On Mon, 21 Jul 2003 19:17:39 GMT, Xcott Craver <caj@B-r-a-i-n-H-z.com>
wrote:

Kavik Kang wrote:

You are the one re-writing it. What it says is ""extraordinary claims
require extraordinary evidence", which is incorrect. Extraordinary
claims require the same exact type and level of evidence as any other
claim.

      The property of extraordinary claims requiring extraordinary
      evidence is a natural and direct consequence of the mathematics
      behind Bayesian hypothesis testing.

Whoa...You are correct, of course. But the person to whom you are
responding, KK Kongo, is a real dumbass and is highly unlikely to
comprehend anything about mathematics beyond the addition of 'to' and
'too'.

      You just do the math, you work out the optimal threshold
      for deciding between a null and alternative hypothesis, and you
      find that the threshold depends in part on the relative
      plausibility of each claim (i.e., the estimated a priori
      probabilities of each at the present time.)  An extraordinary
      claim results in a higher threshold, requiring a lot more
      data points in its favor, or data points which extraordinarily
      favor that claim over the other.  Or both.

      This is not an arbitrary principle some skeptics made up, but
      a plain English restatement of a basic mathematical result in
      hypothesis testing.  Yes, extraordinary claims require
      extraordinary evidence.  Perhaps you can explain in further
      detail why you have concluded the opposite?

Thank you for your post, even if those pearls fall upon the deaf ears
of the swine.


Pete Charest
Insult Yourself, I'm Busy