Dear all,
I have a couple of questions concerning glmm in the package repeated.
After fitting a binomial glmm, summary(model) gives a very different
result from anova(model, test="Chisq"). Why is that? Are both reliable
for glmm, or is it something with my data/model? The anova call produces
a warning message ("non-integer #successes in a binomial glm"). Also,
what does the sd term represent? The calls and outputs follow below. (R
1.2.1 on SuSE Linux 6.3, repeated 0.9, rmutil 0.9).
I would appreciate your help,
Hakan Haggstrom
> ksvm2 #The data frame
no yes Pop Prey Poptype
1 2 17 M C NoPA
2 8 11 S C NoPA
3 11 9 M PA NoPA
4 9 10 S PA NoPA
5 14 5 K C PA
6 8 10 V C PA
7 13 10 K PA PA
8 10 10 V PA PA
>
> ksvm2.glmm <- glmm(cbind(yes,no) ~ Prey*Poptype, nest=Pop,
family=binomial, data=ksvm2)
> summary(ksvm2.glmm)
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.03026 0.36810 2.799 0.00513 **
PreyPA -1.08177 0.48806 -2.216 0.02666 *
PoptypePA -1.41323 0.49769 -2.840 0.00452 **
sd 0.05281 0.16541 0.319 0.74952
PreyPA:PoptypePA 1.32561 0.66631 1.989 0.04665 *
> anova(ksvm2.glmm,test="Chisq")
Analysis of Deviance Table
Df Deviance Resid. Df Resid. Dev P(>|Chi|)
NULL 7 19.0232
Prey -71 1.5022 78 17.5210 1.0000
Poptype 1 4.6068 77 12.9143 0.0318
sd 1 0.0958 76 12.8185 0.7569
Prey:Poptype 73 3.9771 3 8.8414 1.0000
Warning messages:
1: non-integer #successes in a binomial glm! in: eval(expr, envir, enclos)
2: non-integer #successes in a binomial glm! in: eval(expr, envir, enclos)
3: non-integer #successes in a binomial glm! in: eval(expr, envir, enclos)
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._