Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: Multinomial Logit Model - Cramer Ridder Test


From   Nick Cox <n.j.cox@durham.ac.uk>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Multinomial Logit Model - Cramer Ridder Test
Date   Mon, 14 Nov 2011 10:35:34 +0000

-crtest- as cited here long predates the introduction of factor variables in Stata. But you are not explicitly using factor variable notation. So it could well be something else. I'd

. set trace on
. set traced 1 
. crtest 

-- so you can report what is issuing the error. 

Nick 
n.j.cox@durham.ac.uk 

Muhammad Anees

try without the -if year == 2009- as the help file suggests -crtest-
"performs the test for pooling states" after the Multinomial logit
models. The error suggest to not use time operators and factor
variables. Also read the -help crtest- file for more information.

On Mon, Nov 14, 2011 at 11:18 AM, Ian Li <ianz.magic@hotmail.com> wrote:

> I am using Stata 11.2 on a Windows OS.
>
> I am estimating a multinomial logit model with 7 categorical outcomes. As the estimated log odds ratio and marginal effects for some of the outcomes (e.g.) are insignificant in statistical and economic terms, I would like to do a Cramer Ridder test (see Cramer J.S. and Ridder G. (1991), "Pooling States in the Multinomial Logit Model", Journal of Econometrics, 47, 267-272).
>
> There is a user-written program for this test:
>
> package crtest from http://fmwww.bc.edu/RePEc/bocode/c
>
>       Distribution-Date: 20031203
>
>       Author:  Joao Pedro Azevedo, University of Newcastle-upon-Tyne, UK
>       Support: email j.p.azevedo@ncl.ac.uk
>
> However, when I estimate the logit model and use the test I get the following output and error message:
>
>
> . mlogit overed  female noneng nonaust if year == 2009, vce (robust)
>
> Iteration 0:   log pseudolikelihood = -109594.56
> Iteration 1:   log pseudolikelihood = -107547.29
> Iteration 2:   log pseudolikelihood = -106989.03
> Iteration 3:   log pseudolikelihood = -106976.66
> Iteration 4:   log pseudolikelihood = -106976.62
> Iteration 5:   log pseudolikelihood = -106976.62
>
> Multinomial logistic regression                   Number of obs   =      68352
>                                                   Wald chi2(18)   =    5211.03
>                                                   Prob > chi2     =     0.0000
> Log pseudolikelihood = -106976.62                 Pseudo R2       =     0.0239
>
> ------------------------------------------------------------------------------
>              |               Robust
>       overed |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
> -------------+----------------------------------------------------------------
> 1            |  (base outcome)
> -------------+----------------------------------------------------------------
> 2            |
>       female |   .0054676   .0238051     0.23   0.818    -.0411895    .0521248
>       noneng |  -.1713411    .036681    -4.67   0.000    -.2432346   -.0994477
>      nonaust |  -.4023007   .0666538    -6.04   0.000    -.5329397   -.2716617
>        _cons |  -.6301703   .0196192   -32.12   0.000    -.6686234   -.5917173
> -------------+----------------------------------------------------------------
> 3            |
>       female |  -.2061721   .0220245    -9.36   0.000    -.2493394   -.1630049
>       noneng |   .2473055   .0309354     7.99   0.000     .1866732    .3079377
>      nonaust |   .7666135   .0449734    17.05   0.000     .6784672    .8547598
>        _cons |  -.4356624   .0180227   -24.17   0.000    -.4709862   -.4003386
> -------------+----------------------------------------------------------------
> 4            |
>       female |  -.1237307   .0222175    -5.57   0.000    -.1672762   -.0801853
>       noneng |   .1045945   .0315106     3.32   0.001     .0428348    .1663542
>      nonaust |   .6037373   .0464335    13.00   0.000     .5127293    .6947453
>        _cons |  -.4583203   .0183183   -25.02   0.000    -.4942236    -.422417
> -------------+----------------------------------------------------------------
> 5            |
>       female |  -.2056775   .0364563    -5.64   0.000    -.2771306   -.1342244
>       noneng |   .2566406   .0474816     5.41   0.000     .1635783    .3497029
>      nonaust |   1.251198   .0592131    21.13   0.000     1.135142    1.367253
>        _cons |  -1.871285   .0300228   -62.33   0.000    -1.930128   -1.812441
> -------------+----------------------------------------------------------------
> 6            |
>       female |  -.4014462   .0451968    -8.88   0.000    -.4900303   -.3128621
>       noneng |   1.138732   .0587742    19.37   0.000     1.023536    1.253927
>      nonaust |   2.322202   .0632143    36.74   0.000     2.198304      2.4461
>        _cons |  -2.876826   .0415611   -69.22   0.000    -2.958284   -2.795368
> -------------+----------------------------------------------------------------
> 7            |
>       female |  -.0626167   .1889866    -0.33   0.740    -.4330237    .3077902
>       noneng |  -.1017222   .2996702    -0.34   0.734    -.6890651    .4856207
>      nonaust |   .9465996   .3576756     2.65   0.008     .2455683    1.647631
>        _cons |  -5.264885   .1549144   -33.99   0.000    -5.568511   -4.961258
> ------------------------------------------------------------------------------
>
> .
> end of do-file
>
> . crtest
> factor variables and time-series operators not allowed
> r(101);
>
> Could anyone please advise on what is causing the error message, and how I may resolve thi?

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index