Non proportional odds model in GSEM

[Martijn Hogerbrugge]

Does anyone know whether this can be done? I am trying to get marginal effects for a model I ran in GLLAMM, but I cannot get these - either using the margins command, or the gllapred command. Some forum posts suggest to re-estimate models from GLLAMM using Stata's own commands, like meologit, or gsem. However, how an ordered logistic regression model with non proportional odds is created in gsem is still a question to me.

I look forward to your responses.

[Richard Williams]

Hi Martijn. The coding is a little painful and probably error-prone, and you have to use the probit link, but the user-written regoprob2 (available from SSC) might meet your needs. Example:

use https://www.stata-press.com/data/r16/tvsfpors, clear
xtset school
quietly xtoprobit thk prethk i.cc i.tv
margins cc tv

quietly regoprob2 thk prethk cc tv, i(school) pl
forval j = 1/4 {
    margins, at(cc=(0 1)) at(tv=(0 1)) predict(outcome(`j'))
}

regoprob2 is an older program and does not support factor variables. As shown, you can work around that but if you've got interaction or squared terms life may be more complicated.

[Martijn Hogerbrugge]

Hi Richard,

I installed the package, but when trying to fit a model, I receive the following error message:

regoprob2 degree_volunteering female, i(cntry2) pl

Fitting constant-only model:
__00000X not found
r(111);

Any idea what is causing the error message?
Also, when using the sample data, I receive another error message:

. use https://www.stata-press.com/data/r16/tvsfpors, clear

.  regoprob2 thk prethk cc tv, i(school) pl

Fitting constant-only model:

Fitting full model:

Iteration 0:   log likelihood = -2192.2464  (not concave)
Iteration 1:   log likelihood = -2164.8319  (not concave)
no observations
r(2000);

[Richard Williams]

I forgot, I had the exact same problem with regoprob2 a year ago. I wrote to the authors, and they said "As far as I can see, the problem isn’t regoprob2, but regoprob has some problems which can be overcome using the updated regoprob version. Unfortunately, [regoprob's author] has not updated his program. I have attached an updated version of the original regoprob program."

The updated regoprob program can be found at

https://www3.nd.edu/~rwilliam/stataf...rob_update.zip

Extract the files and then place them someplace (e.g. c:\ado\personal\) where Stata will find them before it finds the original regoprob files.

As a sidelight, regoprob and regoprob2 both came out long before xtoprobit. They are slower but I'm impressed with how well they work, giving very similar results.

[Martijn Hogerbrugge]

That did solve the issue with the sample dataset, and I receive the expected output.
Unfortunately, with my dataset, I still have the same error message returning:

regoprob2 degree_volunteering female, i(cntry2) pl

Fitting constant-only model:
__00000X not found
r(111);

[Richard Williams]

Does xtoprobit run ok?

It isn't my program, but if you can email me your data and code I will try it.

[Richard Williams]

Stefan Boes is going to post an updated regoprob soon which will make my workaround unnecessary. I can’t say much about the other problem you are having unless I have the data or a replicable example.

[Martijn Hogerbrugge]

Hi Richard,

xtoprobit works fine on my data, but regoprob2 does not:

. xtset cntry2
       panel variable:  cntry2 (unbalanced)

. xtoprobit degree_volunteering female, nolog

Random-effects ordered probit regression        Number of obs     =     48,073
Group variable: cntry2                          Number of groups  =         27

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =        714
                                                              avg =    1,780.5
                                                              max =      2,868

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(1)      =       1.83
Log likelihood  = -56886.019                    Prob > chi2       =     0.1756

-------------------------------------------------------------------------------------
degree_volunteering |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
             female |   .0150184   .0110893     1.35   0.176    -.0067162     .036753
--------------------+----------------------------------------------------------------
              /cut1 |   .3611059   .0740433                      .2159837    .5062281
              /cut2 |   .7400063   .0740825                      .5948072    .8852054
              /cut3 |   .9599813   .0741238                      .8147012    1.105261
              /cut4 |   1.175925   .0741827                      1.030529     1.32132
              /cut5 |   1.609934   .0743879                      1.464136    1.755731
--------------------+----------------------------------------------------------------
          /sigma2_u |   .1460211   .0400027                      .0853548    .2498062
-------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 4500.07      Prob >= chibar2 = 0.0000

. regoprob2 degree_volunteering female, i(cntry2)

Fitting constant-only model:
__00000X not found
r(111);

Edit: After playing around with the data, selecting only a subsample of cases (max. 450 in each of the 27 countries, so 12,150 cases total) seems to make the model run and give me output. And the model still runs after adding additional covariates.

Increasing the matsize does not solve the issues of not being able to run the model on the complete sample.

[Richard Williams]

What version of Stata do you have? 16 has better memory management. Again if you can give me the data or a replicable example I'll see if there is any easy workaround.

Sometimes you can get by with changing the version # in a program, e.g. change version 8 to version 14. But you have to make sure the program still works correctly then.

