Bayes Home
Jaynes Errata
next up previous


Chapter 4: Elementary hypothesis testing

  • p. 104, last paragraph: ``The exact threshold turns out to be greater than 2/3: $f_t = 0.793951$.'' I get a slightly different result doing the calculation to double precision: 0.794155 (only first six digits printed out). Another reader reports obtaining $f_t = 0.794155259$.

  • p. 115, second-to-last paragraph: ``In the case of a uniform prior pdf for $f$, we may use (4.64) and the result is the incomplete beta-function.'' Should this actually be a reference to (4.67), which is the evaluation of (4.64) for the case $g(f \mid X) = 1, 0 \leq f \leq 1$?

Miscellaneous commentary

  • p. 97, Exercise 4.1. Timothy D. Sanders has contributed a partial solution for the case of three hypotheses and two data points; a version of his solution, to which I have made some minor modifications, is available in postscript or PDF.

  • p. 116, second to fifth full paragraphs: The term ``logit'' is not that recent: see Joseph Berkson (Application of the logistic function to bio-assay. JASA 39: 357-365, 1944); in many fields it is now totally standard, and far more widely used than the other terms mentioned here. [Contributed by Nick Cox.]

next up previous