RE: st: RE: LONEWAY & XTREG

Dear Al,

Thanks for your reply.

When you indicate that "the dependent variable can be any numeric variable"
for -loneway-, do you mean continuous variables, right? How to examine 
variance
components for binomial/ordered/multinomial variables such as occupational
group then?

Best

Lijun

Quoting "Feiveson, Alan H. (JSC-SK311)" <alan.h.feiveson@nasa.gov>:


> -loneway- works only with first and second moments of the data and thus
> will unbiasedly estimate between and within variance components,
> regardless of the distribution of the data (as long as it it numeric).
> The dependent variable can be any numeric variable.
>
> -xtreg- assumes the within- and between-class errors come from normal
> distributions and estimates their variances by maximum likelihood with a
> GLS regression model. So if the data is not normally distributed, you
> can get completely different results and -xtreg- is not to be believed.
> However in this case the value of knowing values of variance components
> is questionable. On the other hand, if the data really are normal,
> -xtreg- should be more efficient than -loneway-.
>
> Al Feiveson
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Lijun Song
> Sent: Friday, December 01, 2006 4:12 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: RE: LONEWAY & XTREG
>
> Hi Nick
>
> Thank you for your information. I already tried them. But I failed to
> find any information about variable types for loneway. Does STATA set it
> as a default that LONEWAY could be used for all kinds of variables?
>
> Lijun
>
> On 1 Dec 2006 at 21:27, Nick Cox wrote:
>
> > If you ask Stata itself what is available, you will see other
> possibilities.
> > . search variance components
> >
> > Keyword search
> >
> >         Keywords:  variance components
> >           Search:  (1) Official help files, FAQs, Examples, SJs, and
> > STBs
> >
> > Search of official help files, FAQs, Examples, SJs, and STBs
> >
> >
> > [R]     loneway . . . . . Large one-way ANOVA, random effects, and
> reliability
> >         (help loneway)
> >
> > [R]     regress postestimation  . . . . . . . Postestimation tools for
> regress
> >         (help regress postestimation)
> >
> > Example . . . . . . . . . . . . . . .  Stata web books:  Regression
> with Stata
> >         . .  Chen, Ender, Mitchell & Wells (UCLA Academic Technology
> Services)
> >         7/06    web book Regression with Stata by (in alphabetical
> >                 order) Xiao Chen, Philip B. Ender, Michael Mitchell
> >                 & Christine Wells
> >                 http://www.ats.ucla.edu/stat/stata/webbooks/reg/
> >
> > SJ-6-1  st0095  . . . . . . . . . . .  Estimating variance components
> in Stata
> >         . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Y.
> Marchenko
> >         Q1/06   SJ 6(1):1--21                                    (no
> commands)
> >         describes using xtmixed to estimate variance components
> >         in linear models
> >
> > SJ-6-1  gn0031  . . Review of Multilevel and Longitudinal Modeling
> Using Stata
> >         . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> R. Wolfe
> >         Q1/06   SJ 6(1):138--143                                 (no
> commands)
> >         book review of Multilevel and Longitudinal Modeling
> >         Using Stata by Rabe-Hesketh and Skrondal
> >
> > SJ-4-4  st0077  . . CIs for the variance comp. of random-effects
> linear models
> >         (help xtvc if installed)  . . . . . . . . . .  M. Bottai and
> N. Orsini
> >         Q4/04   SJ 4(4):429--435
> >         confidence intervals for the variance components of
> >         random-effects linear regression models.
> >
> > STB-60  sg160 . . . . . . . . . . . . On boundary-value
> likelihood-ratio tests
> >         . . . . . . . . . . . .  R. G. Gutierrez, S. Carter, and D. M.
> Drukker
> >         3/01    pp.15--18; STB Reprints Vol 10, pp.269--273      (no
> commands)
> >         discusses likelihood-ratio boundary tests (such as tests for
> >         the presence of overdispersion or random effects) which are
> >         based on a mixture of a point mass at zero and a chi-squared
> >         distribution
> >
> > Nick
> > n.j.cox@durham.ac.uk
> >
> > Lijun Song
> >
> > > I am running variance component models using both loneway and xtreg.
> > >
> > > Could we use both of them to deal with all kinds of variables
> > > including continuous, categorical and discrete dependent variables?
> > >
> > > I have this question because when we run random intercept/slope
> > > models, we can only use xtreg to analyze Continuous outcomes but use
> > > gllamm to analyze categorial or discrete variables, right?
> >
> > *
> > *   For searches and help try:
> > *   http://www.stata.com/support/faqs/res/findit.html
> > *   http://www.stata.com/support/statalist/faq
> > *   http://www.ats.ucla.edu/stat/stata/
>
> ------------------------------------------------------------------------
> ------------------
> Lijun Song
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/