Message-ID: <DB5PR03MB1205FF0188DA8D5F9327D25FC7010@DB5PR03MB1205.eurprd03.prod.outlook.com>
Date: 2016-07-29T20:32:05Z
From: xavier piulachs
Subject: Random slope does not improve hierarchical fitting across time
In-Reply-To: <5dec4b95-5ce6-2bda-7699-664d5630130d@gmail.com>
Dear Ben,
First of all, many thanks for your quick response. Moreover, I'm aware that you are an expertise in this field, so I'm doubly happy of receiving your comments.
I have two doubts about what you say (the clue point is maybe the first):
1) The effect of time is in the model as fixed effect (and it is significant), ok. But I also would expect that each subject, i = 1,...,n, has:
a) His underlying baseline level (ie, a subject-specific baseline effect = beta0 + random intercept = beta0 + ui0) ,and
b) A particular trend-evolution across time (a subject especific slope = fixed effect of time + random slope = beta_t + uit).
It is indeed very common when dealing repeated measurements across time (a particular case of longitudinal models) to have these two significant
effects. In fact, always I have fitted longitudinal measurements over time (with unstructure matrix correlation by default), I got that random intercept
and slope model improves the accuracy of considering a single random intercept. So I think this is compatible with the idea of an autoregressive model.
Is it correct?
2) I have fitted the GLMM with the option corStruct = "full":
glmmADMB.0.int.NB <- glmmadmb(claimyr ~ obstime + (1|ID), corStruct = "full", data = tr.j, family = "nbinom")
And I get the following R error message:
Parameters were estimated, but standard errors were not: the most likely problem is that the curvature at MLE was zero or negative
The function maximizer failed (couldn't find parameter file)
Best,
Xavier
________________________________
De: R-sig-mixed-models <r-sig-mixed-models-bounces at r-project.org> en nombre de Ben Bolker <bbolker at gmail.com>
Enviado: viernes, 29 de julio de 2016 19:48
Para: r-sig-mixed-models at r-project.org
Asunto: Re: [R-sig-ME] Random slope does not improve hierarchical fitting across time
On 16-07-29 03:09 PM, xavier piulachs wrote:
> Dear members of mixed-models list,
>
>
> I'm adressing you in order to ask a question about Hierarchical and ZI counts measured over time.
> To have preliminar results, I'm modeling longitudinal data with a Negative Binomial GLMM, via
> lme4 and glmmADBM packages (very similar results). I have considered two possibilities:
>
>
> 1) A single random intercept:
> glmer.0.int.NB <- glmer.nb(counts ~ obstime + (1|id), data = tr.j) # lme4 package
>
> tr.j$ID <- as.factor(tr.j$id)
> glmmADMB.0.int.NB <- glmmadmb(claimyr ~ obstime + (1|ID), data = tr.j, family = "nbinom")
> Estimate Std. Error z value Pr(>|z|)
> (Intercept) -0.9652 0.1222 -7.9005 0.0000
> obstime 0.0238 0.0073 3.2735 0.0011
>
>
> 2) Random intercept and random slope effects:
> glmer.0.slp.NB <- glmer.nb(counts ~ obstime + (obstime|id), data = tr.j) # lme4 package
>
> glmmADMB.0.slp.NB <- glmmadmb(claimyr ~ obstime + (obstime|ID), data = tr.j, family = "nbinom")
> Estimate Std. Error z value Pr(>|z|)
> (Intercept) -0.9401 0.1190 -7.9005 0.0000
> obstime 0.0230 0.0075 3.0540 0.0023
>
>
> Surprisingly, the anova test indicates non significant improvement by fitting second model:
>
> anova(glmer.0.int.NB, glmer.0.slp.NB) # LRT: p-value = 0.2725 > 0.05
> anova(glmmADMB.0.int.NB, glmmADMB.0.slp.NB) # LRT: p-value = 0.1042 > 0.05
>
>
> As far as I know, when dealing repeated measurements across time, we expect that outcomes closer in time to be
> more correlated (it is indeed a more realistic approach), so I'm totally disconcerted by this result.
> Can anyone explain what could be the reason?
A few comments:
- most important: just because an effect is 'really' in the model (e.g.,
in this case, the effect of time really does vary among individuals)
doesn't mean it will have a statistically significant effect. In most
observational/complex fields (population biology, social sciences),
*all* of the effects are really non-zero. The purpose of significance
tests is to see which effects can be distinguished from noise.
- your explanation ("outcomes closer in time are more correlated") isn't
a very precise description of what the (obstime|ID) term in the model is
doing. Your description is of an autoregressive model; the (obstime|ID)
model is a random-slope model (slopes with respect to time vary among
individuals). You might want to check out the glmmTMB package for
autoregressive models ...
- glmmADMB's default correlation structure is diagonal, glmer.nb's is
unstructured; if you use (obstime||ID) in glmer.nb or corStruct="full"
in glmmadmb you should get more similar results (I would generally
recommend "full" as the default ...)
- likelihood ratio tests (which is what anova() is doing) generally give
conservative p-values when applied to random-effect variances (boundary
issues -- see http://tinyurl.com/glmmFAQ.html or Bolker (2009) or
Pinheiro and Bates 2000 for more discussion) -- so the p-values should
probably be approximately halved
_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
[[alternative HTML version deleted]]