Matthew,
The upload reference # is 875810.
For this featquery analysis I used all the PE's. PE's 1-6 correspond to the 3 EVs and their temporal derivatives. I believe the rest are the motion parameters.
If you look within the featquery directory Hippo_modal_mask_L and then look at the peristimulus graph values for the mean of zstat3, you will see values in the -8,000,000 range. Zstat3 is comparing EV3 (EV2 as specified in the FEAT GUI) against baseline. Since the graph follows the timecourse of EV3, it almost seems like the value is being scaled. Maybe?
There are many subjects for which featquery outputs odd values, it just depends on what mask I use and which zstat I am looking at. One subject might might not have any problems using one mask, and then I use another mask and the values get weird. This also becomes a problem because I am not sure which values to consider "normal." Based on looking through many of the results, it seems that normal values should range from about 6000 to 10 or 110000.
There are other Featquery directories within the main feat folder using different masks, but I ran some of those without specifying all the PEs. I didn't know I HAD to specify all the PE's until Mark pointed that out the other day.
Thank you,
Carlos Faraco
|