Hi Gang,

Thanks for you help. I take it from your answer that the ANOVA won't work for unequal sample sizes in each group (11 13).

I would like to compare the following
two groups: patients and controls
two conditions: colour and achromatic

I'd like to find out if there is a
main effect of group for each condition
main effect for each condition
any group x condition interactions
If I use 3dttest for each of the conditions would I threshold the maps and use 3dcalc to Or the maps from each of the conditions together and then run ROIstats with this combined mask for each condition and run an ANOVA in spss to look for any interactions? Is it possible to get activations where colour > achromatic condition as the stimulus was designed to have the same image characteristics to explore the regions that are activated in response to colour and see if there is a effect in these colour specific areas.

Any advice would be much appreciated .

Thanks again

Erik