# Multiple Imputation in Stata: Imputation Log File

This web page contains the log file from the example imputation discussed in the Imputing section, plus the graphics it creates.

```----------------------------------------------------------------------------------------------------------------------------------
name:
log:  \sscc\pubs\mi\miex.log
log type:  text
opened on:  17 Aug 2012, 10:51:48

.
. use midata

.
. // test missingness of data
. unab numvars: *

. unab missvars: urban-wage

. misstable sum, gen(miss_)
Obs<.
+------------------------------
|                                | Unique
Variable |     Obs=.     Obs>.     Obs<.  | values        Min         Max
-------------+--------------------------------+------------------------------
race |       293               2,707  |      3          0           2
urban |       273               2,727  |      2          0           1
edu |       319               2,681  |      4          1           4
exp |       293               2,707  |   >500          0     47.8623
wage |       299               2,701  |   >500          0    227465.2
-----------------------------------------------------------------------------

.
. foreach var of local missvars {
2.         local covars: list numvars - var
3.         display _newline(3) "logit missingness of `var' on `covars'"
4.         logit miss_`var' `covars'
5.         foreach nvar of local covars {
6.                 display _newline(3) "ttest of `nvar' by missingness of `var'"
7.                 ttest `nvar', by(miss_`var')
8.         }
9. }

logit missingness of urban on female race edu exp wage

Iteration 0:   log likelihood = -613.04047
Iteration 1:   log likelihood = -611.32144
Iteration 2:   log likelihood = -611.31554
Iteration 3:   log likelihood = -611.31554

Logistic regression                               Number of obs   =       1964
LR chi2(5)      =       3.45
Prob > chi2     =     0.6310
Log likelihood = -611.31554                       Pseudo R2       =     0.0028

------------------------------------------------------------------------------
miss_urban |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
female |   .1505333   .1696945     0.89   0.375    -.1820618    .4831284
race |   -.068621   .0980029    -0.70   0.484     -.260703    .1234611
edu |   .0098348   .0973647     0.10   0.920    -.1809964     .200666
exp |  -.0033092   .0094184    -0.35   0.725    -.0217689    .0151504
wage |   3.68e-06   2.57e-06     1.43   0.153    -1.36e-06    8.71e-06
_cons |  -2.513739   .2871859    -8.75   0.000    -3.076613   -1.950865
------------------------------------------------------------------------------

ttest of female by missingness of urban

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2727    .4979831    .0095764    .5000876    .4792053    .5167609
1 |     273    .4761905    .0302826      .50035    .4165725    .5358085
---------+--------------------------------------------------------------------
combined |    3000        .496    .0091299    .5000674    .4780984    .5139016
---------+--------------------------------------------------------------------
diff |            .0217927    .0317471               -.0404556    .0840409
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.6864
Ho: diff = 0                                     degrees of freedom =     2998

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.7538         Pr(|T| > |t|) = 0.4925          Pr(T > t) = 0.2462

ttest of race by missingness of urban

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2456    1.014658    .0163483    .8101878    .9826002    1.046716
1 |     251    1.055777    .0513125    .8129431     .954717    1.156837
---------+--------------------------------------------------------------------
combined |    2707    1.018471    .0155756    .8103808    .9879293    1.049012
---------+--------------------------------------------------------------------
diff |           -.0411189    .0537051               -.1464261    .0641883
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.7656
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.2220         Pr(|T| > |t|) = 0.4440          Pr(T > t) = 0.7780

ttest of edu by missingness of urban

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2442    2.356675    .0184465    .9115617    2.320503    2.392847
1 |     239    2.368201    .0595328    .9203542    2.250922    2.485479
---------+--------------------------------------------------------------------
combined |    2681    2.357702     .017617     .912182    2.323158    2.392247
---------+--------------------------------------------------------------------
diff |            -.011526    .0618353               -.1327757    .1097237
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.1864
Ho: diff = 0                                     degrees of freedom =     2679

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.4261         Pr(|T| > |t|) = 0.8521          Pr(T > t) = 0.5739

ttest of exp by missingness of urban

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2450    15.56019    .1954164     9.67262    15.17699    15.94339
1 |     257    15.69341    .5938361    9.519917    14.52399    16.86284
---------+--------------------------------------------------------------------
combined |    2707    15.57284    .1856003    9.656566    15.20891    15.93677
---------+--------------------------------------------------------------------
diff |           -.1332234    .6332773                -1.37498    1.108533
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.2104
