Skip to content

nlme:lme pseudo-replicate degrees of freedom

2 messages · Edland, Steven, Ben Bolker

Original
Edland, Steven
# 
I notice that nlme:lme (vers. 3.1.145) still has the pseudo-replicate degrees of freedom bug when applied to the usual random slopes model of longitudinal repeated measures.

Has anyone written a lmerTest equivalent wrapper for lme, or is there any other available work-around?

Thank you in advance for any advice on this issue.

Sincerely,
Steve Edland

Steven D. Edland, Ph.D.
Professor
Dept. of Family Medicine & Public Health
Dept. of Neurosciences
University of California, San Diego
9500 Gilman Dr. M/C 0948
La Jolla, CA 92093-0948
http://biostat.ucsd.edu/sedland.htm
1 day later
Ben Bolker
# 
?? Can you point to a reference that explains the issue (e.g. mailing 
list thread or ... something on the R bug tracker or ... ?)? I do know 
that there's an issue with df of random-slopes models which is 
illustrated under "Df alternatives" at 
https://bbolker.github.io/mixedmodels-misc/glmmFAQ.html#why-doesnt-lme4-display-denominator-degrees-of-freedomp-values-what-other-options-do-i-have

 ?I wrote some code (linked from the above discussion, available from 
https://bbolker.github.io/mixedmodels-misc/R/calcDenDF.R ) ; it uses the 
same kind of "semantic"/containment method, nothing with a quantitative 
basis like Satterthwaite/etc., but it does do a somewhat better job on 
random-slopes models. (I'm sure it could be improved.)
On 7/19/20 12:44 PM, Edland, Steven wrote: