Dear Jaivijay,
A simple two-sample t-test should suffice here, with eight images
in group 1 and 4 in group 2.
SPM will, by default, assume the
variance in response is the same in each group (this way it will get
a better estimate of what that variance is - because it has 12 data points
to play with at each voxel). You can then compare the group
means using a [1 -1] t or F- contrast (to look for one or
two-sided differences).
Because you don't have many data points you won't have much power but may
nevertheless find something interesting.
Best,
Will.
Jaivijay Ramu wrote:
> Dear SPM ers,
> I have a very naive question, I have two
> different treatment groups of animal data, one group which has 8, the
> other group has 4, is there a way I can do a simple t-test using SPM to
> find if there is significant difference between the two groups, all the
> animal data have been registered to one standard image and I am thinking
> of averaging all the data in each group to one image, Is it the right
> approach ? Please comment on this !
>
> --
> Jaivijay Ramu
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7475
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|