Klara,
Is your "medication duration" a covariate with the same value for Time1
and Time2 of each subject, or does it differ for Time1 and Time2?
The design proposed by Dr. Auer would test whether the DIFFERENCE
between Time1 and Time2 is modulated by the covariate (medication
duration in this case). If that is what you want to test, great. On
the other hand, if medication duration differs for Time1 and Time2 of
each subject, and you want to know whether medication duration is
related to the measure itself (rather than the difference of Time1 and
Time2), then you would simply include demeaned medication duration as a
covariate (without modulating it by +/- 1). Note that such a model will
only work if medication duration differs for the Time1 and Time2
measures of each subject (otherwise your model would be rank deficient).
cheers,
-MH
On Thu, 2011-09-08 at 07:47 +0200, Auer, Tibor M.D. Ph.D. wrote:
> Hi,
>
> This issue has already been discussed. According to Stephen Smith
> (22.10.2010):
> You should “keep the full paired ttest design, and then add a new EV
> which is like the [ 1 -1 1 -1.... EV, but with each pair of 1 -1
> multiplied by the subject's score” demeaned (i.e. including both the
> EV of effect ([1 1 1 1 -1 -1 -1 -1]) (EV1) and the EV of demeaned
> score of subjects multiplied with EV1 (e.g. [2 4 -5 -1 -2 -4 5 1]))
> (EV2).
>
> S1 S2 S3 S4 EV1 EV2
> 1 0 0 0 1 2
> 0 1 0 0 1 4
> 0 0 1 0 1 -5
> 0 0 0 1 1 -1
> 1 0 0 0 -1 -2
> 0 1 0 0 -1 -4
> 0 0 1 0 -1 5
> 0 0 0 1 -1 1
>
>
> Auer, Tibor M.D. Ph.D.
> Biomedizinische NMR Forschungs GmbH
> am Max-Planck Institut für Biophysikalische Chemie
> Am Fassberg 11
> 37077 Göttingen
> Germany
> Phone/Work: +49-(0)551-201-1725
> Phone/Home: +49-(0)551-387-0076
> Mobile: +49-(0)176-8012-7921
> Mail: [log in to unmask]
>
>
> 2011.09.08. 4:08 keltezéssel, Klara Mareckova írta:
> > Hi,
> >
> >
> > I have a question about the design of my analysis in feat. I have a
> > group of patients taking mecation and I want to explore the effect
> > of medication duration. This would be a simple example of Single-
> > Group Average with Additional Covariate. My problem is that all
> > these patients were tested twice (time1, time2) which would rather
> > suggest using Two-Sample Paired T-Test. Is there a way how to
> > combine these two analysis in feat so that I don't have to desribe
> > the effect of duration in time 1 as "exploration" and the effect of
> > duration in time 2 as "replication"?
> >
> >
> > Would it be ok to simply set up the Two-Sample Paired T-test and add
> > one additionoal EV duration? Meaning "pool" all the data in one
> > analysis and then have a look only at the effects of the last EV
> > duration?
> >
> >
> > Thanks,
> >
> >
> > Klara
> >
> >
> >
|