Dear Tuba,
you should treat family as a factor and the family ID as the levels of
that factor. In the (flexible factorial) model, you would add a main
effect for this factor to get the family means estimated.
Hope this helps,
Volkmar
Am Freitag, dem 15.04.2022 um 10:08 +0300 schrieb Tuba Şahin:
> Dear Volkmar,
>
> Thank you very much for your response. I have a question regarding
> the practical side of it. Let's say, I do a regression analysis then
> how should I enter the family ID? As a covariate or something else to
> be able to model per-family means and get family-specific covariance
> options?
>
> Best Wishes,
> Tuba
>
> Volkmar Glauche <[log in to unmask]>, 11 Nis 2022
> Pzt, 11:55 tarihinde şunu yazdı:
> > Dear Tuba,
> > you may want to introduce a factor "family" and assign each family
> > to
> > a distinct level of this factor. This will let you model per-family
> > means and family-specific error covariance options.
> > Hope this helps,
> > Volkmar
>
>
|