 |
 |
Yes, thank for the quick response... On Mon, May 30, 2016 at 8:57 AM, Charles Leger <[log in to unmask]> wrote: > Thank you for the fast response, > > Best, > Lance > > On Mon, May 30, 2016 at 3:46 AM, Anderson M. Winkler < > [log in to unmask]> wrote: > >> Hi Lance, >> This looks all fine. >> All the best, >> Anderson >> >> >> On 30 May 2016 at 02:01, Lance Stevens <[log in to unmask]> wrote: >> >>> I plan to run a some structural analyses including fsl_vbm and First >>> analyses. I am setting up the design matrix using the glm_gui. As I want >>> to control for age and gender covariates it looks like I would use this >>> design >>> http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/GLM#Two-Group_Difference_Adjusted_for_Covariate >>> rather than the simple design suggested in the fsl course example >>> http://fsl.fmrib.ox.ac.uk/fslcourse/lectures/practicals/seg_struc/index.html#vbm >>> >>> Would it be possible for someone to verify this design and contrats? Any >>> feedback would be much appreciated. >>> >>> My concern is mainly the contrasts. Following the Two group difference >>> adjusted for covariate model, I have entered 1s in the Group column. >>> Two groups of 9: 9 controls and 9 patients (total 18) >>> EVs: controls first followed by patients; EVs 3 and 4 are the covariates >>> and they are demeaned (i.e. (age-mean(age)) but the numbers are not listed >>> here. >>> Group EV 1 EV2 EV3 EV4 >>> Con Pat age sex >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 1 0 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> 1 0 1 >>> >>> C1 con > pat 1 -1 0 0 >>> C2 pat > con -1 1 0 0 >>> C3 pos_eff_age 0 0 1 0 >>> C4 neg_eff_age 0 0 -1 0 >>> C5 pos_eff_sex 0 0 0 1 >>> C6 neg_eff_sex 0 0 0 -1 >>> >> >> >