[Martijn Hogerbrugge]

I have Stata version 15.1. I will send the data by email to you now.
Thanks for your help.

[Neeraj Kumar]

Hi,
Sir Richard Williams, I am using the regoprob2 command for estimation. My dependent variable having 10 categories and I am using 5 year panel data. I used the follopwing command

regoprob2 A170 quantX0472 quantX0473 quantX0474 relative2 relative3 relative4 IU0 X003, i(newid)

I am getting the results too but when I am using the margin command, you have mentioned in #2. it showing the error. Kindly guide me to calculate margins in the case of regoprob2.

[Richard Williams]

Neeraj Kumar My earlier code still works. Please show your commands and output. It could just be you have a syntax problem (e.g. a misspelled option) or it could be a more fundamental problem. Without seeing your code and output, it is impossible to say.

[Neeraj Kumar]

Thank you so much Sir Richard Williams for your reply. Sir here I am attaching the command as well output.

. regoprob2 A170 quantX0472 quantX0473 quantX0474 relative2 relative3 relative4 IU0 X003 X011 marriage, i(newid)

Fitting constant-only model:
rho <= 0, set to rho = 0.01

Fitting full model:

Iteration 0:   log likelihood =  -19799.85  
Iteration 1:   log likelihood = -19013.978  
Iteration 2:   log likelihood = -18984.695  
Iteration 3:   log likelihood =  -18983.71  
Iteration 4:   log likelihood = -18983.708  
Iteration 5:   log likelihood = -18983.708  

Random Effects Generalized Ordered Probit         Number of obs   =       9458
                                                  LR chi2(90)     =      14.62
Log likelihood = -18983.708                       Prob > chi2     =     1.0000