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.4167         Pr(|T| > |t|) = 0.8334          Pr(T > t) = 0.5833

ttest of wage by missingness of urban

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2458    71240.46    763.6437    37860.09    69743.01    72737.91
1 |     243    74058.12     2597.11    40484.95    68942.29    79173.95
---------+--------------------------------------------------------------------
combined |    2701    71493.95    733.1819     38104.3     70056.3    72931.61
---------+--------------------------------------------------------------------
diff |           -2817.665    2562.273               -7841.881    2206.551
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -1.0997
Ho: diff = 0                                     degrees of freedom =     2699

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.1358         Pr(|T| > |t|) = 0.2716          Pr(T > t) = 0.8642

logit missingness of edu on female race urban exp wage

Iteration 0:   log likelihood = -670.64062
Iteration 1:   log likelihood = -669.91049
Iteration 2:   log likelihood = -669.90956
Iteration 3:   log likelihood = -669.90956

Logistic regression                               Number of obs   =       1989
LR chi2(5)      =       1.46
Prob > chi2     =     0.9174
Log likelihood = -669.90956                       Pseudo R2       =     0.0011

------------------------------------------------------------------------------
miss_edu |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
female |  -.0194159   .1557151    -0.12   0.901    -.3246119    .2857801
race |   .0569055   .0903871     0.63   0.529    -.1202499    .2340609
urban |   .0476788    .157765     0.30   0.762    -.2615349    .3568925
exp |  -.0028472   .0086668    -0.33   0.743    -.0198338    .0141393
wage |   1.93e-06   2.25e-06     0.86   0.390    -2.47e-06    6.33e-06
_cons |  -2.314849   .2528625    -9.15   0.000     -2.81045   -1.819248
------------------------------------------------------------------------------

ttest of female by missingness of edu

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2681    .4983215    .0096583    .5000905    .4793831      .51726
1 |     319     .476489    .0280076    .5002316    .4213854    .5315926
---------+--------------------------------------------------------------------
combined |    3000        .496    .0091299    .5000674    .4780984    .5139016
---------+--------------------------------------------------------------------
diff |            .0218325    .0296195               -.0362442    .0799092
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.7371
Ho: diff = 0                                     degrees of freedom =     2998

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.7694         Pr(|T| > |t|) = 0.4611          Pr(T > t) = 0.2306

ttest of race by missingness of edu

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2416    1.016142    .0164994    .8109934    .9837879    1.048497
1 |     291    1.037801    .0472723    .8064058    .9447603    1.130841
---------+--------------------------------------------------------------------
combined |    2707    1.018471    .0155756    .8103808    .9879293    1.049012
---------+--------------------------------------------------------------------
diff |           -.0216583    .0502926                -.120274    .0769574
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.4306
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.3334         Pr(|T| > |t|) = 0.6668          Pr(T > t) = 0.6666

ttest of urban by missingness of edu

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2442    .6588862    .0095956    .4741806    .6400698    .6777025
1 |     285    .6912281    .0274139    .4627995    .6372679    .7451882
---------+--------------------------------------------------------------------
combined |    2727    .6622662    .0090582     .473024    .6445046    .6800278
---------+--------------------------------------------------------------------
diff |           -.0323419    .0296084               -.0903991    .0257153
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -1.0923
Ho: diff = 0                                     degrees of freedom =     2725

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.1374         Pr(|T| > |t|) = 0.2748          Pr(T > t) = 0.8626

ttest of exp by missingness of edu

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2419    15.61121    .1972722    9.702505    15.22437    15.99805
1 |     288    15.25056    .5463414     9.27172    14.17522    16.32591
---------+--------------------------------------------------------------------
combined |    2707    15.57284    .1856003    9.656566    15.20891    15.93677
---------+--------------------------------------------------------------------
diff |            .3606459    .6020106               -.8198013    1.541093
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.5991
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.7254         Pr(|T| > |t|) = 0.5492          Pr(T > t) = 0.2746

ttest of wage by missingness of edu

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2412    71484.16    778.2065    38219.37    69958.14    73010.19
1 |     289    71575.65    2187.928    37194.77     67269.3    75882.01
---------+--------------------------------------------------------------------
combined |    2701    71493.95    733.1819     38104.3     70056.3    72931.61
---------+--------------------------------------------------------------------
diff |           -91.48891    2372.352               -4743.299    4560.321
