I am trying to determine the transform from log odds to the original
scale. Assuming that I have x such that lo(x) is ~N(m,s^2) where lo(x)
= ln[x/(1-x)], is there a closed form to calculate the moments on x
given m and s?
George Scott
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================