afni.nimh.nih.gov

How many time points do you have? For 6 stimuli, I think you'd want at least 6*5*10=300 repetitions at TR=3 to be able to get decent results.

Did you do simple regression (i.e., generate waveforms with waver, then use 3dDeconvolve with minlag=maxlag=0), or do true deconvolution?

bobcox