First of all, this book is great to learn Bayesian stats.
May I ask for help in working out the "results for all hypotheses, after normalization" in P332 (e.g., P(D|H1:S independent of V and D)=0.007).
I can follow the steps in Box 10.9, but cannot work out the results for the hypotheses.
Re: Help on Example 10.9 Race Equality in U.S. ...
Thanks for your kind comments on the book.
The results on page 322 'after normalization' come from the following:
-The joint probability for the data in the case of H1 (line 3, page 331) is shown to be 4.26E-6
-The joint probability for the data in the case of H2 (middle of page 331) is shown to be 4.67E-5
-The joint probability for the data in the case of H3 (line 1, page 332) is shown to be 1.93E-4
-The joint probability for the data in the case of H4 (last line of box, page 331) is shown to be 3.78E-4
When you normalize the four values 4.26E-6, 4.67E-5, 1.93E-4, 3.78E-4 (i.e. scale them to sum to 1) you end up (to 3 decimal places) with 0.007, 0.075, 0.310, 0.608
and these are the four values respectively for the probability of the data given H1, H2, H3, and H4.
What is confusing (and we will put thisin the errata) is that when we write P(D | H1) etc at the bottom of page 332 the D refers not to the Defendant's race but to the data observed.