Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: GLLAMM versus xtmixed


From   martina brandt <[email protected]>
To   [email protected]
Subject   Re: st: GLLAMM versus xtmixed
Date   Mon, 16 Apr 2007 19:56:38 +0200

dear richard, thanks for the hint!

the code that seems to work is:
xi: xtmixed lhstdwoe ce_gesunde_ ce_altere_ both_ ce_spheu endsmeetex i.edu=
kat religion ce_n_kidsk ce_ngeschwk muto vato vaso ce_nentfe_ fineep_ ce_er=
bez ce_isicn if e(sample) ||land: ||hhnr: ||beobnr: , ml variance
i.edukat _Iedukat_1-3 (naturally coded; _Iedukat_1 omitted)

Performing EM optimization:

Performing gradient-based optimization:

Iteration 0: log likelihood =3D -1938.2228
Iteration 1: log likelihood =3D -1934.4078
Iteration 2: log likelihood =3D -1934.305
Iteration 3: log likelihood =3D -1934.3013
Iteration 4: log likelihood =3D -1934.3013

Computing standard errors:

Mixed-effects ML regression Number of obs =3D 1134

-----------------------------------------------------------
| No. of Observations per Group
Group Variable | Groups Minimum Average Maximum
----------------+------------------------------------------
land | 10 20 113.4 278
hhnr | 1039 1 1.1 3
beobnr | 1083 1 1.0 2
-----------------------------------------------------------

Wald chi2(17) =3D 176.35
Log likelihood =3D -1934.3013 Prob > chi2 =3D 0.0000

---------------------------------------------------------------------------=
---
lhstdwoe | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+-------------------------------------------------------------=
---
ce_gesunde_ | .1135772 .0402843 2.82 0.005 .0346214 .1925329
ce_altere_ | .033269 .0075909 4.38 0.000 .0183911 .048147
both_ | -.0755805 .1062736 -0.71 0.477 -.2838728 .1327119
ce_spheu | .2184644 .0555883 3.93 0.000 .1095134 .3274154
endsmeetex | -.1923056 .1115012 -1.72 0.085 -.410844 .0262327
_Iedukat_2 | -.1145846 .1080963 -1.06 0.289 -.3264494 .0972803
_Iedukat_3 | -.1320016 .1165883 -1.13 0.258 -.3605105 .0965073
religion | -.1790786 .1399235 -1.28 0.201 -.4533235 .0951664
ce_n_kidsk | -.0440056 .0362299 -1.21 0.225 -.1150149 .0270036
ce_ngeschwk | -.0629339 .0253449 -2.48 0.013 -.1126089 -.0132589
muto | .3857507 .0897365 4.30 0.000 .2098705 .561631
vato | .1961351 .1258618 1.56 0.119 -.0505496 .4428197
vaso | -.4357497 .1249964 -3.49 0.000 -.6807381 -.1907612
ce_nentfe_ | -.2243865 .0381961 -5.87 0.000 -.2992495 -.1495235
fineep_ | -.0709821 .1534461 -0.46 0.644 -.3717309 .2297666
ce_erbez | -.0105379 .0124366 -0.85 0.397 -.0349131 .0138373
ce_isicn | -.0562341 .0170115 -3.31 0.001 -.0895761 -.0228922
_cons | .497216 .1540194 3.23 0.001 .1953435 .7990885
---------------------------------------------------------------------------=
---

---------------------------------------------------------------------------=
---
Random-effects Parameters | Estimate Std. Err. [95% Conf. Interval]
-----------------------------+---------------------------------------------=
---
land: Identity |
var(_cons) | .0377874 .0315882 .0073416 .1944934
-----------------------------+---------------------------------------------=
---
hhnr: Identity |
var(_cons) | .4514975 .285685 .1306345 1.560461
-----------------------------+---------------------------------------------=
---
beobnr: Identity |
var(_cons) | 1.116612 .290321 .6707928 1.858729
-----------------------------+---------------------------------------------=
---
var(Residual) | .2948126 .0578313 .2007113 .4330322
---------------------------------------------------------------------------=
---
LR test vs. linear regression: chi2(3) =3D 77.25 Prob > chi2 =3D 0.0000

Note: LR test is conservative and provided only for reference

and with gllamm (default link/family) i get:

xi: gllamm lhstdwoe ce_gesunde_ ce_altere_ both_ ce_spheu endsmeetex i.eduk=
at religion ce_n_kidsk ce_ngeschwk muto vato vaso ce_nentfe_ fineep_ ce_erb=
ez ce_isicn if e(sample), i(beobnr hhnr land) nip(10) adapt robust
i.edukat _Iedukat_1-3 (naturally coded; _Iedukat_1 omitted)
insufficient observations
r(2001);


any ideas???

THANKS! martina


> -----Urspr=FCngliche Nachricht-----
> Von: [email protected]
> Gesendet: 14.04.07 20:44:24
> An: [email protected]
> Betreff: Re: st: GLLAMM versus xtmixed


> At 01:10 PM 4/14/2007, martina brandt wrote:
>>can anyone imagine why gllamm (adapt) should not be able to estimate
>>a linear multilevel model with 4 levels and 1100 1.level, 700 2.
>>level, 500 3. level and 10 4.level observations , error:
>>"insufficient observations" - whereas with xtmixed (ml) it works
>>without problems?
>>
>>thanks for helping! martina
>
> It is always possible your code is wrong.  Why don't you post the
> xtmixed code that works, and the gllamm code (with error message)
> that does not?
>
>
> -------------------------------------------
> Richard Williams, Notre Dame Dept of Sociology
> OFFICE: (574)631-6668, (574)631-6463
> HOME:   (574)289-5227
> EMAIL:  [email protected]
> WWW:    http://www.nd.edu/~rwilliam
>
> *
> *   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index