------------------------------------------------------------------------------
        A170 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
mleq1        |
  quantX0472 |   .3712199   .0703646     5.28   0.000     .2333078    .5091319
  quantX0473 |   .2990958   .1052161     2.84   0.004      .092876    .5053155
  quantX0474 |   .5384643   .1444039     3.73   0.000     .2554379    .8214907
   relative2 |  -.0387631   .0731855    -0.53   0.596     -.182204    .1046779
   relative3 |   .1104649   .0987894     1.12   0.263    -.0831588    .3040886
   relative4 |   .1872434   .1362469     1.37   0.169    -.0797956    .4542824
         IU0 |   .8879326   .3449306     2.57   0.010      .211881    1.563984
        X003 |  -.0047363   .0016921    -2.80   0.005    -.0080528   -.0014199
        X011 |  -.0020559   .0140705    -0.15   0.884    -.0296337    .0255219
    marriage |   .1488653   .0695044     2.14   0.032     .0126391    .2850914
       _cons |   .8420028   .2696995     3.12   0.002     .3134014    1.370604
-------------+----------------------------------------------------------------
mleq2        |
  quantX0472 |   .2980582   .0613732     4.86   0.000      .177769    .4183475
  quantX0473 |   .2180192   .0910305     2.40   0.017     .0396027    .3964358
  quantX0474 |   .3965289   .1135108     3.49   0.000     .1740518     .619006
   relative2 |   .0826411   .0645248     1.28   0.200    -.0438251    .2091074
   relative3 |   .2806092   .0846074     3.32   0.001     .1147818    .4464365
   relative4 |     .25531   .1074024     2.38   0.017     .0448052    .4658148
         IU0 |   .1374752   .3274172     0.42   0.675    -.5042508    .7792011
        X003 |  -.0035726   .0015252    -2.34   0.019     -.006562   -.0005833
        X011 |  -.0052946   .0124854    -0.42   0.672    -.0297656    .0191764
    marriage |   .1408971     .06285     2.24   0.025     .0177133    .2640808
       _cons |   1.199536    .255893     4.69   0.000     .6979952    1.701077
-------------+----------------------------------------------------------------
mleq3        |
  quantX0472 |   .3243286   .0476655     6.80   0.000      .230906    .4177513
  quantX0473 |   .5907402   .0751643     7.86   0.000     .4434208    .7380596
  quantX0474 |   .9798843   .0933194    10.50   0.000     .7969817    1.162787
   relative2 |  -.0617923   .0494261    -1.25   0.211    -.1586657    .0350811
   relative3 |  -.0799129   .0628837    -1.27   0.204    -.2031626    .0433368
   relative4 |  -.1988519   .0861433    -2.31   0.021    -.3676896   -.0300141
         IU0 |   2.645902    .231547    11.43   0.000     2.192078    3.099726
        X003 |  -.0031191   .0011826    -2.64   0.008    -.0054369   -.0008014
        X011 |  -.0506203   .0093759    -5.40   0.000    -.0689967   -.0322438
    marriage |   .2080788   .0486246     4.28   0.000     .1127763    .3033812
       _cons |  -1.167085    .181235    -6.44   0.000    -1.522299   -.8118708
-------------+----------------------------------------------------------------
mleq4        |
  quantX0472 |    .178823   .0465012     3.85   0.000     .0876823    .2699638
  quantX0473 |   .4139642   .0721104     5.74   0.000     .2726304    .5552979
  quantX0474 |   .6895927   .0882476     7.81   0.000     .5166306    .8625548
   relative2 |   .0951083   .0479796     1.98   0.047       .00107    .1891466
   relative3 |   .1154234     .06167     1.87   0.061    -.0054476    .2362944
   relative4 |   .0831767   .0825284     1.01   0.314     -.078576    .2449294
         IU0 |   1.401816   .2258157     6.21   0.000     .9592257    1.844407
        X003 |  -.0043627    .001138    -3.83   0.000    -.0065933   -.0021322
        X011 |  -.0209891    .009074    -2.31   0.021    -.0387738   -.0032044
    marriage |   .1530804   .0470582     3.25   0.001     .0608481    .2453127
       _cons |   -.417688   .1778758    -2.35   0.019    -.7663183   -.0690578
-------------+----------------------------------------------------------------
mleq5        |
  quantX0472 |   .2982999   .0434344     6.87   0.000     .2131701    .3834297
  quantX0473 |   .7951107   .0650926    12.22   0.000     .6675315    .9226899
  quantX0474 |   1.356479   .0774001    17.53   0.000     1.204778    1.508181
   relative2 |  -.0719204   .0439514    -1.64   0.102    -.1580637    .0142228
   relative3 |  -.2888017   .0553227    -5.22   0.000    -.3972322   -.1803713
   relative4 |  -.5684913   .0720125    -7.89   0.000    -.7096332   -.4273495
         IU0 |   2.598029   .2098799    12.38   0.000     2.186672    3.009386
        X003 |  -.0031936   .0010509    -3.04   0.002    -.0052533    -.001134
        X011 |  -.0481928   .0085047    -5.67   0.000    -.0648618   -.0315239
    marriage |    .177438   .0434556     4.08   0.000     .0922666    .2626094
       _cons |  -2.060834   .1659657   -12.42   0.000    -2.386121   -1.735548
-------------+----------------------------------------------------------------
mleq6        |
  quantX0472 |   .1573262    .043758     3.60   0.000      .071562    .2430904
  quantX0473 |   .4141032   .0646828     6.40   0.000     .2873271    .5408792
  quantX0474 |   .8384677   .0760297    11.03   0.000     .6894522    .9874832
   relative2 |   .0577221   .0442551     1.30   0.192    -.0290162    .1444604
   relative3 |  -.0733708   .0556145    -1.32   0.187    -.1823732    .0356315
   relative4 |  -.0919314   .0711935    -1.29   0.197     -.231468    .0476052
         IU0 |   1.869935   .2134152     8.76   0.000     1.451649    2.288221
        X003 |  -.0025437   .0010614    -2.40   0.017     -.004624   -.0004635
        X011 |  -.0453112   .0087243    -5.19   0.000    -.0624105   -.0282119
    marriage |   .1260287   .0440101     2.86   0.004     .0397706    .2122869
       _cons |   -1.69491   .1689797   -10.03   0.000    -2.026104   -1.363715
-------------+----------------------------------------------------------------
mleq7        |
  quantX0472 |   .1296422   .0472218     2.75   0.006     .0370892    .2221953
  quantX0473 |   .3736992   .0691113     5.41   0.000     .2382436    .5091548
  quantX0474 |   .7959402   .0799526     9.96   0.000      .639236    .9526444
   relative2 |  -.0560071   .0480069    -1.17   0.243    -.1500989    .0380847
   relative3 |  -.1580012   .0604743    -2.61   0.009    -.2765287   -.0394737
   relative4 |  -.2092494   .0748396    -2.80   0.005    -.3559323   -.0625665
         IU0 |   2.980595   .2617242    11.39   0.000     2.467624    3.493565
        X003 |  -.0029187   .0011421    -2.56   0.011    -.0051573   -.0006801
        X011 |  -.0383926    .009545    -4.02   0.000    -.0571004   -.0196849
    marriage |   .0742837   .0476494     1.56   0.119    -.0191073    .1676747
       _cons |  -2.883938   .2061239   -13.99   0.000    -3.287933   -2.479943
-------------+----------------------------------------------------------------
mleq8        |
  quantX0472 |  -.0686852   .0517338    -1.33   0.184    -.1700816    .0327112
  quantX0473 |   .1030861   .0753398     1.37   0.171    -.0445773    .2507494
  quantX0474 |   .4448737   .0852212     5.22   0.000     .2778432    .6119042
   relative2 |   .0468859   .0527204     0.89   0.374    -.0564441    .1502159
   relative3 |  -.0389005   .0671361    -0.58   0.562    -.1704848    .0926837
   relative4 |   .0145576   .0801179     0.18   0.856    -.1424706    .1715857
         IU0 |    1.81802   .3013497     6.03   0.000     1.227385    2.408654
        X003 |  -.0013129   .0012581    -1.04   0.297    -.0037787    .0011528
        X011 |  -.0157169   .0105124    -1.50   0.135    -.0363209     .004887
    marriage |     .02127   .0514969     0.41   0.680    -.0796621    .1222021
       _cons |   -2.31973   .2351128    -9.87   0.000    -2.780543   -1.858918
-------------+----------------------------------------------------------------
mleq9        |
  quantX0472 |  -.2054642   .0561782    -3.66   0.000    -.3155714    -.095357
  quantX0473 |  -.1367442   .0824348    -1.66   0.097    -.2983133    .0248249
  quantX0474 |   .0396952   .0916723     0.43   0.665    -.1399791    .2193696
   relative2 |   .1036618   .0570458     1.82   0.069     -.008146    .2154696
   relative3 |   .1314401   .0733415     1.79   0.073    -.0123065    .2751868
   relative4 |   .1700958   .0866897     1.96   0.050     .0001872    .3400044
         IU0 |   2.230735   .3526587     6.33   0.000     1.539537    2.921933
        X003 |  -.0000519   .0013914    -0.04   0.970    -.0027789    .0026752
        X011 |  -.0031044   .0113738    -0.27   0.785    -.0253966    .0191877
    marriage |  -.0652516   .0560576    -1.16   0.244    -.1751224    .0446193
       _cons |  -2.801932   .2749377   -10.19   0.000      -3.3408   -2.263064
-------------+----------------------------------------------------------------
rho          |
       _cons |   .0357694   .0098023     3.65   0.000     .0165573    .0549815
------------------------------------------------------------------------------

