 |
 |
Hi, Yes, see the sections on GLM and contrasts in the FEAT manual, including the appendix for some info on the GLM. The difference between the t-test and the multiple regression (GLM) with added confounds is that the extra covariates (confounds) soak up variance in the data and reduce residuals (which generally increases Z-stats), and also may correlate partially with the regressors of interest (unless you orthogonalise), which can reduce Z-stats. Cheers. On 7 Apr 2008, at 22:30, Liberty Hamilton wrote: > Hi Steve- > > Thank you so much for your speedy reply. I was wondering if you (or > any of > the FSL people) could clarify what type of statistic is calculated > for the > contrasts when covariates are included in the model. I ask this > because when > I ran an unpaired T-test and calculated the patients>controls > contrast and > controls>patients contrast, the result was similar, but not exactly > the > same, as the result I obtained for the contrast I described in my > previous > message (sex and age as covariates, but not included in the > contrast). So, > to clarify, here is the original "contrast" (scenario 1) and the t- > test > (scenario 2): > > SCENARIO 1: > Group control patient age sex > 1 1 0 21 1 > 1 1 0 24 -1 > 1 1 0 35 -1 > 1 1 0 42 1 > 1 0 1 23 1 > 1 0 1 43 -1 > 1 0 1 22 1 > > Contrast control patient age sex > ctrl>pat 1 -1 0 0 > pat>ctrl -1 1 0 0 > avg ctrl 1 0 0 0 > avg pat 0 1 0 0 > > SCENARIO 2 (age and sex covariates removed from the model) > Group control patient > 1 1 0 > 1 1 0 > 1 1 0 > 1 1 0 > 1 0 1 > 1 0 1 > 1 0 1 > > Contrast control patient > ctrl>pat 1 -1 > pat>ctrl -1 1 > avg ctrl 1 0 > avg pat 0 1 > > Could you explain whether (and how) age and sex are included in the > model > when I calculate a contrast (without an F test), and why this would > or would > not be the same as performing a T-test without these covariates? > > Thank you again for your time. Your help is truly appreciated! > > Liberty > --------------------------------------------------------------------------- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ---------------------------------------------------------------------------