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.0386
Ho: diff = 0                                     degrees of freedom =     2699

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.4846         Pr(|T| > |t|) = 0.9692          Pr(T > t) = 0.5154

logit missingness of exp on female race urban edu wage

Iteration 0:   log likelihood = -654.79701
Iteration 1:   log likelihood = -653.43555
Iteration 2:   log likelihood = -653.43222
Iteration 3:   log likelihood = -653.43222

Logistic regression                               Number of obs   =       1982
LR chi2(5)      =       2.73
Prob > chi2     =     0.7416
Log likelihood = -653.43222                       Pseudo R2       =     0.0021

------------------------------------------------------------------------------
miss_exp |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
female |   .0225336   .1628237     0.14   0.890     -.296595    .3416622
race |  -.0595427   .0946498    -0.63   0.529    -.2450529    .1259675
urban |   .0187602   .1634337     0.11   0.909     -.301564    .3390845
edu |   .1058189   .0930734     1.14   0.256    -.0766016    .2882394
wage |  -2.42e-06   2.20e-06    -1.10   0.271    -6.73e-06    1.89e-06
_cons |  -2.216882   .2563097    -8.65   0.000     -2.71924   -1.714524
------------------------------------------------------------------------------

ttest of female by missingness of exp

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2707    .4931659    .0096109    .5000457    .4743204    .5120114
1 |     293    .5221843    .0292315    .5003622    .4646532    .5797154
---------+--------------------------------------------------------------------
combined |    3000        .496    .0091299    .5000674    .4780984    .5139016
---------+--------------------------------------------------------------------
diff |           -.0290184    .0307552               -.0893219    .0312851
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.9435
Ho: diff = 0                                     degrees of freedom =     2998

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.1727         Pr(|T| > |t|) = 0.3455          Pr(T > t) = 0.8273

ttest of race by missingness of exp

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2448    1.020425    .0163788    .8103788    .9883071    1.052543
1 |     259           1    .0504388    .8117356    .9006758    1.099324
---------+--------------------------------------------------------------------
combined |    2707    1.018471    .0155756    .8103808    .9879293    1.049012
---------+--------------------------------------------------------------------
diff |            .0204248    .0529598               -.0834209    .1242705
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.3857
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.6501         Pr(|T| > |t|) = 0.6998          Pr(T > t) = 0.3499

ttest of urban by missingness of exp

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2450    .6628571    .0095526    .4728306    .6441251    .6815892
1 |     277    .6570397    .0285735    .4755575    .6007901    .7132894
---------+--------------------------------------------------------------------
combined |    2727    .6622662    .0090582     .473024    .6445046    .6800278
---------+--------------------------------------------------------------------
diff |            .0058174    .0299902               -.0529884    .0646233
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.1940
Ho: diff = 0                                     degrees of freedom =     2725

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.5769         Pr(|T| > |t|) = 0.8462          Pr(T > t) = 0.4231

ttest of edu by missingness of exp

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2419    2.355105    .0185189    .9108219    2.318791     2.39142
1 |     262    2.381679    .0572124    .9260638    2.269023    2.494336
---------+--------------------------------------------------------------------
combined |    2681    2.357702     .017617     .912182    2.323158    2.392247
---------+--------------------------------------------------------------------
diff |            -.026574    .0593371               -.1429251    .0897771
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.4478
Ho: diff = 0                                     degrees of freedom =     2679

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.3271         Pr(|T| > |t|) = 0.6543          Pr(T > t) = 0.6729

ttest of wage by missingness of exp

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2432    71682.77    773.3145    38136.25    70166.35     73199.2
1 |     269    69786.84    2307.266    37841.97    65244.17    74329.51
---------+--------------------------------------------------------------------
combined |    2701    71493.95    733.1819     38104.3     70056.3    72931.61
---------+--------------------------------------------------------------------
diff |            1895.932    2448.559               -2905.309    6697.173
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.7743
Ho: diff = 0                                     degrees of freedom =     2699

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.7806         Pr(|T| > |t|) = 0.4388          Pr(T > t) = 0.2194

logit missingness of wage on female race urban edu exp

Iteration 0:   log likelihood = -647.94103
Iteration 1:   log likelihood = -645.05158
Iteration 2:   log likelihood =  -645.0361
Iteration 3:   log likelihood =  -645.0361

Logistic regression                               Number of obs   =       1979
LR chi2(5)      =       5.81
