Anil:
I'm not sure how much more succinct I can be. I teach intro. to
statistics, and my students complain that I don't explain things in
English very well. They want me to teach a mathematics-free statistics
class! I don't think such a thing can be done. I would suggest you
read a book like Intro. to Prob. and Math. Stat. by Bain and Engelhardt.
Beyond that, I think I have a bit more to add to the solution. But,
this doesn't seem to be particularly Bayesian as I have constructed it.
H_0: \mu=0 vs H_1: \mu NE 0
(sum of {X_i}^2)/n
T = -------------------
(sum of {Y_j}^2)/m
If you have n A events, and m B events, then T will be distributed as F(n, m) under
H_0. This tests something like: B is more likely to be near R than A.
--
Rodney Sparapani Medical College of Wisconsin
Sr. Biostatistician Patient Care & Outcomes Research
[EMAIL PROTECTED] http://www.mcw.edu/pcor
Was 'Name That Tune' rigged? WWLD -- What Would Lombardi Do
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
