hi SPMers,
I have one question regarding the parametric modulation in spm2. in my study, i want to explore the neural change as a function of time within one LONG block. in order to do this, i modelled each 3s as a single event, and use time as parametric regressors. there are two blocks in one scanner session, and each was dealt with the same method. but I noticed that the regressor in the first block looked steper than the second block (model attached). compared the results to that of SPM99 with mean & exponential function, the activation map of the first block is similar, but it's different for the second block. for the results of second block, the results from SPM99 at the threshold of .01 (not .001) is similar to that of SPM2 at the threshold of .001. This seems like the decreased slope for the regressor in the second block (SPM2) actually increase the statistical value.
But shall we use two different regressor? Comments and suggestions are appreciated.
gui
2005-05-17
__________________________________________
Gui Xue, Ph.D.
Department of Psychology,
Franz Hall, Box 951563
University of California, Los Angeles.
CA90095-1563
Tel: (+1)310-825-5169
Fax: (+1)310-206-5895
email: [log in to unmask]
___________________________________________
|