RE: More Hard Problems Using Bayes' Theorem, Please

From: Lee Corbin (lcorbin@tsoft.com)
Date: Tue Jul 08 2003 - 19:13:28 MDT

  • Next message: Lee Corbin: "RE: Tranquility"

    I wrote

    > I can't answer your question, but did you check out problem
    > number three on my list? Apparently Bayesian statisticians
    > can solve it whereas those like me trained in classical
    > mathematics, do not have enough information. I'm still in
    > shock that not in all instances does P(A & B) = P(A|B)*P(B)
    > according to the Bayesians.

    Those remarks, as well as the problem below, come from a link
    http://www.apa.org/journals/xge/press_releases/september_2001/xge1303380.html
    in an email written by Damien Sullivan on January 26, in the
    thread entitled "Teaching Bayesian Inference in Two Hours".
    I continued:

    > Anyway, problem number three is hopefully my long sought
    > Rosetta Stone that does not require talking of distributions
    > to see Bayesianism rear its head. (I do have one smaller
    > and much easier problem that lies bare the differences, but
    > this here might be much hotter stuff.)

    Well, now it seems to me that in this particular problem, it's
    easy to think, wrongly, that an answer is possible. So I apologize
    for putting it on the list (at the time, I thought that I just
    wasn't quick enough.) I still maintain that there is not enough
    information in it. But now I believe that that goes for
    Bayesians as well as for everyone else.

    > > 3. The probability that a newborn will have deformities
    > > traceable to a sickness of its mother during pregnancy is 1%.
    > > If a child is born healthy and normal, the probability that
    > > the mother had rubella during her pregnancy is 10%. If a
    > > child is born with deformities that can be traced to a
    > > sickness of the mother, the probability that the mother had
    > > rubella during her pregnancy is 50%. What is the probability
    > > that a child will be born with deformities if its mother had
    > > rubella during her pregnancy?

    In this stupid problem, the authors badly misstate at least one
    of the premises, in my opinion. If you draw a picture of the
    problem, perhaps, you may be less likely to misread it than if
    you plug the numbers into a formula.

    And the heresy P(AB) is not equal to P(A|B)*P(B) appears in this
    quote from the above paper:

          In Germany, every expectant mother must have an
          obligatory test for rubella infection because
          children born to women who have rubella while
          pregnant are often born with terrible deformities.
          The following information is at your disposal:

            The probability that a newborn will have deformities
            traceable to a sickness of its mother during pregnancy is 1%.

            If a child is born healthy and normal, the probability
            that the mother had rubella during her pregnancy is 10%.

            If a child is born with deformities and it can be traced
            to some sickness of the mother, the probability that the
            mother had rubella during her pregnancy is 50%.

            What is the probability that a child will be born with
            deformities if its mother had rubella during her pregnancy?

          The Bayesian solution p (H | D) is .048. But participants
          who use one of two non-Bayesian algorithms, computing
          p (H&D) = .005 or picking p (H) = .01, will produce
          estimates that lie in the interval of ±5 percentage points
          around the Bayesian solution.

    But why shouldn't p(H&D) --- that is, p(defect & rubella) be equal
    to .005? What other conclusion can one come to from the first
    statement together with the third statement?

    Unless, they fail to believe that p(H&D) = p(H|D)*p(D). Sure
    enough, go back a couple of paragraphs in that paper, and find

          The most frequent non-Bayesian algorithms they identified
          include computing p (H&D) by multiplying p (H) and p (D | H);

    No wonder I can't grasp Bayesianism. ;-)

    Of course, it is a libel to say that Bayesians would be guilty of
    believing that p(H&D) was not equal to p(H|D)*p(D). But I was
    hoping that it would turn out to be the case that in some problems
    they do not---now I simply believe that the authors are full of shit.
    And I think that their study is deeply flawed too, as a predictable
    result of inflicting this problem on a lot of innocent victims.

    Lee



    This archive was generated by hypermail 2.1.5 : Tue Jul 08 2003 - 19:24:04 MDT