Hi Elaine,

Yes, you are exactly right, but to be precise that analysis would implement a omnibus/multivariate test evaluating those two simple main effects jointly so it would perhaps more properly be refered to as the omnibus/multivariate main effect of condition. In addition, you can also evaluate the individual simple main effects of condition (separately for each group) using the between-subject contrasts [1 0 0 0 0 0] and [0 1 0 0 0 0] separately in two different analyses, or the standard/univariate main effect of condition using the contrast [.5 .5 0 0 0 0]. 

Hope this helps

Alfonso

Originally posted by Elain Liao:

Dear Colleagues

I have two between-subjects contrasts: Group A & Group B and two conditions: Pre- & Post-.

I  used CONN to do a Time*Group ANCOVA and added covariates of no interest (i.e Age, Education year, Baseline clinical score1, Baseline clinical score2).

Firstly, I selected

a)"Group A", "Group B", "Age of A&B", "Education Year of A&B", "clinical score 1", and "clinical score 2" in the subject-effects list, and entered a [1 -1 0 0 0 0 ] between-subjects contrast

b) 'Pre' and 'Post' in the conditions list and entered a [-1 1] between-conditions contrast

and then successfully got the significant group by scan interaction.

Now I wonder how to set the model to determine the simple main effects of conditions, simple main effects of group, main group effects, and main condition effects.

Take "simple main effects of conditions" for example,

is that correct that I select

a)"Group A", "Group B", "Age of A&B", "Education Year of A&B", "clinical score 1", and "clinical score 2" in the subject-effects list, and enter a [1 0 0 0 0 0; 0 1 0 0 0 0] between-subjects contrast

b) 'Pre' and 'Post' in the conditions list and enter a [-1 1] between-conditions contrast.

All your comments are highly appreciated,
Elaine