Prob > chi2     =     0.3252
Log likelihood =  -645.0361                       Pseudo R2       =     0.0045

------------------------------------------------------------------------------
miss_wage |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
female |   -.191566   .1570953    -1.22   0.223    -.4994672    .1163353
race |  -.1705262   .0959515    -1.78   0.076    -.3585876    .0175352
urban |  -.1708259   .1599631    -1.07   0.286    -.4843478     .142696
edu |   .0710834   .0886472     0.80   0.423     -.102662    .2448288
exp |   .0040734   .0079491     0.51   0.608    -.0115065    .0196534
_cons |  -2.049828   .2771956    -7.39   0.000    -2.593121   -1.506535
------------------------------------------------------------------------------

ttest of female by missingness of wage

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2701    .5012958    .0096225    .5000909    .4824277     .520164
1 |     299    .4481605    .0288081    .4981391    .3914674    .5048537
---------+--------------------------------------------------------------------
combined |    3000        .496    .0091299    .5000674    .4780984    .5139016
---------+--------------------------------------------------------------------
diff |            .0531353     .030468                -.006605    .1128755
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   1.7440
Ho: diff = 0                                     degrees of freedom =     2998

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.9594         Pr(|T| > |t|) = 0.0813          Pr(T > t) = 0.0406

ttest of race by missingness of wage

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2442    1.020885    .0164342    .8121201    .9886582    1.053111
1 |     265    .9962264    .0488572    .7953373    .9000271    1.092426
---------+--------------------------------------------------------------------
combined |    2707    1.018471    .0155756    .8103808    .9879293    1.049012
---------+--------------------------------------------------------------------
diff |            .0246581    .0524204               -.0781299    .1274461
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.4704
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.6809         Pr(|T| > |t|) = 0.6381          Pr(T > t) = 0.3191

ttest of urban by missingness of wage

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2458    .6647681    .0095237    .4721675    .6460928    .6834434
1 |     269    .6394052    .0293312    .4810681    .5816562    .6971542
---------+--------------------------------------------------------------------
combined |    2727    .6622662    .0090582     .473024    .6445046    .6800278
---------+--------------------------------------------------------------------
diff |            .0253629    .0303797               -.0342066    .0849324
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.8349
Ho: diff = 0                                     degrees of freedom =     2725

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.7981         Pr(|T| > |t|) = 0.4039          Pr(T > t) = 0.2019

ttest of edu by missingness of wage

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2412    2.357794    .0185831    .9126566    2.321354    2.394235
1 |     269    2.356877    .0554598    .9096083    2.247685     2.46607
---------+--------------------------------------------------------------------
combined |    2681    2.357702     .017617     .912182    2.323158    2.392247
---------+--------------------------------------------------------------------
diff |             .000917     .058647                -.114081    .1159151
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =   0.0156
Ho: diff = 0                                     degrees of freedom =     2679

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.5062         Pr(|T| > |t|) = 0.9875          Pr(T > t) = 0.4938

ttest of exp by missingness of wage

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 |    2432    15.51836    .1952193    9.627299    15.13555    15.90117
1 |     275    16.05461    .5979892    9.916529    14.87737    17.23184
---------+--------------------------------------------------------------------
combined |    2707    15.57284    .1856003    9.656566    15.20891    15.93677
---------+--------------------------------------------------------------------
diff |           -.5362457    .6143811                -1.74095    .6684581
------------------------------------------------------------------------------
diff = mean(0) - mean(1)                                      t =  -0.8728
Ho: diff = 0                                     degrees of freedom =     2705

Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
Pr(T < t) = 0.1914         Pr(|T| > |t|) = 0.3828          Pr(T > t) = 0.8086

.
.
. // set up trial imputation command just to get the individual regression commands
. mi set wide

. mi register imputed race-wage

. mi register regular female

. mi impute chained (logit) urban (mlogit) race (ologit) edu (regress) exp wage = i.female, dryrun

Conditional models:
urban: logit urban i.race exp wage i.edu i.female
race: mlogit race i.urban exp wage i.edu i.female
exp: regress exp i.urban i.race wage i.edu i.female
wage: regress wage i.urban i.race exp i.edu i.female
edu: ologit edu i.urban i.race exp wage i.female

.
. // test imputation model for race
. mlogit race exp wage i.edu i.urban i.female

Iteration 0:   log likelihood =   -1953.43
Iteration 1:   log likelihood = -1879.0566
Iteration 2:   log likelihood = -1877.6678
Iteration 3:   log likelihood = -1877.6668
