[R-sig-ME] GLMM with only one predictor variable

[Douglas Bates]

It is certainly possible to fit such a model.  The formula would be like

y ~ 1 + (1 | grp)

We define the distribution of the random effects to have a mean of zero so
the (Intercept) term generated by the first 1 is needed to allow for a
nonzero mean.

You may use such a model for screening but the advantage of being able to
examine the effect of one variable in the presence of multiple influences
on the response.

On Mon, Mar 11, 2024, 14:57 Ana Hernandez <anahmdlr using gmail.com> wrote:

> Dear all,
>
> Is it possible to run a GLMM with a random effects variable as the only
> predictor variable? Would it make sense to do this to assess the impact of
> a single variable on the response variable?
>
> Thank you,
> Ana
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models using r-project.org mailing list
>
> https://urldefense.com/v3/__https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models__;!!Mak6IKo!KjBNHCSaUPpyxJ6fFDOkfhxSdA8IIeb1-VrEORYKrOyq7Ic9YVjX_YYNtSaC5plih7cChjrEca8vHAyy6A$
>

	[[alternative HTML version deleted]]