 |
 |
See inline responses below. On Mon, Feb 25, 2013 at 5:26 PM, Daniel Cole <[log in to unmask]> wrote: > Hi Donald thanks for the response, > > >> You can't do these types of statistics on 2 subjects. > > Do these types of designs simply lack power for 2 subjects? I apologize as > I'm not particularly savvy about design efficiency and such things. > Wouldn't the linear trend analysis be an "ok" analysis for a small subject > pool since it's comparing within subjects? Are there other types of > analyses that you could recommend? >>> There really aren't any designs that work with 2 subjects. Jeanette Mumford has posted on this before. The statistics in FSL are the same as any statistics program. Think of a one-sample t-test, you wouldn't be allowed to use it with 2 subjects. Most people would say you need at least 5-10 subjects before the t-test would work. For small samples, randomise will be better, but with 2 subjects, it still wouldn't work. When you collect your 10 subjects, then the tests will be possible. > > I think you may have been mistaken about my EV values for the 2nd design. > The 3rd and 4th session EVs are technically -0.5 and +0.5 but are written in > scientific notation (-5.000000e-01). >>> Yes. I missed that change in notation. The linear trends are fine. You should put both subjects linear columns in the same column. This analysis will use the within-subject variance, so you'll have a fixed effects analysis as opposed to a random effects analysis. One solution, once you have more subjects is to compute the slope for each subject and then feed the slope into one-sample t-test. > > Thanks again for your advice! > > Daniel > > > > On Mon, Feb 25, 2013 at 2:53 PM, MCLAREN, Donald <[log in to unmask]> > wrote: >> >> Daniel, >> >> You can't do these types of statistics on 2 subjects. It looks like #1 >> is correct, but #2 is odd as S3 is the first scan collected and S4 is >> the last scan collected. If the scans go from S1 to S6, then you will >> want the covariate to go in the same direction across all scans. [-2.5 >> -1.5 -.5 .5 1.5 2.5] >> >> Best Regards, Donald McLaren >> ================= >> D.G. McLaren, Ph.D. >> Research Fellow, Department of Neurology, Massachusetts General Hospital >> and >> Harvard Medical School >> Postdoctoral Research Fellow, GRECC, Bedford VA >> Website: http://www.martinos.org/~mclaren >> 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 Mon, Feb 25, 2013 at 1:20 PM, Daniel Cole <[log in to unmask]> >> wrote: >> > Hello FSLers, >> > I'll start by explaining my experiment. We have 2 monkey subjects who >> > have >> > been receiving training on an identical test over the period of months. >> > I >> > currently have 6 structural scans (6 test sessions) and would like to 1) >> > see >> > if any one session is different from the last (6th) scan and 2) see if >> > there >> > is a linear growth or shrinking of grey matter over the 6 sessions. >> > >> > >> > 1) ANOVA 1 Factor 4-Levels (Repeated Measures) >> > From looking at the FEAT Manual >> > (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FEAT/UserGuide#FEAT_Basics) I've >> > created the following design matrix and contrasts. The first two EVs >> > are >> > separate mean EVs for each subject and the rest model each session >> > (except >> > the 6th) across both subjects. >> > >> > Matrix >> > 1.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 >> > 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 >> > >> > Contrasts >> > C1:1-6 0.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 * >> > C2:2-6 0.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 >> > 0.000000e+00 0.000000e+00 * >> > C3:3-6 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 >> > 0.000000e+00 0.000000e+00 * >> > C4:4-6 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 * >> > C5:5-6 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 * >> > C6:6-1 0.000000e+00 0.000000e+00 -1.000000e+00 0.000000e+00 >> > 0.000000e+00 >> > 0.000000e+00 0.000000e+00 ** >> > C7:6-2 0.000000e+00 0.000000e+00 0.000000e+00 -1.000000e+00 >> > 0.000000e+00 >> > 0.000000e+00 0.000000e+00 ** >> > C8:6-3 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > -1.000000e+00 >> > 0.000000e+00 0.000000e+00 ** >> > C9:6-4 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > -1.000000e+00 0.000000e+00 ** >> > C10:6-5 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > 0.000000e+00 -1.000000e+00 ** >> > * and ** denote F-Tests >> > >> > My questions for the ANOVA are a) Is this the correct setup for the >> > question >> > I posed (looking for any difference effect with N scan and the 6th >> > scan)? >> > and b) What exactly would the -1 contrasts be showing? The activation >> > for >> > corresponding contrasts (C1 and C6) have no overlap, which might be >> > expected, but I'm still unsure "what question" this contrast is asking. >> > >> > >> > 2) Linear Trend >> > >> > Matrix >> > 1.000000e+00 0.000000e+00 -2.500000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 -1.500000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 -5.000000e-01 0.000000e+00 >> > 1.000000e+00 0.000000e+00 5.000000e-01 0.000000e+00 >> > 1.000000e+00 0.000000e+00 1.500000e+00 0.000000e+00 >> > 1.000000e+00 0.000000e+00 2.500000e+00 0.000000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 -2.500000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 -1.500000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 -5.000000e-01 >> > 0.000000e+00 1.000000e+00 0.000000e+00 5.000000e-01 >> > 0.000000e+00 1.000000e+00 0.000000e+00 1.500000e+00 >> > 0.000000e+00 1.000000e+00 0.000000e+00 2.500000e+00 >> > >> > Contrasts >> > C1: Sub1 Mean 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 >> > C2: Sub2 Mean 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 >> > C3: Sub1 Linear+ 0.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 >> > C4: Sub2 Linear+ 0.000000e+00 0.000000e+00 0.000000e+00 1.000000e+00 >> > C5: Sub1 Linear- 0.000000e+00 0.000000e+00 -1.000000e+00 0.000000e+00 >> > C6: Sub2 Linear- 0.000000e+00 0.000000e+00 0.000000e+00 -1.000000e+00 >> > >> > Does this design look appropriate for the question- 'Where is grey >> > matter >> > increasing linearly across sessions and where is it decreasing?' >> > >> > >> > Thank you very much for your time and help in teaching me to understand >> > this >> > program. >> > >> > Daniel >> > -- >> > Daniel Cole >> > University of Rochester >> > Brain and Cognitive Sciences >> > [log in to unmask] > > > > > -- > Daniel Cole > University of Rochester > Brain and Cognitive Sciences > [log in to unmask]