Hi, and thank you for your reply.
The values are similar but in some cases are off by 2 or 3 (14.1 versus 11.5, for example). I did use FLAME at the third level but not outlier deweighting, I can rerun my analyses to see how the outlier deweighting changes the values.
The biggest issue I discovered this morning is that my contrast values don't match my condition values. At the first level I have set up my GLM to model condition A, condition B, condition C, etc.... as well as the difference between these conditions (A minus B). When I extract the cope values for both the conditions and contrasts at the 2nd level (single subject), the values for the contrasts do not match up with the subtraction of the individual conditions that make up these contrasts. At the first level (at the run level for a single subject), these values to match. Is it to be expected that they wouldn't match up at the 2nd level (I am using fixed effects at the second level)?
Jill
|