Iteration 4:   log likelihood = -1877.6668

Multinomial logistic regression                   Number of obs   =       1779
LR chi2(14)     =     151.53
Prob > chi2     =     0.0000
Log likelihood = -1877.6668                       Pseudo R2       =     0.0388

------------------------------------------------------------------------------
race |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
0            |
exp |  -.0160644   .0075448    -2.13   0.033     -.030852   -.0012767
wage |   5.80e-06   2.07e-06     2.80   0.005     1.75e-06    9.85e-06
|
edu |
2  |  -.8129621   .1829418    -4.44   0.000    -1.171521   -.4544027
3  |  -1.593897   .1971316    -8.09   0.000    -1.980268   -1.207526
4  |   -2.72232   .2886243    -9.43   0.000    -3.288013   -2.156626
|
1.urban |   .7865707   .1339259     5.87   0.000     .5240808    1.049061
1.female |   .2893221   .1342653     2.15   0.031      .026167    .5524772
_cons |    .226939    .225613     1.01   0.314    -.2152544    .6691324
-------------+----------------------------------------------------------------
1            |
exp |   .0052958   .0071395     0.74   0.458    -.0086974    .0192891
wage |   6.69e-07   1.97e-06     0.34   0.734    -3.19e-06    4.53e-06
|
edu |
2  |  -.5144888   .1832349    -2.81   0.005    -.8736226    -.155355
3  |  -1.125629   .1949919    -5.77   0.000    -1.507806    -.743452
4  |  -1.307677   .2464598    -5.31   0.000    -1.790729   -.8246246
|
1.urban |   .4772458   .1266699     3.77   0.000     .2289775    .7255142
1.female |   .1276518   .1290494     0.99   0.323    -.1252803     .380584
_cons |    .253432    .220844     1.15   0.251    -.1794143    .6862783
-------------+----------------------------------------------------------------
2            |  (base outcome)
------------------------------------------------------------------------------

. // test for misspecification by adding interactions
. mlogit race (c.exp c.wage i.edu)##(i.female i.urban)

Iteration 0:   log likelihood =   -1953.43
Iteration 1:   log likelihood = -1873.2138
Iteration 2:   log likelihood = -1871.3005
Iteration 3:   log likelihood = -1871.2474
Iteration 4:   log likelihood = -1871.2473

Multinomial logistic regression                   Number of obs   =       1779
LR chi2(34)     =     164.37
Prob > chi2     =     0.0000
Log likelihood = -1871.2473                       Pseudo R2       =     0.0421

-------------------------------------------------------------------------------
race |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
0             |
exp |  -.0032793   .0152468    -0.22   0.830    -.0331624    .0266038
wage |   7.15e-06   4.01e-06     1.78   0.075    -7.13e-07     .000015
|
edu |
2  |  -.7354223   .3525541    -2.09   0.037    -1.426416    -.044429
3  |  -1.842053   .4005341    -4.60   0.000    -2.627085    -1.05702
4  |  -3.830948   1.095232    -3.50   0.000    -5.977563   -1.684332
|
1.female |   .1884419   .4158562     0.45   0.650    -.6266212    1.003505
1.urban |   1.056967   .4258448     2.48   0.013     .2223265    1.891608
|
female#c.exp |
1  |  -.0064319   .0151943    -0.42   0.672    -.0362122    .0233484
|
urban#c.exp |
1  |  -.0133441   .0161867    -0.82   0.410    -.0450693    .0183812
|
female#c.wage |
1  |   1.91e-06   4.18e-06     0.46   0.648    -6.29e-06    .0000101
|
urban#c.wage |
1  |  -3.07e-06   4.27e-06    -0.72   0.472    -.0000114    5.30e-06
|
edu#female |
2 1  |  -.0181898   .3925029    -0.05   0.963    -.7874813    .7511017
3 1  |  -.0172244   .4175976    -0.04   0.967    -.8357006    .8012518
4 1  |   .4977631   .6962046     0.71   0.475    -.8667728    1.862299
|
edu#urban |
2 1  |  -.0763182   .3951518    -0.19   0.847    -.8508015    .6981651
3 1  |   .3914775    .440372     0.89   0.374    -.4716358    1.254591
4 1  |   .8667237   1.155459     0.75   0.453    -1.397934    3.131381
|
_cons |   .0469868   .3959872     0.12   0.906    -.7291337    .8231074
--------------+----------------------------------------------------------------
1             |
exp |   .0117503   .0139703     0.84   0.400     -.015631    .0391316
wage |   6.92e-07   3.72e-06     0.19   0.852    -6.59e-06    7.98e-06
|
edu |
2  |  -.4485059   .3458792    -1.30   0.195    -1.126417    .2294049
3  |   -1.31316   .3798982    -3.46   0.001    -2.057747   -.5685735
4  |  -1.904266   .5852199    -3.25   0.001    -3.051275   -.7572556
|
1.female |   .1574212   .4125146     0.38   0.703    -.6510925    .9659349
1.urban |   .3765925    .421229     0.89   0.371    -.4490011    1.202186
|
female#c.exp |
1  |  -.0173227   .0144194    -1.20   0.230    -.0455843    .0109389
|
urban#c.exp |
1  |   .0034701    .015149     0.23   0.819    -.0262214    .0331615
|
female#c.wage |
1  |   3.64e-06   4.00e-06     0.91   0.363    -4.20e-06    .0000115
|
urban#c.wage |
1  |  -2.70e-06   4.03e-06    -0.67   0.503    -.0000106    5.20e-06
|
edu#female |
2 1  |   -.227974   .3945182    -0.58   0.563    -1.001215    .5452674
3 1  |  -.0181844    .414751    -0.04   0.965    -.8310815    .7947127
4 1  |   .4709082   .5673412     0.83   0.407    -.6410601    1.582876
|
edu#urban |
2 1  |   .0936678   .3968078     0.24   0.813    -.6840612    .8713968
3 1  |   .3483946   .4271216     0.82   0.415    -.4887483    1.185538
4 1  |   .3778603   .6590117     0.57   0.566     -.913779      1.6695
|
_cons |   .2636095   .3807272     0.69   0.489    -.4826021    1.009821
--------------+----------------------------------------------------------------
2             |  (base outcome)
-------------------------------------------------------------------------------

.
