Tom's post is here:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?
A2=ind1102&L=FSL&P=R35177&1=FSL&9=A&J=on&d=No+Match%3BMatch%
3BMatches&z=4
What is it that are you trying to do? Do you want to make any across
group comparisons on either the group means or the HPA slopes? If you
are solely interested in just the regression (correlation) of your HPA
regressor for each group separately, and the groups have no hypothesized
relationship to each other, then why not just run each group in an
separate model?
cheers,
-MH
On Thu, 2011-02-24 at 22:58 +0000, Hongyu Yang wrote:
> Hello Michael,
>
> If I just want the specific correlation for each group separately, was what I did correct?
>
> For I did not find the recent post/comment on this from Tom Nichols, could you give me the link?
>
> For I do not need to think the interaction of group and HPA, can I just do this for each group to get their specific correlation based on the description in the FSL website, although I know that will cost double times.
>
>
> Thanks,
> Hongyu
|