Hi - yes this makes sense. If EV2 explains some variance in the data
and doesn't highly correlate with EV1 then yes you would expect
significance testing on PE1 to go up.
Cheers.
On 29 May 2009, at 03:37, Isabelle Haberling wrote:
> Dear Fsl people
>
> I’m using TBSS to see if FA values correlate to a behavioural
> measures while controlling for a confound variable. I’ve already
> made a regression with 2 EV’s (both demeaned) and used the –D option.
>
> Mat file:
>
> /NumWaves 2
> /NumPoints 102
> /PPheights 9.00000e-01 2.00000e+02
>
> /Matrix
>
> EV1 EV2
> 2.6 -8
> 2 4
> -1 2
> 1 3
> .. ..
>
> Contrasts
> 1 0 (positive effect of EV1 controlled for EV2?)
> -1 0 (negative effect of EV1 controlled for EV2?)
>
> Is this the right way to control for EV2? My results seemed to get
> more significant, which I found strange...
>
> Or is there a possibility to do a partial correlation?
>
> Any help would be appreciated!
>
> Kind regards
>
> Isabelle
>
>
>
>
> __________ Information from ESET NOD32 Antivirus, version of virus
> signature database 4113 (20090528) __________
>
> The message was checked by ESET NOD32 Antivirus.
>
> http://www.eset.com
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|