Hello,
I am studying the effect of different coviariates (score1, score2) on different seeds (seed1 seed2), on a seed to voxels analysis (usually I use an uncorrected p-value for the height of the cluster and an FDR corrected p-value for its size, using parametric analysis).
I looked at the independent effects of score1 and score2 on seed1 and seed2 independently, at the significant differences between the effect of the two scores using the contrast [1 -1], but would also like to see the overlap of the effects of the scores on each seed. Does [1 1] represent this overlap, ie where both scores significantly correlate with the connectivity of one seed? (the results seem different than if I had taken the maps corresponding to each effect independently and only plot where both maps showed significant results - it is possible that it makes sense statistically, I am wondering if it does)
I would also like to display the effects of one score on the two seeds, where there are significant results for both seeds and for at least one of them. Actually it is the same question: how to plot the AND and the OR (the 'eye' function that appears by default seem to plot the regions where there is an effect for one seed and not the other so exclusive OR, does inclusive OR corresponds to the contrast [1 0; 0 1; 1 1]?)
I hope it's clear enough.
Thanks in advance for any answer.
Lena
I am studying the effect of different coviariates (score1, score2) on different seeds (seed1 seed2), on a seed to voxels analysis (usually I use an uncorrected p-value for the height of the cluster and an FDR corrected p-value for its size, using parametric analysis).
I looked at the independent effects of score1 and score2 on seed1 and seed2 independently, at the significant differences between the effect of the two scores using the contrast [1 -1], but would also like to see the overlap of the effects of the scores on each seed. Does [1 1] represent this overlap, ie where both scores significantly correlate with the connectivity of one seed? (the results seem different than if I had taken the maps corresponding to each effect independently and only plot where both maps showed significant results - it is possible that it makes sense statistically, I am wondering if it does)
I would also like to display the effects of one score on the two seeds, where there are significant results for both seeds and for at least one of them. Actually it is the same question: how to plot the AND and the OR (the 'eye' function that appears by default seem to plot the regions where there is an effect for one seed and not the other so exclusive OR, does inclusive OR corresponds to the contrast [1 0; 0 1; 1 1]?)
I hope it's clear enough.
Thanks in advance for any answer.
Lena