. // test imputation model for exp
. regress exp i.race wage i.edu i.urban i.female

Source |       SS       df       MS              Number of obs =    1779
-------------+------------------------------           F(  8,  1770) =   93.75
Model |  49906.8412     8  6238.35514           Prob > F      =  0.0000
Residual |   117780.77  1770  66.5428078           R-squared     =  0.2976
Total |  167687.611  1778  94.3124921           Root MSE      =  8.1574

------------------------------------------------------------------------------
exp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
race |
1  |   1.391616   .4799807     2.90   0.004      .450227    2.333004
2  |   1.031061   .4947771     2.08   0.037     .0606522     2.00147
|
wage |   .0001343   5.71e-06    23.50   0.000     .0001231    .0001455
|
edu |
2  |   -1.96545   .5537782    -3.55   0.000    -3.051578   -.8793225
3  |  -5.058849   .5979374    -8.46   0.000    -6.231586   -3.886111
4  |  -7.905853   .8106653    -9.75   0.000    -9.495815   -6.315891
|
1.urban |  -.5730682   .4234363    -1.35   0.176    -1.403556    .2574196
1.female |  -1.111798   .4256352    -2.61   0.009    -1.946598   -.2769972
_cons |   9.243531    .716064    12.91   0.000     7.839111    10.64795
------------------------------------------------------------------------------

. // test for misspecification with rvfplot
. // constraint line indicates exp>=0
. rvfplot, ylabel(-40 -20 0 20 40)

. graph export exp1.png, replace
(file exp1.png written in PNG format)

. predict exphat
(option xb assumed; fitted values)
(1018 missing values generated)

. predict expres, res
(1221 missing values generated)

. gen y=-exphat
(1018 missing values generated)

. scatter expres exphat || line y exphat, legend(order(2 "Exp>=0 Constraint"))

. graph export exp2.png, replace
(file exp2.png written in PNG format)

. drop expres exphat y

. //test for misspecification by adding interactions
. regress exp (i.race i.urban i.female)##(c.wage i.edu)

Source |       SS       df       MS              Number of obs =    1779
-------------+------------------------------           F( 24,  1754) =   32.28
Model |  51376.4689    24   2140.6862           Prob > F      =  0.0000
Residual |  116311.142  1754  66.3119396           R-squared     =  0.3064
Total |  167687.611  1778  94.3124921           Root MSE      =  8.1432

-------------------------------------------------------------------------------
exp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
race |
1  |   .8903704   1.271895     0.70   0.484     -1.60422    3.384961
2  |   1.418534    1.46223     0.97   0.332    -1.449363    4.286431
|
1.urban |  -1.905077    1.19761    -1.59   0.112     -4.25397    .4438155
1.female |  -.0698405   1.188445    -0.06   0.953    -2.400758    2.261077
wage |   .0001473   .0000136    10.87   0.000     .0001207    .0001739
|
edu |
2  |  -3.806439   1.314344    -2.90   0.004    -6.384285   -1.228592
3  |  -6.196198   1.456027    -4.26   0.000     -9.05193   -3.340466
4  |  -8.003504   2.559556    -3.13   0.002    -13.02361   -2.983403
|
race#c.wage |
1  |  -8.72e-06   .0000133    -0.65   0.513    -.0000349    .0000174
2  |  -.0000135    .000013    -1.04   0.296     -.000039    .0000119
|
race#edu |
1 2  |   2.488045   1.259945     1.97   0.048      .016893    4.959198
1 3  |  -.1736131   1.376548    -0.13   0.900     -2.87346    2.526234
1 4  |    2.29836    2.10793     1.09   0.276    -1.835959    6.432679
2 2  |   1.029303    1.46414     0.70   0.482    -1.842342    3.900947
2 3  |  -.0586898   1.510437    -0.04   0.969    -3.021136    2.903756
2 4  |   2.118492   2.158164     0.98   0.326    -2.114354    6.351337
|
urban#c.wage |
1  |   8.06e-06   .0000115     0.70   0.482    -.0000144    .0000306
|
urban#edu |
1 2  |   .8233918   1.191193     0.69   0.490    -1.512916      3.1597
1 3  |   1.802902    1.29168     1.40   0.163    -.7304935    4.336297
1 4  |  -3.443128    2.15643    -1.60   0.111    -7.672571    .7863152
|
female#c.wage |
1  |  -.0000233   .0000115    -2.02   0.044     -.000046   -6.71e-07
|
female#edu |
1 2  |   .5626791   1.188512     0.47   0.636    -1.768369    2.893728
1 3  |   .4449296   1.252109     0.36   0.722    -2.010853    2.900712
1 4  |   2.712876     1.8349     1.48   0.139    -.8859457    6.311698
|
_cons |   9.321907   1.382308     6.74   0.000     6.610763    12.03305
-------------------------------------------------------------------------------

.
.
. // test imputation model for wage
. regress wage i.race exp i.edu i.urban i.female

Source |       SS       df       MS              Number of obs =    1779
-------------+------------------------------           F(  8,  1770) =  145.49
Model |  1.0214e+12     8  1.2767e+11           Prob > F      =  0.0000
Residual |  1.5532e+12  1770   877509504           R-squared     =  0.3967
Total |  2.5746e+12  1778  1.4480e+09           Root MSE      =   29623

------------------------------------------------------------------------------
wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
race |
1  |  -4353.207   1744.074    -2.50   0.013    -7773.869   -932.5451
2  |  -4939.278   1795.106    -2.75   0.006    -8460.029   -1418.526
|
exp |   1771.135   75.35312    23.50   0.000     1623.344    1918.925
|
edu |
2  |   8345.912   2008.366     4.16   0.000     4406.894    12284.93
3  |   26875.02   2120.707    12.67   0.000     22715.66    31034.37
4  |   44200.82   2833.404    15.60   0.000     38643.65    49757.99
|
1.urban |    3737.22     1535.9     2.43   0.015     724.8518    6749.588
1.female |  -19496.65   1477.669   -13.19   0.000    -22394.81   -16598.49
_cons |   37847.82   2566.897    14.74   0.000     32813.35    42882.29
------------------------------------------------------------------------------

. // test for misspecification with rvfplot
. // constraint line indicates wage>=0
. rvfplot

. graph export wage.png, replace

(file wage.png written in PNG format)

. // test interactions
. regress wage (i.race i.urban i.female)##(c.exp i.edu)

