R help - Jan 2012 - Is using glht with "Tukey" for lme post-hoc comparisons an appropriate substitute to TukeyHSD?

Hello,

I am trying to determine the most appropriate way to run post-hoc  
comparisons on my lme model.  I had originally planned to use Tukey  
HSD method as I am interested in all possible comparisons between my  
treatment levels.  TukeyHSD, however, does not work with lme.  The  
only other code that I was able to find, and which also seems to be  
widely used, is glht specified with Tukey:

summary(glht(model, linfct=mcp(Treatment="Tukey")))

Out of curiosity, I ran TukeyHSD and the glht code for a simple ANOVA  
and found that they had quite different p-values.  If the glht code is  
not running TukeyHSD, what does the "Tukey" in the code specify?  Is  
using glht code appropriate if I am interested in a substitute for  
TukeyHSD?  Are there any other options for multiple comparisons for  
lme models?  I am really interested in knowing if the Tukey contrasts  
generated from the glht code is providing me with appropriate p-values  
for my post-hoc comparisons.

I feel like I have reached a dead end and am not sure where else to  
look for information on this issue. I would be grateful for any  
feedback on this matter.


Anne Cheverie
M.Sc. Candidate
Dalhousie University

glht is probably what you should be using.  Both TukeyHSD and glht give
essesntially identical confidence intervals for
the example in ?glht.  What aren't you satisfied with?

amod <- aov(breaks ~ tension, data = warpbreaks)
confint(glht(amod, linfct = mcp(tension = "Tukey")))
TukeyHSD(amod)
On Mon, Jan 2, 2012 at 6:19 PM, Anne Aubut <an438516@dal.ca> wrote:
> Hello,
>
> I am trying to determine the most appropriate way to run post-hoc
> comparisons on my lme model.  I had originally planned to use Tukey HSD
> method as I am interested in all possible comparisons between my treatment
> levels.  TukeyHSD, however, does not work with lme.  The only other code
> that I was able to find, and which also seems to be widely used, is glht
> specified with Tukey:
>
> summary(glht(model, linfct=mcp(Treatment="Tukey"))**)
>
> Out of curiosity, I ran TukeyHSD and the glht code for a simple ANOVA and
> found that they had quite different p-values.  If the glht code is not
> running TukeyHSD, what does the "Tukey" in the code specify?  Is
using glht
> code appropriate if I am interested in a substitute for TukeyHSD?  Are
> there any other options for multiple comparisons for lme models?  I am
> really interested in knowing if the Tukey contrasts generated from the glht
> code is providing me with appropriate p-values for my post-hoc comparisons.
>
> I feel like I have reached a dead end and am not sure where else to look
> for information on this issue. I would be grateful for any feedback on this
> matter.
>
>
> Anne Cheverie
> M.Sc. Candidate
> Dalhousie University
>
> ______________________________**________________
> R-help@r-project.org mailing list
>
https://stat.ethz.ch/mailman/**listinfo/r-help<https://stat.ethz.ch/mailman/listinfo/r-help>
> PLEASE do read the posting guide http://www.R-project.org/**
> posting-guide.html <http://www.r-project.org/posting-guide.html>
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]