Johanna,

I am not sure if you can use 3dDeconvolve to get the change from onset. Doug Ward will have to answer that for sure, but I doubt it can. I have only used that when the trials are spaced far enough apart to allow the signal to come back down to 'baseline' and I have done it manually with formulas in Excel.

I have high-pass filtered my SCR before using a filter I created with Matlab's special filter designer. After filtering, alot of the baseline variability I was talking about was gone so I presume a combination of high-pass filtering and 3dDeconvolve may work ok for you. It seems deconvolution is the only way since you are using a rapid presentation design. We had always designed our studies to be slow ER designs where there was enough time between trials for our GSR to return to baseline.

Let me know if you want a look at my filter if you are familiar with Matlab and if you dont have another filtering approach handy. I custom designed it to best filter out the SC level while leaving my SC Responses intact. I think it was based on a butterworth filter originally.

Christine Smith