[SPAM] - Re: How to assess significance of random effect in lme4 - Bayesian Filter detected spam
Actually, I re-read the post and think it needs clarification. We may both be right. If the question is "I am building a model and want to know if I should retain this random effect?" (or something like that) then the LRT should be used to compare the fitted model against another model. This would be accomplished via anova(). In other multilevel programs, the variance components are often associated with a chi-square statistic and a test statistic associated with the variance. lmer() does not report this test statistic. Now, if the question is something like "I want to know if a specific realization of the random variable (i.e., a specific empirical Bayes estimate) is different from a population value?" then one would need the posterior means. So, Shige, I hope this hasn't been confusing, but there are many things happening in these models and it is easy to get confused. Maybe if you could clarify your question. -----Original Message----- From: Spencer Graves [mailto:spencer.graves at pdf.com] Sent: Thursday, August 18, 2005 8:26 AM To: Doran, Harold Cc: Shige Song; r-help at stat.math.ethz.ch Subject: [SPAM] - Re: [R] How to assess significance of random effect in lme4 - Bayesian Filter detected spam Hi, Harold: Thanks for the clarification. I thought I had read the original post. Obviously, I had misread it. Thanks again. spencer graves
Doran, Harold wrote:
Yes, it is a different issue. ranef() extracts the empirical Bayes
estimates, which are the empirical posterior modes. The bVar slot
holds the corresponding posterior variances of these modes.
Technically, (according to D. Bates) the values in the bVar slot are
the the diagonal elements of (Z'Z+\Omega)^{-1}.
The original post was asking how to test and compare a specific random
effect, not a general assessment of how much information is provided by the data via LRT. Shige asked how to test whether a specific EB estimate is different than some other value. LRT doesn't answer this question, but the values in the bVar slot do. -----Original Message----- From: Spencer Graves [mailto:spencer.graves at pdf.com] Sent: Wed 8/17/2005 10:08 PM To: Doran, Harold Cc: Shige Song; r-help at stat.math.ethz.ch Subject: Re: [R] How to assess significance of random effect in
lme4
Is there some reason you are NOT using "anova", as in
"Examples"
section of "?lmer"?
Permit me to summarize what I know about this, and I'll be
pleased if someone else who thinks they know different would kindly
enlighten me and others who might otherwise be misled if anything I
say is inconsistent with the best literature available at the moment:
1. Doug Bates in his PhD dissertation and later in his book
with Don Watts (1988) Nonlinear Regression Analysis and Its Applications (Wiley) split approximation errors in nonlinear least squares into "intrinsic curvature" and "parameter effects curvature".
He quantified these two problems in the context of roughly three dozen
published examples, if my memory is correct, and found that in not
quite all cases, the parameter effects were at least an order of
magnitude greater than the intrinsic curvature.
2. In nonnormal situations, maximum likelihood is subject
to more approximation error -- intrinsic curvature -- than "simple"
nonlinear least squares. However, I would expect this comparison to
still be fairly accurate, even if the differences may not be quite as
stark.
3. The traditional use of "standard errors" to judge
statistical significance is subject to both intrinsic and parameter
effects errors, while likelihood ratio procedures such as anova are
subject only to the intrinsic curvature (assuming there are no
substantive problems with nonconvergence). Consequently, to judge
statistical significance of an effect, anova is usually substantially
better than the so-called Wald procedure using approximate standard
errors, and is almost never worse.
If anyone knows of a case where this is NOT true, I'd like to know.
4. With parameters at a boundary as with variance
components, the best procedure seems to double the p-value from a
nested anova (unless the reported p-value is already large). This is
because the 2*log(likelihood ratio) in such cases is roughly a 50-50
mixture of 0 and chi-square(1) [if testing only 1 variance component
parameter].
This is supported by a substantial amount of research, including simulations discussed in a chapter in Pinheiro and Bates (2000) Mixed-Effects Models in S and S-Plus (Springer). The may be more accurate procedures available in the literature, but none so simple as
this as far as I know.
Comments?
spencer graves
p.s. It looks like fm at bVars is a list containing vectors of length 29
and 6 in your example. I don't know what they are, but I don't see how they can be standard errors in the usual sense. Doran, Harold wrote:
> These are the posterior variances of the random effects (I think
more > properly termed "empirical" posteriors). Your model apparently includes > three levels of random variation (commu, bcohort, residual). The first > are the variances associated with your commu random effect and the > second are the variances
associated with the bcohort random effect.
> > Accessing either one would require > > fm at bVar$commu or fm at bVar$bcohort > > Obviously, replace "fm" with the name of your fitted model. > > -----Original Message----- > From: r-help-bounces at stat.math.ethz.ch >
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Shige Song > Sent: Wednesday, August 17, 2005 7:50 AM > To: r-help at stat.math.ethz.ch > Subject: Re: [R] How to assess significance of random effect in lme4 > > Hi Harold, > > Thanks for the reply. I looked at my outputs using str() as you > suggested,
here is the part you mentioned:
> > ..@ bVar :List of 2 > .. ..$ commu : num [1, 1, 1:29] 5e-10 5e-10 5e-10 5e-10 5e-10
...
> .. ..$ bcohort: num [1, 1, 1:6] 1.05e-05 7.45e-06 6.53e-06
8.25e-06
> 7.11e-06 ... > > where commu and bcohort are the two second-level units. Are these standard errors? Why the second vector contains a series of
different > numbers?
> > Thanks! > > Shige > > On 8/17/05, Doran, Harold <HDoran at air.org> wrote: >
>> >> >>You can extract the posterior variance of the random effect from
the >>bVar slot of the fitted lmer model. It is not a hidden option, but a >>part of the fitted model. It just doesn't show up when you use > > summary().
>
>> >> Look at the structure of your object to see what is available
using > > str().
>
>> >> However, your comment below seems to imply that it is incorrect
for >>lmer to report SDs instead of the standard error, which is not
true.
>>That is a quantity of direct interest. >> >> Other multilevel programs report the same exact statistics (for
the >>most part). For instance, HLM reports the variances as well. If
you >>want the posterior variance of an HLM model you need to extract
it.
>> >> >> >> -----Original Message----- >> From: r-help-bounces at stat.math.ethz.ch on behalf of >>Shige Song >> Sent: Wed 8/17/2005 6:30 AM >> To: r-help at stat.math.ethz.ch >> Cc: >> Subject: [R] How to assess significance of random effect in
> > lme4 >
>> >> Dear All, >> >> With kind help from several friends on the list, I am getting
close.
>> Now here are something interesting I just realized: for random effects, lmer reports standard deviation instead of standard error!
Is > > >>there a hidden option that tells lmer to report standard error of >>random effects, like most other multilevel or mixed modeling software, > > >>so that we can say something like "randome
effect for xxx is >>significant, while randome effect for xxx is not significant"? Thanks!
>> >> Best, >> Shige >> >> ______________________________________________ >> R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help >> PLEASE do read the posting guide! >>http://www.R-project.org/posting-guide.html >> >> >> >> >>
> > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list >
https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! > http://www.R-project.org/posting-guide.html > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA spencer.graves at pdf.com www.pdf.com <http://www.pdf.com> Tel: 408-938-4420 Fax: 408-280-7915
-- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA spencer.graves at pdf.com www.pdf.com <http://www.pdf.com> Tel: 408-938-4420 Fax: 408-280-7915