Hello All,
I'm still rather new to FSL and fMRI in general, so I apologize in advance if I don't use the right terms or give all the relevant details. I am analyzing a dataset in which participants played a videogame (Space Fortress) inside the scanner. It is a block design, with blocks of watching the game being played (passive) and playing the game (active). There is also a fixation period used as a baseline.
Using FEAT, I found clusters in the LDLPFC and RDLPFC that are active while participants play. I used these clusters to create ROI masks, and then used featquery to extract signal change results. I created an active>passive contrast by manually subtracting my active and passive signal change results, and then correlated this measure with scores on the game, to see if activity in either of these areas correlated with performance. To double check my results, I ran the same analysis with both PE and % signal change to compare the two. The problem is, the results are dramatically different.
Correlations between my % signal change and PE results range from -0.2 to 0.2, and not a single result holds across both measures. That is, I get statistically significant correlations between signal change and performance with both % signal change and PE, but they are entirely different.
Can anyone shed any light on why this might be happening? I'd be grateful for any help.
Best,
~Aldi
|