R help - Sep 2007 - Getting intervals for within-group standard errors for each group using nlme and varIdent

I am using lme from the nlme package to fit a mixed model.  We have observations
nested in patients(encounters) and patients nested in groups (2 different
treatments).  We are interested in the differences between the 2 groups, both
the means and the standard deviations (are patients in group A less variable
than those in group B? both within patient and between patient within group).
 
Here is the call from my lme object:
 
> fit3$call
lme.formula(fixed = BG ~ group, data = bg, random = list(Encounter =
pdDiag(~group -
    1)), weights = varIdent(form = ~1 | group))

then I call the intervals function:
 
Approximate 95% confidence intervals
 Fixed effects:
                lower      est.     upper
(Intercept) 123.40005 126.88445 130.36885
groupB       17.44991  23.60049  29.75107
attr(,"label")
[1] "Fixed effects:"
 Random Effects:
  Level: Encounter 
              lower     est.    upper
sd(groupA) 16.62769 19.13351 22.01698
sd(groupB) 24.01774 27.63703 31.80171
 Variance function:
      lower      est.     upper
A 0.8270135 0.8534733 0.8807797
attr(,"label")
[1] "Variance function:"
 Within-group standard error:
   lower     est.    upper 
33.56082 34.29237 35.03987 

and this gives me most of what I want, the only exception is that I (actully my
client) want the interval for the Within-group standard error of group A (the
last interval shows it for group B and the Variance function shows the interval
for their ratio).
 
Any suggestions?
 
Thanks,

	[[alternative HTML version deleted]]

Just to follow-up on this for the archive and anyone else who is
interested in this problem.  I ended up solving this by running lme
seperately on subsets of the data (one for group=='A' and one for
group=='B'), the intervals for the statistics computed both ways were
within rounding of each other, so I felt confident in using the standard
deviation estimate from the "A" only subset group.


-- 
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at intermountainmail.org
(801) 408-8111
 
 
> -----Original Message-----
> From: r-help-bounces at r-project.org 
> [mailto:r-help-bounces at r-project.org] On Behalf Of Greg Snow
> Sent: Thursday, September 27, 2007 12:20 PM
> To: r-help at r-project.org
> Subject: [R] Getting intervals for within-group standard 
> errors for each group using nlme and varIdent
> 
> I am using lme from the nlme package to fit a mixed model.  
> We have observations nested in patients(encounters) and 
> patients nested in groups (2 different treatments).  We are 
> interested in the differences between the 2 groups, both the 
> means and the standard deviations (are patients in group A 
> less variable than those in group B? both within patient and 
> between patient within group).
>  
> Here is the call from my lme object:
>  
> > fit3$call
> lme.formula(fixed = BG ~ group, data = bg, random = 
> list(Encounter = pdDiag(~group - 
>     1)), weights = varIdent(form = ~1 | group))
> 
> then I call the intervals function:
>  
> Approximate 95% confidence intervals
>  Fixed effects:
>                 lower      est.     upper
> (Intercept) 123.40005 126.88445 130.36885
> groupB       17.44991  23.60049  29.75107
> attr(,"label")
> [1] "Fixed effects:"
>  Random Effects:
>   Level: Encounter 
>               lower     est.    upper
> sd(groupA) 16.62769 19.13351 22.01698
> sd(groupB) 24.01774 27.63703 31.80171
>  Variance function:
>       lower      est.     upper
> A 0.8270135 0.8534733 0.8807797
> attr(,"label")
> [1] "Variance function:"
>  Within-group standard error:
>    lower     est.    upper 
> 33.56082 34.29237 35.03987 
> 
> and this gives me most of what I want, the only exception is 
> that I (actully my client) want the interval for the 
> Within-group standard error of group A (the last interval 
> shows it for group B and the Variance function shows the 
> interval for their ratio).
>  
> Any suggestions?
>  
> Thanks,
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>