[R-sig-ME] binomial fixed-effect p-values by simulation

[Douglas Bates]

On Mon, Aug 25, 2008 at 3:38 AM, Daniel Ezra Johnson
<danielezrajohnson at gmail.com> wrote:
> Sorry if this has been covered elsewhere, but if my interest is in
> testing a single fixed effect _term_ (all coefficients at once) is
> there an appropriate statistic to simulate for a binomial model?

> In other words, if I fit a linear model "glmodel" I can simulate one
> of the F-statistics from anova(glmodel). If there's only one
> coefficient for the term then F = t^2...

Is there an F statistic reported in anova(glmmodel)?  I had a bug in
the code that allowed such a calculation in some circumstances but
Antje Nuthmann pointed out to me  that it was incorrect and could
provoke an obscure error message.

The good news is that I changed the single-argument form of the anova
method in 0.999375-25 so the wrong results are not reported.  The bad
news is that no results are reported.  For a binomial or Poisson GLMM
I think the appropriate statistic would be a chi-square but I haven't
thought of exactly how it would be calculated.

> If I have a "glmmodel" I can do anova(glmmodel) but I wanted to make
> sure the F-statistic reported there was a sensible thing to look at
> since it wasn't quite the square of the z-statistic in the simple
> case.
>
> Maybe it doesn't have an F-distribution but it would still work well
> as a single-number stand-in for the 'size of a fixed effect' in a
> simulation...
>
> Thanks,
> D
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>