Dear FSL users,
I have what is likely a very basic question, but nonetheless valuable for me to try and clarify.
We have used featquery to calculate COPE as %-signal change values from a functionally-defined ROI. In our case, featquery is pointed to a gfeat directory - single-subject data, averaged across four functional runs.
For each COPE, a table of values is generated, and includes, for example:
pe: mean = 0.3526, stddev = 0.203
zstat: mean = 2.721, stddev = 1.405
Can we straightforwardly interpret the mean zstat?
Specifically, I interpret this value as the z-statistic of the specified COPE (which reflects the average across all four runs - the gfeat cope) averaged across all voxels in the ROI. Is this correct?
Further, continuing with the example values above, would it then be appropriate to compute a p-value based on the mean zstat of 2.721 (two-tailed) = 0.0065.
If so, does this then suggest that the mean parameter estimate of 0.3526 is statistically non-zero, p < 0.01?
Thank you in advance, we greatly appreciate your input, and the time you've taken to share it.
Sincerely,
Ken Valyear
|