Hi Giora!!

Thanks a lot for your interest! 

You are right, my question is not well formulated. 

For instance, if I just compare 2 groups, the corresponding contrast will be: [ 1 -1 ] and it is a 2 sample t-test.

My questions is, if I got 3 groups and I need to perform an ANOVA (main effect of group), How should I do the corresponding contrast? 

Thanks again..

Marisol