dear all
I would need a confirmation over the correct p-value correction i have to apply to my DR results. consider two groups with the following model
A B
1 0
1 0
1 0
0 1
0 1
0 1
with contrasts
1: 1 -1
2: -1 1
if I interpreted correctly the duaregr Wiki, my p-values should be < 0.05 * (2*#RSN) ???
>>>>You ... looked at the outputs of randomise...2, with your model being a two-group test (controls and patients). However, you didn't know whether you were looking for increases or decreases in RSN connectivity, and so you ran the two-group contrast both ways for each RSN. In this case, instead of your corrected p-values needing to be <0.05 for full significance,they really need to be < 0.05 / (12 * 2) = 0.002 !
why not < 0.05 * #RSN ?
Should the two contrast already corrected? as defined in the same GLM.
Is this request specific for melodic analysis? if I had to apply the same model with a randomise over a VBM/DTI analysis the threshold would be 0.975 ?,
moreover, concerning a 3 groups (controls vs 2 patients type) model.
do i have to extend the previous as such?
A B C
1 0 0
1 0 0
0 1 0
0 1 0
0 0 1
0 0 1
A>B: 1 -1 0
B>A: -1 1 0
.......: 1 0 -1
.......: -1 0 1
.......: 0 1 -1
........: 0 -1 1
or could I divide in three GLM A vs B, B vs C, B vs C ??
or should I create an ANOVA following the example of ANOVA 1factor - 4 levels in FEAT guide ??
A B C
1 1 0
1 1 0
1 0 1
1 0 1
1 0 0
1 0 0
with contrasts
A-C: 0 1 0
B-C: 0 0 1
mean: 1 0.33 0.33
A-mean: 0 0.66 -0.33
B-mean: 0 -0.33 0.66
C-mean: 0 -0.33 -0.33
regardless of its possible usage here in randomize, is this AVOVA 1 factor- 3levels version correct???
how would be the contrast A-B ?? : 0 1 -1
but if applied to dual regression randomize, should I also look for C-A, C-B ?
thanks in advance
Alberto
|