There most of the variable used in the above output are dummy variable (having value 0 and 1) except IU0, X003 and X011. These three variables are continuous. My dependent variable A170 having 10 categories.

. forval j = 1/10{
  2.
.     margins, at(quantX0472=(0 1)) at(quantX0473=(0 1))  at(quantX0474=(0 1)) at(relative2=(0 1)) at(relative3=(0 1)) at(relativ
> e4=(0 1)) at(marriage=(0 1))at(IU0=(0.747)) at(X011=(2.60)) at (X003= (39.17))  predict (outcome(`j'))
  3.
. }
command margins not appropriate after mean
r(322);

The values mentioned in the bracket of the variables IU0, X011 and X003 are the mean values.
Kindly guide my further on the issue.

[Richard Williams]

Is there an intermediate command before margins (E.g. mean) that you are not showing us?

sometimes commands are super-picky about spaces. Make sure there is always one space before each at, and no spaces between the at and the following parentheses.

you are allowing all coefficients to differed between panels. I always think gologit models are pointless in such situations. They gain you no parsimony, so you might as well use better known mlogit models.

if you still can’t get the model to run, you can either try posting an extract of your data, or, if you are free to share, send your data to me. My first guess, though, is that either you aren’t showing us all your commands, or there is a trivial syntax error which is zapping things. Even if the model did run I don’t think it would be worth much as is, but it might be worth something if several variables meet the parallel lines constraint.

[Neeraj Kumar]

Yes, I have calculated the mean of my three continuous variables, these variable I have used in the margin also. I am sharing the entire result now. Thank you so much for your guidance. One question is it correct to use the mean value which I have used in the margin for my continuous variable. Actually I had run the wald test earlier. That result skipped from my mind to save.

 
. regoprob2 A170 quantX0472 quantX0473 quantX0474 relative2 relative3 relative4 IU0 X003 X011 marriage, i(newid)

Fitting constant-only model:
rho <= 0, set to rho = 0.01

Fitting full model:

Iteration 0:   log likelihood =  -19799.85  
Iteration 1:   log likelihood = -19013.978  
Iteration 2:   log likelihood = -18984.695  
Iteration 3:   log likelihood =  -18983.71  
Iteration 4:   log likelihood = -18983.708  
Iteration 5:   log likelihood = -18983.708  

Random Effects Generalized Ordered Probit         Number of obs   =       9458
                                                  LR chi2(90)     =      14.62
Log likelihood = -18983.708                       Prob > chi2     =     1.0000

------------------------------------------------------------------------------
        A170 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
mleq1        |
  quantX0472 |   .3712199   .0703646     5.28   0.000     .2333078    .5091319
  quantX0473 |   .2990958   .1052161     2.84   0.004      .092876    .5053155
  quantX0474 |   .5384643   .1444039     3.73   0.000     .2554379    .8214907
   relative2 |  -.0387631   .0731855    -0.53   0.596     -.182204    .1046779
   relative3 |   .1104649   .0987894     1.12   0.263    -.0831588    .3040886
   relative4 |   .1872434   .1362469     1.37   0.169    -.0797956    .4542824
         IU0 |   .8879326   .3449306     2.57   0.010      .211881    1.563984
        X003 |  -.0047363   .0016921    -2.80   0.005    -.0080528   -.0014199
        X011 |  -.0020559   .0140705    -0.15   0.884    -.0296337    .0255219
    marriage |   .1488653   .0695044     2.14   0.032     .0126391    .2850914
       _cons |   .8420028   .2696995     3.12   0.002     .3134014    1.370604
-------------+----------------------------------------------------------------
mleq2        |
  quantX0472 |   .2980582   .0613732     4.86   0.000      .177769    .4183475
  quantX0473 |   .2180192   .0910305     2.40   0.017     .0396027    .3964358
  quantX0474 |   .3965289   .1135108     3.49   0.000     .1740518     .619006
   relative2 |   .0826411   .0645248     1.28   0.200    -.0438251    .2091074
   relative3 |   .2806092   .0846074     3.32   0.001     .1147818    .4464365
   relative4 |     .25531   .1074024     2.38   0.017     .0448052    .4658148
         IU0 |   .1374752   .3274172     0.42   0.675    -.5042508    .7792011
        X003 |  -.0035726   .0015252    -2.34   0.019     -.006562   -.0005833
        X011 |  -.0052946   .0124854    -0.42   0.672    -.0297656    .0191764
    marriage |   .1408971     .06285     2.24   0.025     .0177133    .2640808
       _cons |   1.199536    .255893     4.69   0.000     .6979952    1.701077
-------------+----------------------------------------------------------------
mleq3        |
  quantX0472 |   .3243286   .0476655     6.80   0.000      .230906    .4177513
  quantX0473 |   .5907402   .0751643     7.86   0.000     .4434208    .7380596
  quantX0474 |   .9798843   .0933194    10.50   0.000     .7969817    1.162787
   relative2 |  -.0617923   .0494261    -1.25   0.211    -.1586657    .0350811
   relative3 |  -.0799129   .0628837    -1.27   0.204    -.2031626    .0433368
   relative4 |  -.1988519   .0861433    -2.31   0.021    -.3676896   -.0300141
         IU0 |   2.645902    .231547    11.43   0.000     2.192078    3.099726
        X003 |  -.0031191   .0011826    -2.64   0.008    -.0054369   -.0008014
        X011 |  -.0506203   .0093759    -5.40   0.000    -.0689967   -.0322438
    marriage |   .2080788   .0486246     4.28   0.000     .1127763    .3033812
       _cons |  -1.167085    .181235    -6.44   0.000    -1.522299   -.8118708
-------------+----------------------------------------------------------------
mleq4        |
  quantX0472 |    .178823   .0465012     3.85   0.000     .0876823    .2699638
  quantX0473 |   .4139642   .0721104     5.74   0.000     .2726304    .5552979
  quantX0474 |   .6895927   .0882476     7.81   0.000     .5166306    .8625548
   relative2 |   .0951083   .0479796     1.98   0.047       .00107    .1891466
   relative3 |   .1154234     .06167     1.87   0.061    -.0054476    .2362944
   relative4 |   .0831767   .0825284     1.01   0.314     -.078576    .2449294
         IU0 |   1.401816   .2258157     6.21   0.000     .9592257    1.844407
        X003 |  -.0043627    .001138    -3.83   0.000    -.0065933   -.0021322
        X011 |  -.0209891    .009074    -2.31   0.021    -.0387738   -.0032044
    marriage |   .1530804   .0470582     3.25   0.001     .0608481    .2453127
       _cons |   -.417688   .1778758    -2.35   0.019    -.7663183   -.0690578
-------------+----------------------------------------------------------------
mleq5        |
  quantX0472 |   .2982999   .0434344     6.87   0.000     .2131701    .3834297
  quantX0473 |   .7951107   .0650926    12.22   0.000     .6675315    .9226899
  quantX0474 |   1.356479   .0774001    17.53   0.000     1.204778    1.508181
   relative2 |  -.0719204   .0439514    -1.64   0.102    -.1580637    .0142228
   relative3 |  -.2888017   .0553227    -5.22   0.000    -.3972322   -.1803713
   relative4 |  -.5684913   .0720125    -7.89   0.000    -.7096332   -.4273495