Source |       SS       df       MS              Number of obs =    1779
-------------+------------------------------           F( 24,  1754) =   51.27
Model |  1.0615e+12    24  4.4230e+10           Prob > F      =  0.0000
Residual |  1.5130e+12  1754   862625195           R-squared     =  0.4123
Total |  2.5746e+12  1778  1.4480e+09           Root MSE      =   29370

------------------------------------------------------------------------------
wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
race |
1  |   409.0987   4761.366     0.09   0.932    -8929.451    9747.648
2  |    40.7097     5452.8     0.01   0.994    -10653.96    10735.38
|
1.urban |   2955.393   4512.636     0.65   0.513     -5895.32     11806.1
1.female |  -5986.333    4393.73    -1.36   0.173    -14603.83    2631.167
exp |   2086.047   188.9033    11.04   0.000     1715.547    2456.546
|
edu |
2  |   12596.53   4787.695     2.63   0.009     3206.339    21986.72
3  |    33416.6   5179.464     6.45   0.000     23258.02    43575.17
4  |   29270.41   9233.095     3.17   0.002     11161.38    47379.44
|
race#c.exp |
1  |  -367.0251   180.2227    -2.04   0.042     -720.499   -13.55117
2  |  -53.25982   182.8739    -0.29   0.771    -411.9335    305.4139
|
race#edu |
1 2  |  -482.0457   4541.364    -0.11   0.915    -9389.103    8425.011
1 3  |   1861.724   4875.674     0.38   0.703    -7701.021    11424.47
1 4  |   7840.737   7555.631     1.04   0.300    -6978.253    22659.73
2 2  |  -7391.542   5296.316    -1.40   0.163     -17779.3    2996.213
2 3  |  -4044.694   5390.587    -0.75   0.453    -14617.35    6527.957
2 4  |   3039.309   7774.405     0.39   0.696    -12208.77    18287.38
|
urban#c.exp |
1  |   119.1172   157.4763     0.76   0.450    -189.7437    427.9781
|
urban#edu |
1 2  |   644.4913   4304.863     0.15   0.881    -7798.712    9087.694
1 3  |  -7540.896   4566.843    -1.65   0.099    -16497.92    1416.132
1 4  |   25090.89   7625.385     3.29   0.001     10135.09    40046.69
|
female#c.exp |
1  |   -547.008   151.0542    -3.62   0.000    -843.2732   -250.7427
|
female#edu |
1 2  |  -6295.529   4285.177    -1.47   0.142    -14700.12    2109.064
1 3  |  -3410.924   4406.216    -0.77   0.439    -12052.91    5231.064
1 4  |  -19915.24   6326.051    -3.15   0.002    -32322.64   -7507.849
|
_cons |   29172.73   5272.194     5.53   0.000     18832.28    39513.17
------------------------------------------------------------------------------

.
.
. // test imputation model for edu
. ologit edu i.race exp wage i.urban i.female

Iteration 0:   log likelihood = -2295.6305
Iteration 1:   log likelihood =   -2021.07
Iteration 2:   log likelihood = -2013.1176
Iteration 3:   log likelihood = -2013.1071
Iteration 4:   log likelihood = -2013.1071

Ordered logistic regression                       Number of obs   =       1779
LR chi2(6)      =     565.05
Prob > chi2     =     0.0000
Log likelihood = -2013.1071                       Pseudo R2       =     0.1231

------------------------------------------------------------------------------
edu |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
race |
1  |   .5023646   .1109327     4.53   0.000     .2849405    .7197886
2  |   1.220383   .1136064    10.74   0.000     .9977182    1.443047
|
exp |  -.0604264   .0055268   -10.93   0.000    -.0712587    -.049594
wage |   .0000253   1.50e-06    16.89   0.000     .0000224    .0000282
1.urban |   .8322351   .0973248     8.55   0.000      .641482    1.022988
1.female |   .9733093   .0976488     9.97   0.000     .7819211    1.164697
-------------+----------------------------------------------------------------
/cut1 |   .6530809   .1597114                      .3400522    .9661095
/cut2 |   2.796932   .1712768                      2.461236    3.132628
/cut3 |    5.04024   .2008955                      4.646492    5.433988
------------------------------------------------------------------------------

. // test for misspecification by adding interactions
. ologit edu (i.race i.urban i.female)##(c.exp c.wage)

Iteration 0:   log likelihood = -2295.6305
Iteration 1:   log likelihood =  -2013.335
Iteration 2:   log likelihood = -2004.1162
Iteration 3:   log likelihood = -2004.0949
Iteration 4:   log likelihood = -2004.0949

Ordered logistic regression                       Number of obs   =       1779
LR chi2(14)     =     583.07
Prob > chi2     =     0.0000
Log likelihood = -2004.0949                       Pseudo R2       =     0.1270

-------------------------------------------------------------------------------
edu |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
race |
1  |   .1327298   .2534179     0.52   0.600    -.3639601    .6294198
2  |   1.003797   .2497778     4.02   0.000     .5142413    1.493352
|
1.urban |   1.366727   .2189527     6.24   0.000     .9375876    1.795866
1.female |    .618023   .2149086     2.88   0.004     .1968099    1.039236
exp |  -.0465445   .0139356    -3.34   0.001    -.0738578   -.0192312
wage |   .0000226   3.55e-06     6.38   0.000     .0000157    .0000296
|
race#c.exp |
1  |   .0009364   .0132133     0.07   0.944    -.0249611    .0268339
2  |  -.0018853   .0134862    -0.14   0.889    -.0283179    .0245472
|
race#c.wage |
1  |   5.02e-06   3.42e-06     1.47   0.142    -1.68e-06    .0000117
2  |   3.62e-06   3.35e-06     1.08   0.280    -2.95e-06    .0000102
|
urban#c.exp |
1  |  -.0137359   .0115411    -1.19   0.234    -.0363561    .0088844
|
urban#c.wage |
1  |  -4.85e-06   2.88e-06    -1.68   0.092    -.0000105    7.92e-07
|
female#c.exp |
1  |  -.0057173   .0108306    -0.53   0.598    -.0269449    .0155103
|
female#c.wage |
1  |   6.55e-06   2.81e-06     2.33   0.020     1.05e-06    .0000121
--------------+----------------------------------------------------------------
/cut1 |   .6270383   .2723209                      .0932992    1.160777
/cut2 |   2.779004   .2811126                      2.228033    3.329974
/cut3 |   5.047488   .2980929                      4.463237     5.63174
-------------------------------------------------------------------------------

.
