Dear Libby and Eugene,
Following up here, I am also finding some discrepancies between PE versus COPE values pulled from featquery - only when I'm apply featquery to single run - (feat) - data; when i run featquery at the level of multiple runs - gfeat - the PE and COPE values for the simple contrasts (i.e. condition predictors vs. rest) they match up, as I would expect.
So, do I understand your response correctly, Eugene:
- do the COPE values reflect the influence of temporal derivatives, whereas the PE's do not?
- In my data, for most subjects, for most conditions, the PE's and COPE values (again for these simple contrasts of condition vs. rest) are similar. However, in some cases they are very different (even flipping sign, + to - for example). Does this make sense.
Thanks in advance. I certainly appreciate your time and consideration.
Ken
|