         IU0 |   2.598029   .2098799    12.38   0.000     2.186672    3.009386
        X003 |  -.0031936   .0010509    -3.04   0.002    -.0052533    -.001134
        X011 |  -.0481928   .0085047    -5.67   0.000    -.0648618   -.0315239
    marriage |    .177438   .0434556     4.08   0.000     .0922666    .2626094
       _cons |  -2.060834   .1659657   -12.42   0.000    -2.386121   -1.735548
-------------+----------------------------------------------------------------
mleq6        |
  quantX0472 |   .1573262    .043758     3.60   0.000      .071562    .2430904
  quantX0473 |   .4141032   .0646828     6.40   0.000     .2873271    .5408792
  quantX0474 |   .8384677   .0760297    11.03   0.000     .6894522    .9874832
   relative2 |   .0577221   .0442551     1.30   0.192    -.0290162    .1444604
   relative3 |  -.0733708   .0556145    -1.32   0.187    -.1823732    .0356315
   relative4 |  -.0919314   .0711935    -1.29   0.197     -.231468    .0476052
         IU0 |   1.869935   .2134152     8.76   0.000     1.451649    2.288221
        X003 |  -.0025437   .0010614    -2.40   0.017     -.004624   -.0004635
        X011 |  -.0453112   .0087243    -5.19   0.000    -.0624105   -.0282119
    marriage |   .1260287   .0440101     2.86   0.004     .0397706    .2122869
       _cons |   -1.69491   .1689797   -10.03   0.000    -2.026104   -1.363715
-------------+----------------------------------------------------------------
mleq7        |
  quantX0472 |   .1296422   .0472218     2.75   0.006     .0370892    .2221953
  quantX0473 |   .3736992   .0691113     5.41   0.000     .2382436    .5091548
  quantX0474 |   .7959402   .0799526     9.96   0.000      .639236    .9526444
   relative2 |  -.0560071   .0480069    -1.17   0.243    -.1500989    .0380847
   relative3 |  -.1580012   .0604743    -2.61   0.009    -.2765287   -.0394737
   relative4 |  -.2092494   .0748396    -2.80   0.005    -.3559323   -.0625665
         IU0 |   2.980595   .2617242    11.39   0.000     2.467624    3.493565
        X003 |  -.0029187   .0011421    -2.56   0.011    -.0051573   -.0006801
        X011 |  -.0383926    .009545    -4.02   0.000    -.0571004   -.0196849
    marriage |   .0742837   .0476494     1.56   0.119    -.0191073    .1676747
       _cons |  -2.883938   .2061239   -13.99   0.000    -3.287933   -2.479943
-------------+----------------------------------------------------------------
mleq8        |
  quantX0472 |  -.0686852   .0517338    -1.33   0.184    -.1700816    .0327112
  quantX0473 |   .1030861   .0753398     1.37   0.171    -.0445773    .2507494
  quantX0474 |   .4448737   .0852212     5.22   0.000     .2778432    .6119042
   relative2 |   .0468859   .0527204     0.89   0.374    -.0564441    .1502159
   relative3 |  -.0389005   .0671361    -0.58   0.562    -.1704848    .0926837
   relative4 |   .0145576   .0801179     0.18   0.856    -.1424706    .1715857
         IU0 |    1.81802   .3013497     6.03   0.000     1.227385    2.408654
        X003 |  -.0013129   .0012581    -1.04   0.297    -.0037787    .0011528
        X011 |  -.0157169   .0105124    -1.50   0.135    -.0363209     .004887
    marriage |     .02127   .0514969     0.41   0.680    -.0796621    .1222021
       _cons |   -2.31973   .2351128    -9.87   0.000    -2.780543   -1.858918
-------------+----------------------------------------------------------------
mleq9        |
  quantX0472 |  -.2054642   .0561782    -3.66   0.000    -.3155714    -.095357
  quantX0473 |  -.1367442   .0824348    -1.66   0.097    -.2983133    .0248249
  quantX0474 |   .0396952   .0916723     0.43   0.665    -.1399791    .2193696
   relative2 |   .1036618   .0570458     1.82   0.069     -.008146    .2154696
   relative3 |   .1314401   .0733415     1.79   0.073    -.0123065    .2751868
   relative4 |   .1700958   .0866897     1.96   0.050     .0001872    .3400044
         IU0 |   2.230735   .3526587     6.33   0.000     1.539537    2.921933
        X003 |  -.0000519   .0013914    -0.04   0.970    -.0027789    .0026752
        X011 |  -.0031044   .0113738    -0.27   0.785    -.0253966    .0191877
    marriage |  -.0652516   .0560576    -1.16   0.244    -.1751224    .0446193
       _cons |  -2.801932   .2749377   -10.19   0.000      -3.3408   -2.263064
-------------+----------------------------------------------------------------
rho          |
       _cons |   .0357694   .0098023     3.65   0.000     .0165573    .0549815
------------------------------------------------------------------------------

