We would like to know which the difference between predicted and
adjusted response in a 'Fitted responses' plot is (SPM5). And the
meaning of the plots labeled 'fitted' and 'plus error' when one of the
previous options is selected.
We have a group of 50 images which their voxel values are between 0 and
1. We use a 'One-sample t-test' with a covariate. So our design matrix
has two columns: the mean and the covariate. We would like to plot the
correlation between the voxel and the covariate values. Therefore, we
select the 'plot' option in the interface (the one which shows after
pressing 'Results' button), 'fitted responses', the F-contrast [1 1],
'adjusted response', plot against explanatory variable (the covariate).
That plot shows values upper than 1 (e.g. 1.08) in some cases, for
example in the voxels which have a better adjustment to the straight
line. We supposed the maximum value in our case had to be 1.