To correlate the beta maps for each subject with a questionnaire, you can use
3Tcorr1D. Some instructions on that method
here.
You can also use 3dMEMA or 3dttest++ with only a single group defined and convert the t-statistic to a correlation. Instructions for that are
here.
