Gus Gassmann wrote: > Two quick observations. First, the solution to b) must be wrong, > because > the expression is > 1 if p is > 1/3. Second, the states of nature A > and C(A) > are very strangely chosen, as there is a lot of middle ground. C(A) > should be > "There exists a cow that is not white". If you then try to apply > Bayes' theorem, > you find that P(C|comp(A)) is no longer 1/2.
You are right, it should be 2p/(1+p). About the middle ground, imagine you are on a different planet :-)