Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nick Cox <njcoxstata@gmail.com> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | Re: st: Re: Marginsplot on backtransformed data |
Date | Fri, 20 Dec 2013 09:52:07 +0000 |
<> Quite so. My stance on GLMs is very positive. Nick njcoxstata@gmail.com On 20 December 2013 02:47, Richard Williams <richardwilliams.ndu@gmail.com> wrote: > Perhaps I misunderstand him, but I got the impression Nick preferred the glm > approach. Of course, if it makes little difference in the results, you may > want to go with whatever approach makes it easiest to generate the graph you > want. At 09:10 PM 12/19/2013, Daniel Herbert Opi wrote: >> From the explanation by Nick Cox on -glm- versus -regress- I think I >> am right in assuming that I would rather first square transform the >> data then run the regress.