 |
 |
Dear Xiang, > One more question, just curious, what's the difference between inputting TIV as global calculation and covariate, are they actually the same? In addition, for global calculation, it is suggested to set "Proportional" for global normalization. Is this necessary for covariate? If you set TIV as a global calculation it essentially corrects for a single influence of TIV across the whole brain (there is just a single scaling involved). If you set TIV as a covariate, then it can have a different effect at each voxel—i.e., you account for, at each voxel, the variance that can be explained by TIV. It's not really a matter of one method being "correct", but they are telling you slightly different things. There is more on this in the Barnes et al. (2010) paper, in which I believe they recommend generally including it as a covariate. As to your second question, when entering TIV as a covariate you don't have to worry about calculating a proportion—just add the raw TIV values. Hope this helps! Jonathan -- Dr. Jonathan Peelle Department of Neurology University of Pennsylvania 3 West Gates 3400 Spruce Street Philadelphia, PA 19104 USA http://jonathanpeelle.net/