. forval j = 1/10 {
  2. 
.     margins, at(quantX0472=(0 1)) at(quantX0473=(0 1))  at(quantX0474=(0 1)) at(relative2=(0 1)) at(relative3=(0 1)) at(relativ
> e4=(0 1)) at(marriage=(0 1))at(IU0=(0.747)) at(X011=(2.60)) at (X003=(39.17)) predict (outcome (`j'))
  3. 
. }

Predictive margins                              Number of obs     =      9,458
Model VCE    : OIM

Expression   : Pr(A170==1), predict(outcome (1))

1._at        : quantX0472      =           0

2._at        : quantX0472      =           1

3._at        : quantX0473      =           0

4._at        : quantX0473      =           1

5._at        : quantX0474      =           0

6._at        : quantX0474      =           1

7._at        : relative2       =           0

8._at        : relative2       =           1

9._at        : relative3       =           0

10._at       : relative3       =           1

11._at       : relative4       =           0

12._at       : relative4       =           1

13._at       : marriage        =           0

14._at       : marriage        =           1

15._at       : IU0             =        .747

16._at       : X011            =         2.6

17._at       : X003            =       39.17

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0606092   .0045526    13.31   0.000     .0516864    .0695321
          2  |   .0286091   .0027828    10.28   0.000     .0231549    .0340632
          3  |   .0488252   .0029216    16.71   0.000     .0430991    .0545514
          4  |   .0261725   .0051535     5.08   0.000     .0160719    .0362732
          5  |    .052846   .0040175    13.15   0.000     .0449719    .0607201
          6  |   .0162795   .0046103     3.53   0.000     .0072435    .0253154
          7  |   .0435483   .0025678    16.96   0.000     .0385154    .0485811
          8  |   .0470542   .0056855     8.28   0.000     .0359108    .0581975
          9  |   .0457005   .0024691    18.51   0.000     .0408612    .0505398
         10  |   .0364689   .0068043     5.36   0.000     .0231328     .049805
         11  |   .0467201   .0028847    16.20   0.000     .0410661    .0523741
         12  |   .0316686   .0081775     3.87   0.000      .015641    .0476961
         13  |   .0573123   .0071453     8.02   0.000     .0433076    .0713169
         14  |   .0427777   .0021926    19.51   0.000     .0384803    .0470751
         15  |   .0440387   .0021183    20.79   0.000     .0398869    .0481904
         16  |   .0443512   .0021256    20.87   0.000     .0401852    .0485172
         17  |   .0434627   .0021188    20.51   0.000     .0393099    .0476154
------------------------------------------------------------------------------

Predictive margins                              Number of obs     =      9,458
Model VCE    : OIM

Expression   : Pr(A170==2), predict(outcome (2))

1._at        : quantX0472      =           0

2._at        : quantX0472      =           1

3._at        : quantX0473      =           0

4._at        : quantX0473      =           1

5._at        : quantX0474      =           0

6._at        : quantX0474      =           1

7._at        : relative2       =           0

8._at        : relative2       =           1

9._at        : relative3       =           0

10._at       : relative3       =           1

11._at       : relative4       =           0

12._at       : relative4       =           1

13._at       : marriage        =           0

14._at       : marriage        =           1

15._at       : IU0             =        .747

16._at       : X011            =         2.6

17._at       : X003            =       39.17

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |    .015423   .0026154     5.90   0.000      .010297    .0205491
          2  |   .0144205   .0021437     6.73   0.000     .0102189     .018622
          3  |   .0156091   .0016876     9.25   0.000     .0123014    .0189168
          4  |    .015949   .0039668     4.02   0.000     .0081744    .0237237
          5  |   .0162197    .002594     6.25   0.000     .0111355    .0213039
          6  |   .0148059    .004372     3.39   0.001      .006237    .0233748
          7  |   .0195844   .0017969    10.90   0.000     .0160625    .0231063
          8  |   .0068872   .0031367     2.20   0.028     .0007393    .0130351
          9  |   .0195646   .0017148    11.41   0.000     .0162036    .0229255
         10  |   .0010094    .003904     0.26   0.796    -.0066424    .0086611
         11  |   .0193331   .0022446     8.61   0.000     .0149337    .0237325
         12  |     .00828   .0057569     1.44   0.150    -.0030033    .0195633
         13  |    .019118   .0043981     4.35   0.000     .0104979     .027738
         14  |   .0161526   .0013567    11.91   0.000     .0134935    .0188116
         15  |   .0167697   .0013285    12.62   0.000     .0141658    .0193736
         16  |   .0164002   .0013078    12.54   0.000     .0138369    .0189634
         17  |   .0165172   .0013249    12.47   0.000     .0139204     .019114
------------------------------------------------------------------------------

Predictive margins                              Number of obs     =      9,458
Model VCE    : OIM

Expression   : Pr(A170==3), predict(outcome (3))

1._at        : quantX0472      =           0

2._at        : quantX0472      =           1

3._at        : quantX0473      =           0

4._at        : quantX0473      =           1

5._at        : quantX0474      =           0

6._at        : quantX0474      =           1

7._at        : relative2       =           0

8._at        : relative2       =           1

9._at        : relative3       =           0

10._at       : relative3       =           1

11._at       : relative4       =           0

12._at       : relative4       =           1

13._at       : marriage        =           0

14._at       : marriage        =           1

15._at       : IU0             =        .747

16._at       : X011            =         2.6

17._at       : X003            =       39.17

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1181924   .0055021    21.48   0.000     .1074084    .1289764
          2  |    .080503   .0044121    18.25   0.000     .0718554    .0891506
          3  |   .1168936     .00411    28.44   0.000     .1088382    .1249491
          4  |   .0311195   .0055732     5.58   0.000     .0201963    .0420427
          5  |   .1368172   .0058796    23.27   0.000     .1252935     .148341
          6  |     .00925   .0043176     2.14   0.032     .0007877    .0177122
          7  |   .0924863   .0035634    25.95   0.000     .0855022    .0994704
          8  |   .1157135   .0076099    15.21   0.000     .1007985    .1306286
          9  |   .0914359    .003171    28.84   0.000     .0852208    .0976509
         10  |   .1375691   .0110952    12.40   0.000      .115823    .1593153
         11  |   .0874887   .0032713    26.74   0.000      .081077    .0939004
         12  |   .1604886   .0154795    10.37   0.000     .1301493     .190828
         13  |   .1273619   .0097832    13.02   0.000     .1081872    .1465366
         14  |   .0947404   .0031169    30.40   0.000     .0886314    .1008495
         15  |   .0937732   .0029652    31.62   0.000     .0879614     .099585
         16  |   .0965522   .0029929    32.26   0.000     .0906862    .1024183
         17  |   .0976772    .003022    32.32   0.000     .0917543    .1036002
------------------------------------------------------------------------------

