A. W. F. Edwards's scientific contributions

Citations

... The significance of the last term really involves the prior probability that such a term will be present. The usual thing is to keep it if it is some arbitrary multiple of its standard error, but I think it ought to be possible to frame a rule with some sort of argument behind it..." Sir Harold Jeffreys, in a letter to Sir Ronald Fisher, 1934(Bennett 1990 The Bayes factor rule that Jeffreys later derived turned out to be different from "the usual thing": the strength of the Bayes factor is not proportional to a constant multiple of the standard error, but also involves sample size. This is the paradox. ...