I ran 3dDeconvolve with 3 predictors on 27 subjects and am now looking at the full model F. What is the best way to generate a group stat at every voxel based on the F value from the full model?
Because 3dttest works on the t-values, should i do the following?
convert F to t values; sub-brik 1 is the full model F value:
3dcalc -a "file[1]" -expr "sqrt (a)" -prefix fileFtoT
glue F back on (to make sure that t^2 = F):
3dbucket
generate group t-map:
3dttest
Is this proper?