Hi Joel
>Hi,
>How can I look for a linear trend between three (or an odd number of)
>conditions?
>EG Conditions A, B, C. Contrast [-1 0 1] finds regions where C>A independ
>ent of B (I think) but I'd like to find C>B>A.
>(I'm not sure that masked contrasts or conjunctions will work but I can't
>quite get my head around that...)
>
>
try a simple correlation analysis where you enter your 3 types of con images
the contrast [2 0 -2] for A, B, C should reveal, if any, a linear
increase from C to A :-)
best
cyril
|