Xu, Jiansong wrote:
> Dear Experts:
>
>
>
> I am working on nBack smoking study. I have 3 groups, i.e., non-smokers,
> smokers with satiety condition, smokers with withdraw condition. Now, I
> am trying to compare the difference between groups. If I use the basic
> model, I can compare any two groups, but I do not know how to compare
> the 3 groups ( something similar to anova). Could anyone give me some
> suggestions?
>
Dear Xu,
If you have (only) 3 groups (ie. conditions)
then you can implement an ANOVA using the
default F-contrast labelled 'effects of interest'.
Do this:
1. RESULTS
1b. Select SPM.mat file
2. Show [F-contrasts]
3. Select [001{F}:effects of interest]
4. Press Done
5. etc.
Note - if you press the green question mark button at the
bottom right of the contrast manager window you'll get a
quite thorough description of what t- and F- contrasts are.
If you have more than 3 conditions in your
experiment then you'll have to hand-craft your F-contrast
so that it has support only over your 3 groups described
above
ie.
F=[1 0 0 0; 0 1 0 0; 0 0 1 0] (matlab notation)
will just do an ANOVA for the first 3 conditions (which may be
your 3 groups).
Best wishes,
Will.
>
>
> Thanks
>
>
>
> Jiansong
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7478
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|