 |
 |
Dear Nick, the design specification is similar in both cases. What seems to be dramatically different are the data you entered in the batch or manually. In the "1st" case, there is a huge influence of temporal filtering on your data and design (high pass filter and AR(1) correction). This explains the differences in the displayed design matrix. Also, the bottom lines of the results table differ for both cases. In the "1st" case, SPM reports an estimated error smoothness FWHM that is three times higher than in the "2nd" case. You should go back and have a look at the image files SPM used for each analysis. Use e.g. "Review Design" and select "Files and Factors" or load each SPM.mat in turn and look at SPM.xY.P. It seems most likely that you used different images each time. Best, Volkmar