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

[Douglas Bates]

Grr. I shouldn't try typing on a tablet.  The last sentence was supposed to
say something about allowing for multiple terms in the formula can provide
information about the effect of a condition for accounting for other
conditions.

On Mon, Mar 11, 2024, 15:30 Douglas Bates <dmbates using gmail.com> wrote:

> 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]]