Hello,
>It seems that the newest version of lme4 (0.999999-0) now provides P values
in the summary tables.
Nope, they were only removed for Gaussian models, and still are, try:
fm1 <- lmer(Reaction ~ Days + (Days|Subject), sleepstudy)
summary(fm1)
#other attached packages:
#[1] lme4_0.999999-0
whereas they were never removed for binomial/poisson GLMMs.
Cheers,
Luca
-----Original Message-----
From: Christoph Scherber
To: "r-sig-mixed-models@r-project.org"
Date: Thu, 12 Jul 2012 14:09:36 +0200
Subject: [R-sig-ME] P values in summary tables from lmer objects
Dear all,
It seems that the newest version of lme4 (0.999999-0) now provides P values
in the summary tables.
Given the many previous discussions about the derivation of P values in
lmer, I was just wondering
how "safe" it would now be to use these values (and how they´re
calculated). I guess they´re just
taken internally from a call to glm?
Example:
##
gm1 <- glmer(cbind(incidence, size - incidence) ~ period + (1 | herd),
family = binomial, data = cbpp)
summary(gm1)
coef(summary(gm1))
##
Many thanks,
Christoph
[running R 2.15.1 on Windows 7 32-Bit]
--
PD Dr Christoph Scherber
Georg-August University Goettingen
Department of Crop Science
Agroecology
Grisebachstrasse 6
D-37077 Goettingen
Germany
phone 0049 (0)551 39 8807
fax 0049 (0)551 39 8806
http://www.gwdg.de/~cscherb1
[[alternative HTML version deleted]]