Hoi Rutger,
Heb jij nou afdoende antwoord gekregen op je vraag? wel of niet een regressor ?
In de first level dus wel en higher order niet ?
goeie vraag
groet
Anda
-----Oorspronkelijk bericht-----
Van: Goekoop, R. [mailto:[log in to unmask]]
Verzonden: do 31-7-2003 20:44
Aan: [log in to unmask]
CC:
Onderwerp: [FSL] Is FEAT statistical analysis 'Multi-level'?
Dear FSL users,
In a higher level analysis of the effect of medication on brain activation patterns, we have controlled for task version (1, 2 or 3) and scanorder (3 successive sessions: 1, 2, or 3) by specifying the proper EVs. We see that some part of the variance of BOLD-signal changes is explained by taskversion and some by scansession. Can we now conclude that FEAT has 'regressed out' these confounder effects, leaving a residual effect that is more representative of the medication effect, or is there still some interaction between them? If I understand correctly, multi-level statistical analysis works this way. So: is FEAT multi-level, and if not, is there perhaps a way of eliminating such effects in higher-level analyses?
Thanks,
Rutger.
Drs. R. Goekoop, MD.
Department of Neurology
Vrije Universiteit Medical Centre
P.O. Box 7057, 1007 MB
Amsterdam, the Netherlands
Phone: +31 20 444 0316
E-mail: [log in to unmask] <mailto:[log in to unmask]>
|