Predictive margins                              Number of obs     =      9,458
Model VCE    : OIM

Expression   : Pr(A170==4), predict(outcome (4))

1._at        : quantX0472      =           0

2._at        : quantX0472      =           1

3._at        : quantX0473      =           0

4._at        : quantX0473      =           1

5._at        : quantX0474      =           0

6._at        : quantX0474      =           1

7._at        : relative2       =           0

8._at        : relative2       =           1

9._at        : relative3       =           0

10._at       : relative3       =           1

11._at       : relative4       =           0

12._at       : relative4       =           1

13._at       : marriage        =           0

14._at       : marriage        =           1

15._at       : IU0             =        .747

16._at       : X011            =         2.6

17._at       : X003            =       39.17

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0347948   .0024048    14.47   0.000     .0300814    .0395083
          2  |   .0581399   .0036259    16.03   0.000     .0510332    .0652465
          3  |   .0427779   .0022321    19.17   0.000     .0384032    .0471527
          4  |   .0521217   .0062655     8.32   0.000     .0398414    .0644019
          5  |   .0381155   .0028514    13.37   0.000     .0325269    .0437041
          6  |   .0468384   .0064716     7.24   0.000     .0341543    .0595225
          7  |   .0584928   .0028906    20.24   0.000     .0528273    .0641583
          8  |   .0193515   .0033232     5.82   0.000     .0128382    .0258647
          9  |   .0553594   .0026378    20.99   0.000     .0501893    .0605294
         10  |   .0068916   .0035612     1.94   0.053    -.0000882    .0138713
         11  |    .057551   .0032621    17.64   0.000     .0511574    .0639447
         12  |  -.0113386    .007386    -1.54   0.125     -.025815    .0031378
         13  |   .0420291    .005462     7.69   0.000     .0313237    .0527345
         14  |   .0491203   .0022829    21.52   0.000     .0446458    .0535948
         15  |   .0512374   .0022946    22.33   0.000     .0467401    .0557346
         16  |    .049534   .0022248    22.26   0.000     .0451736    .0538944
         17  |   .0475185   .0021603    22.00   0.000     .0432844    .0517525
------------------------------------------------------------------------------

Predictive margins                              Number of obs     =      9,458
Model VCE    : OIM

Expression   : Pr(A170==5), predict(outcome (5))

1._at        : quantX0472      =           0

2._at        : quantX0472      =           1

3._at        : quantX0473      =           0

4._at        : quantX0473      =           1

5._at        : quantX0474      =           0

6._at        : quantX0474      =           1

7._at        : relative2       =           0

8._at        : relative2       =           1

9._at        : relative3       =           0

10._at       : relative3       =           1

11._at       : relative4       =           0

12._at       : relative4       =           1

13._at       : marriage        =           0

14._at       : marriage        =           1

15._at       : IU0             =        .747

16._at       : X011            =         2.6

17._at       : X003            =       39.17

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .2747913   .0072037    38.15   0.000     .2606723    .2889103
          2  |   .2171905   .0081945    26.50   0.000     .2011295    .2332515
          3  |    .278204   .0055006    50.58   0.000      .267423     .288985
          4  |   .1143138   .0129579     8.82   0.000     .0889169    .1397108
          5  |   .3238196    .007307    44.32   0.000      .309498    .3381411
          6  |   .0532676   .0102085     5.22   0.000     .0332593    .0732759
          7  |   .2405222   .0054814    43.88   0.000     .2297787    .2512656
          8  |   .2917396   .0114233    25.54   0.000     .2693505    .3141288
          9  |   .2333229   .0049372    47.26   0.000     .2236461    .2429997
         10  |   .3675802   .0159331    23.07   0.000     .3363518    .3988085
         11  |   .2132454   .0052922    40.29   0.000     .2028728    .2236179
         12  |   .4265534   .0179458    23.77   0.000     .3913802    .4617266
         13  |   .2728784   .0139771    19.52   0.000     .2454838     .300273
         14  |   .2513393   .0046439    54.12   0.000     .2422374    .2604412
         15  |   .2537436   .0044441    57.10   0.000     .2450334    .2624538
         16  |   .2537048   .0044225    57.37   0.000     .2450368    .2623728
         17  |    .253968   .0044559    57.00   0.000     .2452346    .2627014
------------------------------------------------------------------------------



.