[R] dichotomous independent variables in binomial glmm model

[Andrew Perrin]

This question is a followup on the one I posted Friday
(http://www.r-project.org/nocvs/mail/r-help/2002/4485.html).

I believe I have narrowed the problem to the observation that models fail
to converge when they have dichotomous predictors; to wit:

glmmPQL(random = ~1 | groupid/participantid,
                       fixed=r.logic.morality ~ realage,
                       data=fgdata.10statements.df,
                       na.action=na.omit,
                       niter=50,
                       family=binomial)

works fine, but

glmmPQL(random = ~1 | groupid/participantid,
                       fixed=r.logic.morality ~ is.prot
                       data=fgdata.10statements.df,
                       na.action=na.omit,
                       niter=50,
                       family=binomial)

fails with:

iteration 1 
Error in MEEM(object, conLin, control$niterEM) : 
        Singularity in backsolve at level 0, block 1


This makes sense to me, as a dichotomous variable could be expected to
provide insufficient variability in a transformed matrix.

Unfortunately, though, I need to use dummy variables as fixed effects. Any
advice?

Thanks.

----------------------------------------------------------------------
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U of North Carolina, Chapel Hill
clists at perrin.socsci.unc.edu * andrew_perrin (at) unc.edu


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._