Re: [R] Robust or Sandwich estimates in lmer2
Abdus Sattar <[EMAIL PROTECTED]> asked: > > I am trying to find the robust (or sandwich) estimates of the standard > error of fixed effects parameter estimates using the package "lmer2". > Others have already pointed out that this is not implemented in lmer2. You could try a delete-n jackknife, which you would have to implement yourself. An ordinary delete-1 jackknife does not work for clustered/correlated data. This would also give you bias-corrected point estimates. I should add that, in the limited simulations I have done (for Gaussian mixed model analysis of pedigree data), the jackknife standard errors seemed a bit conservative (too big). David Duffy. -- | David Duffy (MBBS PhD) ,-_|\ | email: [EMAIL PROTECTED] ph: INT+61+7+3362-0217 fax: -0101 / * | Epidemiology Unit, Queensland Institute of Medical Research \_,-._/ | 300 Herston Rd, Brisbane, Queensland 4029, Australia GPG 4D0B994A v __ R-help@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.
Re: [R] Robust or Sandwich estimates in lmer2
Hi Doran, My interests to estimate fixed effects parameters in a mixed model (not the random effects parameters) and, in addition, in the model there are some nuisace parameters. For this estimation I am using pseudo-maximum likelihood methods. First I am estimating nuisance parameters then using these estimates in my likelihood function to get the estimates of fixed effects parameters using lmer2. Therefore, to get the correct estimates of the standard errors, I need robust (or sandwich) estiamtes of the SE. Thank you and thanks to Thomas his remarks, Sattar - Original Message From: Thomas Lumley <[EMAIL PROTECTED]> To: "Doran, Harold" <[EMAIL PROTECTED]> Cc: Sundar Dorai-Raj <[EMAIL PROTECTED]>; Abdus Sattar <[EMAIL PROTECTED]>; [EMAIL PROTECTED]; [EMAIL PROTECTED] Sent: Wednesday, September 19, 2007 1:13:20 PM Subject: Re: [R] Robust or Sandwich estimates in lmer2 On Wed, 19 Sep 2007, Doran, Harold wrote: > This has come up before and I'll again ask the question "why would you > want robust standard errors in lmer"? And I'll again answer: using lmer() does not automatically guarantee correct model specification, either for the correlation structure or for the marginal variance. -thomas Thomas LumleyAssoc. Professor, Biostatistics [EMAIL PROTECTED]University of Washington, Seattle ttp://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 [[alternative HTML version deleted]] __ R-help@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.
Re: [R] Robust or Sandwich estimates in lmer2
On Wed, 19 Sep 2007, Doran, Harold wrote: > This has come up before and I'll again ask the question "why would you > want robust standard errors in lmer"? And I'll again answer: using lmer() does not automatically guarantee correct model specification, either for the correlation structure or for the marginal variance. -thomas Thomas Lumley Assoc. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle __ R-help@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.
Re: [R] Robust or Sandwich estimates in lmer2
This has come up before and I'll again ask the question "why would you want robust standard errors in lmer"? Traditional econometric thinking suggests that there is model mispecification if OLS is used and there is a violation to the assumption of independence. So, one may still get the point estimates via OLS but then get robust standard errors. This makes sense. But, mixed models are designed to account for violations to the iid assumption via correctly specified random effects. So, if your lmer model is correctly specified, the standard errors should yield an accurate estimate of the true sampling variance. > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Sundar Dorai-Raj > Sent: Wednesday, September 19, 2007 11:42 AM > To: Abdus Sattar > Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] > Subject: Re: [R] Robust or Sandwich estimates in lmer2 > > Abdus Sattar said the following on 9/19/2007 7:03 AM: > > Dear R-Users: > > > > I am trying to find the robust (or sandwich) estimates of > the standard error of fixed effects parameter estimates using > the package "lmer2". In model-1, I used "robust=TRUE" on the > other, in model-2, I used "robust=FALSE". Both models giving > me the same estimates. So my question is, does the robust > option works in lmer2 to get the robust estimates of the > standard error? If anybody could offer me a suggestion I > would greatly appreciate it. Thank you. > > > > Model-1: > > > >> p.mle<-lmer2(ddimer~race+steroid+psi+sofa+apache + > (apache|subject), > >> method="ML", data=final, robust=TRUE, cluster="id", > weights=final$w) > >> beta=fixef(p.mle) > >> Vcov=vcov(p.mle, useScale=FALSE) > >> se=sqrt(diag(Vcov)) > >> beta > > (Intercept) race steroid psi > sofa apache > > 5.826489820 -0.001920670 -0.242040171 0.005293996 0.075468340 > > 0.009245152 > >> se > > [1] 0.108325229 0.058921371 0.055975547 0.001285687 0.018119089 > > 0.002559902 > > > > Model-2: > > > >> p.mle<-lmer2(ddimer~race+steroid+psi+sofa+apache + > (apache|subject), > >> method="ML", data=final, robust=FALSE, cluster="id", > weights=final$w) > >> beta=fixef(p.mle) > >> Vcov=vcov(p.mle, useScale=FALSE) > >> se=sqrt(diag(Vcov)) > >> beta > > (Intercept) race steroid psi > sofa apache > > 5.826489820 -0.001920670 -0.242040171 0.005293996 0.075468340 > > 0.009245152 > >> se > > [1] 0.108325229 0.058921371 0.055975547 0.001285687 0.018119089 > > 0.002559902 > > > > > > Best Regards, > > > > Sattar > > > > > > The help page to ?lmer2 in the lme4 package makes no mention > of "cluster" or "robust" arguments. To me, that would mean > these arguments are ignored. > > HTH, > > --sundar > > __ > R-help@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. > __ R-help@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.
Re: [R] Robust or Sandwich estimates in lmer2
Abdus Sattar said the following on 9/19/2007 7:03 AM: > Dear R-Users: > > I am trying to find the robust (or sandwich) estimates of the standard error > of fixed effects parameter estimates using the package "lmer2". In model-1, I > used "robust=TRUE" on the other, in model-2, I used "robust=FALSE". Both > models giving me the same estimates. So my question is, does the robust > option works in lmer2 to get the robust estimates of the standard error? If > anybody could offer me a suggestion I would greatly appreciate it. Thank you. > > > Model-1: > >> p.mle<-lmer2(ddimer~race+steroid+psi+sofa+apache + (apache|subject), >> method="ML", data=final, robust=TRUE, cluster="id", weights=final$w) >> beta=fixef(p.mle) >> Vcov=vcov(p.mle, useScale=FALSE) >> se=sqrt(diag(Vcov)) >> beta > (Intercept) race steroid psi sofa apache > 5.826489820 -0.001920670 -0.242040171 0.005293996 0.075468340 0.009245152 >> se > [1] 0.108325229 0.058921371 0.055975547 0.001285687 0.018119089 0.002559902 > > Model-2: > >> p.mle<-lmer2(ddimer~race+steroid+psi+sofa+apache + (apache|subject), >> method="ML", data=final, robust=FALSE, cluster="id", weights=final$w) >> beta=fixef(p.mle) >> Vcov=vcov(p.mle, useScale=FALSE) >> se=sqrt(diag(Vcov)) >> beta > (Intercept) race steroid psi sofa apache > 5.826489820 -0.001920670 -0.242040171 0.005293996 0.075468340 0.009245152 >> se > [1] 0.108325229 0.058921371 0.055975547 0.001285687 0.018119089 0.002559902 > > > Best Regards, > > Sattar > > The help page to ?lmer2 in the lme4 package makes no mention of "cluster" or "robust" arguments. To me, that would mean these arguments are ignored. HTH, --sundar __ R-help@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.