Rick,
Thank you for your response. That was what caused the error.
I am pretty new to fMRI analysis; I just wanted to also include the baseline as though it were another active condition. This means that the boxcar function for the baseline condition will then also be convolved with an HRF before fitting to the data. This seems incorrect. However, since I suspect the subject might do something else during the baseline condition (not a complete rest), I wanted to see if there is any effect. Do you think this is a problem practically?
Thanks again,
Duong