Dear James
Sorry for delayed response I was out of city for few weeks.
To perform compare-groups analysis under SPMWin you should create new
document (*stl) using menu item File New Group's Statistics with proper
number of groups and fill in it with proper data file names (*hdr or *vhd)
by dragging and dropping. After that you should set parameters and
contrasts with menu item Calculate Settings... and perform analysis with
menu item Calculate Start.
We have checked this type of statistical analysis many times and always got
the same results as in the original SPM96. So, it is very surprising that
you got different results.
To say more I need to have your *.stl file for compare-groups analysis as
well as the set of data files used in groups.
Sergey Pakhomov
>
> I am having problems with both SPM96NT and SPMWin v1.01. I am trying
> to do a simple compare-groups analysis of SPET scans. I have a large
> group of normal scans to which I compare single patient scans. I have
> carried these analyses out on a SUN based system with no problems, and
> am familiar with the results.
>
> Using the same dataset of 24 normals and one brain-injured patient,
> I get robust results as expected on the SUN. On the PC, I get the
..........................................
>
> SPMWin v 1.01 - using groups comparision with parameters and
> contrasts set the same as in SPM96, it tells me that there is no
> significant data!
>
>
> I am reasonably certain that SPM96NT is set up identically to SPM,
> as the interface is identical for all intents and purposes. I am less
> sure with SPMWin 1.01, as I just recently began using it, and am not
> familar with the interface. However, it does not seem very complicated.
>
> All the images have been visually checked. They were coregistered,
> spatially normalized, and smoothed using SPM96NT.
>
>
> Any help anyone can give me would be greatly appreciated. I can
> always go back to the SUN, I suppose.
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|