Hi Ines,
I am experiencing exactly the same problem. I realized that if I estimated the contrasts separetely (and not entering the indices in the batch tool), the results are as they should be (A > B is exactly the inverse of B> A). Thus I guess that something goes wrong when entering the contrast index in the batch module. In my case I just entered as 1 2 3 4... Althoug it seems to compute the contrasts, I am sure that something goes wrong. So my question here is, HOW should the contrast index be introduced? May be 001 002 003 ? Any one found the proper solution?
Thanks!
|