[R-sig-ME] lme approximation method for dfs

[Maarten Jung]

Dear Sala,

I *think* the containment method that is mentioned in the "Models
supported by emmeans" vignette is something similar to the containment
degrees of freedom approximation in SAS [1] but if you want to use
this method, you should ask the author of emmeans (Russell Lenth) to
make sure that my guess is correct.

I guess your emmeans() call throws an error because you are missing
the "specs" argument which specifies the predictor variables over
which the estimated marginal means should be calculated. Also, the
last argument is "sigmaAdjust" instead of "adjustSigma" but it
defaults to TRUE anyway. This should work (if the emmeans package is
loaded, otherwise use emmeans:emmeans()):
emmeans(lme_fit, ~ Group * TIME, mode = "appx-satterthwaite",
sigmaAdjust = TRUE)

The function joint_tests() which constructs Type-III-ANOVA-like tables
of all predictors and the function contrast() could be useful for your
analysis, too. See also the nice vignettes here [2].

[1]  https://documentation.sas.com/?docsetId=statug&docsetTarget=statug_glimmix_details38.htm&docsetVersion=15.1&locale=en
[2]  https://cran.r-project.org/web/packages/emmeans/vignettes/

Best,
Maarten


On Sun, May 24, 2020 at 4:55 AM Salahadin Lotfi
<salahadin.lotfi using gmail.com> wrote:
>
> Wow! Such a thorough answer, Phillip. Thanks very much.
> I misspoken when I said the lmer would generate p values for fixed effects (yes it is indeed lmerTest package which does it).
> Just to make sure I understand it correctly:
> lme uses ML/REML to approximate beta weights of fixed effects,
> it uses Wald to approximate t/z values (fixed effects), and
> it uses "inner-outer" rule to get denominator of df (ddf) which is apparently the least optimal method to estimate ddf among other methods (etc., Satterthwaite's). Lastly, is this "inner-outer" rule the same as "containment method"?
>
> Maarten,
> First of all thank you for the nice pointer. The emmeans is elegant. But, I could not get it to work. Could you point out what I am missing in the emmeans function?
>   Here is my code:
> ## Group and Time are categorical IVs with three levels.
> lme_fit<- lme(DV~  factor (Group) * factor(TIME), random = ~ TIME | SubjectID, # factor(Group) * factor(TIME)
>                   correlation = corSymm (form = ~ 1 | SubjectID),
>                   weights = varIdent (form = ~ 1 | TIME),
>                   data = my_data,
>                   method= "REML",
>                   na.action = "na.omit")
> emmeans::emmeans(lme.fit, mode= "appx-satterthwaite", adjustSigma = TRUE) ## NOT working
>
>
> I appreciate for taking time and replying.
> Sala
>
> *************
> Salahadin (Sala) Lotfi
>
> PhD Candidate of Cognitive Neuroscience
>
> University of Wisconsin-Milwaukee
>
> Anxiety Disorders Laboratory
>
> President, Association of Clinical and Cognitive Neuroscience, UWM
>
>
> On Sat, May 23, 2020 at 5:01 PM Maarten Jung <Maarten.Jung using mailbox.tu-dresden.de> wrote:
>>
>> Dear Sala,
>>
>> As Phillip Alday explained, there is no implementation of the Satterthwaite approximation in the nlme package. If you want to stick with this package, the only way I know to get something similar (for lme objects) is to use functions of the emmeans package with the argument "mode" set to "appx-satterthwaite" (see [1]).
>>
>> [1] https://cran.r-project.org/web/packages/emmeans/vignettes/models.html#K
>>
>> Best,
>> Maarten
>>
>>
>> On Sat, 23 May 2020, 17:07 Phillip Alday <phillip.alday using mpi.nl> wrote:
>>>
>>> Hi,
>>>
>>> On 23/5/20 9:11 am, Salahadin Lotfi wrote:
>>> > Dear all,
>>> > I have a very simple question but, have been having a hard time to figure
>>> > it out.
>>> > I am using a mixed model with random intercept and slope using lme function
>>> > with an unstructured covariance matrix. I know lmer uses Satterthwaite's
>>> > approximation method to approximate dfs of fixed effects,
>>>
>>> This is not accurate. lme4 by default doesn't even try to figure out the
>>> df and doesn't report p-values. The lmerTest package adds in options to
>>> use Satterthwaite or Kenward-Roger approximations for p-values, but
>>> depending on who you ask around here, the sentiment for those
>>> approximations ranges from "of course" to "hmrpf, why would you bother?"
>>> to "the heretics must be purged!".
>>>
>>> The GLMM FAQ (https://bbolker.github.io/mixedmodels-misc/glmmFAQ.html)
>>> has some info on each of these, but I'll copy and paste something
>>> relevant that I wrote on a different mailing list:
>>>
>>> Treating the t-values as z-values is as reasonable as using the
>>> t-distribution with some estimated degrees of freedom for studies with
>>> 20-30 subjects and 10s of observations per condition per subject for two
>>> reasons. One is that a t-distribution with dozens of degrees of freedom
>>> is essentially a normal distribution, and so even if you could figure
>>> out what the "right" number of degrees of freedom were, it wouldn't be
>>> far off from the number you get from the normal distribution. The other
>>> reason is that none of these asymptotic results are guaranteed to be
>>> particularly great for anything other than very well behaved linear
>>> mixed models, which is why things like parametric bootstrap are the gold
>>> standard for figuring out coverage intervals. And for large models,
>>> bootstrapping is about as fast as KR (because KR as implemented in
>>> pbkrmodcomp, which lmerTest depends on, computes the inverse of a large
>>> n x matrix).
>>>
>>> > but I am not sure
>>> > what is the preferred method that lme uses. Is it Wald or Likelihood ratio?
>>>
>>> Wald and likelihood ratio are not degrees of freedom estimates. The
>>> likelihood-ratio tests do have a df, which corresponds to the difference
>>> in the number of free parameters between the models, but this not the
>>> relevant df. (It's numerator degrees of freedom in the ANOVA framework,
>>> while what you need are the denominator degrees of freedom.) The Wald
>>> tests are just the things you see in the table of the fixed effects,
>>> i.e. the tests corresponding to the t- or z-values (or more generally
>>> the ANOVA-style tests / tests of linear hypotheses you then construct
>>> from the fixed effects).
>>>
>>>
>>> > I don't think lme offers such an option to specify an approximation method
>>> > for dfs of fixed effects. Does it?
>>>
>>> The dfs in nlme are computed using the "inner-outer" rule which doesn't
>>> work well for many types of designs common in cognitive neuroscience.
>>> More information on this is in the GLMM FAQ, search for "Df
>>> alternatives" on that page.
>>>
>>>
>>> Hope that helps!
>>>
>>> Phillip
>>>
>>>
>>> >
>>> > I appreciate any response in advance.
>>> > Sala
>>> >
>>> >
>>> > *************
>>> > Salahadin (Sala) Lotfi
>>> >
>>> > PhD Candidate of Cognitive Neuroscience
>>> >
>>> > University of Wisconsin-Milwaukee
>>> >
>>> > Anxiety Disorders Laboratory
>>> >
>>> > President, Association of Clinical and Cognitive Neuroscience, UWM
>>> >
>>> >       [[alternative HTML version deleted]]
>>> >
>>> > _______________________________________________
>>> > R-sig-mixed-models using r-project.org mailing list
>>> > https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>>>
>>> _______________________________________________
>>> R-sig-mixed-models using r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models