> Dear Fish-folks,
>
> I want to compare linear regressions (for example ln length - ln weigth)
> with respect
> to differences in slope and intercept. We have found several
> complicate formulae in statistical textbooks. However, I know that one could
> use
> also the ANOVA-approach - but how? I have intensively studied the manual of
>
Thomas:
I could not get through to your e-mail address, so I'm sending this to the list.
Although it appears complicated, it actually is not too complicated to do the
correct comparison of regressions.  You essentially set up the full data set,
and separate sets to force through same slope or same intercept.  See Weisberg
1985 Applied Linear Regression pgs 179-184 for a very clear description.  It is
the "correct" and I think best approach to the problem.


Ray Newman, Assoc. Prof.             PhoneNet: (612) 625-5704
Fisheries and Wildlife               Internet: [log in to unmask]
University of Minnesota              BITNET:rmn%finsandfur.fw.umn.edu@UMNACVX

           Le hasard ne favorise que les esprits prepares
           Chance favors only the prepared mind  -Pasteur