As Eugene mentioned, if the subjects are not group1 are different from group 2, then you'll need a different model. Additionally, since you have both a between-subject factor and a within-subject factor, your new model will be able to address the effect of time and the group*time effects, but not the effect of group. The reason for this is that there is a single error term for the model which is the within-subject error. If you use this error term, you get inflated statistics. For more details on this issue see: http://www.martinos.org/~mclaren/ftp/presentations/OHBM2011_v3.pdf Best Regards, Donald McLaren ================= D.G. McLaren, Ph.D. Postdoctoral Research Fellow, GRECC, Bedford VA Research Fellow, Department of Neurology, Massachusetts General Hospital and Harvard Medical School Office: (773) 406-2464 ===================== This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is intended only for the use of the individual or entity named above. If the reader of the e-mail is not the intended recipient or the employee or agent responsible for delivering it to the intended recipient, you are hereby notified that you are in possession of confidential and privileged information. Any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited and may be unlawful. If you have received this e-mail unintentionally, please immediately notify the sender via telephone at (773) 406-2464 or email. On Sat, Oct 22, 2011 at 5:50 AM, Jon Brock <[log in to unmask]> wrote: > Hi, > > I apologise in advance as this subject seems to be getting a lot of air time at the moment > (on this and other lists) but as a relative novice I am still a little confused about > how best to proceed. > > A number of earlier emails have included questions related to a 2x2 repeated measures design > (with groups (e.g. controls vs patients) as a between subjects factor and time > (e.g pre and post) as the within subjects factor. I have attached what I believe to be the > correct design matrix to probe the main effects of group and time and the group x time > interaction (inputs are lower level feat directories, and the subjects are entered as: > Subject1-Group1-Scan1 > Subject2-Group1-Scan1 > Subject3-Group1-Scan1 > Subject4-Group1-Scan1 > Subject1-Group2-Scan1 > Subject2-Group2-Scan1 > Subject3-Group2-Scan1 > Subject4-Group2-Scan1 > Subject1-Group1-Scan2 > Subject2-Group1-Scan2 > Subject3-Group1-Scan2 > Subject4-Group1-Scan2 > Subject1-Group2-Scan2 > Subject2-Group2-Scan2 > Subject3-Group2-Scan2 > Subject4-Group2-Scan2 > > My naive questions are: > > 1) Does the attached design correctly probe the above questions? > 2) If so, are there any caveats that need to be noted when reporting the > results using this sort of approach? > 3) If not the above then what is the optimum way to ask these questions with this type > of design? A big ask I know but perhaps an image of the required design matrix/matrices > as an attachment to list or added to the FSL website. > > > Apologies again. > > Jon > >