Hi all and hi Nick,

I would like to seek for some advice in interpreting the result from Qdec (btw, it is a very helpful tool for thickness analysis). Let me describe the problem here. Let's say we are interested in the relationship of thickness with two variables, age and gender, one is continuous variable and one is categorical. Qdec will produce maps for 4 contrasts. We have no question about the first two contrasts. What we don't understand are the contrast of [1 -1 0 0] and [0 0 1 -1]. They always will give very similar pattern, but one is the opposite sign of the other (refer to attachments, this can be seen clearly in the histogram). This happened to other variables. With one more continuous variable involved in the model, say IQ. The maps given by contrast [1 -1 0 0 0 0] and [0 0 1 -1 0 0] will be similar.

Can some one expert in stats or Qdec share some lights in interpreting this? Thanks.

Regards,
Zheng Hui