probability of causes, [...] the essential problem of the experimental method''[19]: from the observed effects we try to rank in probability the alternative causes that might have produced them.

Giulio D'Agostini 2012-01-02