 |
 |
Hi, > I did an analysis with 3 gamma bases model. The result shows that > each of the gamma function is not significant, however, the f stat > of c1 &c2 &c3 is significant. > First, what is the meaning of the f stat of (c1 & c2 &c3)? (both > statistically and physiologically) F1 looks at the total variance explained by the 3 regressors EV1-3. Therefore it has found a linear combination of EV1-3 that fits better than any single one does, hence has reached significance. Ideally this means that this linear combination represents a better estimate of the true HRF than any single EV - this is the point of using basis functions and feeding them into an F. > Second, I only see cope1 cope2 and cope3 ( of C1, C2 and C3) and no > cope for f stat of c1&c2&c3. How should I do a fixed factor > analysis for the f stat? cope1,2,3 get converted into tstat1,2,3 and are also used to generate fstat1. Then these are converted into zstat1,2,3 and zfstat1, respectively. Thresholding is then carried out on these Z images. We don't in the GUI support fixed-effects analyses. You can carry out a "simplified fixed-effects analysis" on the zfstat1 image, using the normal equation (and avwmaths) of zfixed=sum(z)/sqrt(n) - good luck. (ps - don't forget to put everything in standard space before combining!) Cheers, Steve. Stephen M. Smith DPhil Associate Director, FMRIB and Analysis Research Coordinator Oxford University Centre for Functional MRI of the Brain John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve