Let=92s say you have 200 data points. CB comes up with a Chi Square
value. But, how do you calculate correctly the number of Degrees of
Freedom to come up with a P value. I personally assumed that CB
treats each data points as a different group. So, in this case I
would come up with 200 =96 1 =3D 199 groups. But, if CB actually groups
the data in fewer groups I have way overstated the number of DF and
end up with a P value that is a lot higher. By doing so, I would have
significantly overstated the fit of the distribution. However, do you
know for a fact that my assumption is wrong? Do you know how to
correctly figure out the number of groups?
Thanks for any info on this.
Guy