. // test imputation model for urban
. logit urban i.race exp wage i.edu i.female

Iteration 0:   log likelihood = -1142.4725
Iteration 1:   log likelihood = -1075.0707
Iteration 2:   log likelihood = -1073.6056
Iteration 3:   log likelihood = -1073.6034
Iteration 4:   log likelihood = -1073.6034

Logistic regression                               Number of obs   =       1779
LR chi2(8)      =     137.74
Prob > chi2     =     0.0000
Log likelihood = -1073.6034                       Pseudo R2       =     0.0603

------------------------------------------------------------------------------
urban |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
race |
1  |  -.2954482   .1311706    -2.25   0.024    -.5525378   -.0383585
2  |  -.7792647   .1334625    -5.84   0.000    -1.040846    -.517683
|
exp |  -.0088777   .0063875    -1.39   0.165     -.021397    .0036416
wage |   4.29e-06   1.77e-06     2.42   0.015     8.23e-07    7.76e-06
|
edu |
2  |    .606303   .1394181     4.35   0.000     .3330484    .8795575
3  |    1.03064   .1574484     6.55   0.000     .7220471    1.339233
4  |   1.994752   .2554113     7.81   0.000     1.494155    2.495349
|
1.female |  -.0520909   .1138005    -0.46   0.647    -.2751358     .170954
_cons |   .1577303   .1863126     0.85   0.397    -.2074357    .5228963
------------------------------------------------------------------------------

. // test for misspecification by adding interactions
. logit urban (i.race i.female)##(c.exp c.wage i.edu)

Iteration 0:   log likelihood = -1142.4725
Iteration 1:   log likelihood = -1020.9319
Iteration 2:   log likelihood = -1015.6326
Iteration 3:   log likelihood = -1015.3365
Iteration 4:   log likelihood = -1015.3347
Iteration 5:   log likelihood = -1015.3347

Logistic regression                               Number of obs   =       1779
LR chi2(23)     =     254.28
Prob > chi2     =     0.0000
Log likelihood = -1015.3347                       Pseudo R2       =     0.1113

-------------------------------------------------------------------------------
urban |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
race |
1  |  -.7713392   .3795508    -2.03   0.042    -1.515245   -.0274332
2  |  -1.156397   .4315566    -2.68   0.007    -2.002233   -.3105619
|
1.female |   -1.51098   .3321957    -4.55   0.000    -2.162072   -.8598889
exp |  -.0231263   .0138578    -1.67   0.095    -.0502871    .0040344
wage |   5.59e-06   3.70e-06     1.51   0.131    -1.66e-06    .0000129
|
edu |
2  |  -.2098759   .2750943    -0.76   0.446    -.7490508     .329299
3  |  -.0925368   .3264801    -0.28   0.777    -.7324261    .5473525
4  |   .4987983   1.103814     0.45   0.651    -1.664638    2.662234
|
race#c.exp |
1  |   .0257714   .0168647     1.53   0.126    -.0072828    .0588256
2  |   .0187349   .0166967     1.12   0.262      -.01399    .0514598
|
race#c.wage |
1  |  -5.02e-07   4.54e-06    -0.11   0.912    -9.40e-06    8.39e-06
2  |   3.07e-06   4.36e-06     0.70   0.481    -5.47e-06    .0000116
|
race#edu |
1 2  |   .1629017   .3408843     0.48   0.633    -.5052193    .8310227
1 3  |   .0019093   .4077384     0.00   0.996    -.7972432    .8010619
1 4  |  -.5458701   1.203649    -0.45   0.650    -2.904979    1.813238
2 2  |   .0964406   .3983642     0.24   0.809     -.684339    .8772201
2 3  |  -.3518724   .4347531    -0.81   0.418    -1.203973     .500228
2 4  |  -.9799584   1.159742    -0.84   0.398     -3.25301    1.293093
|
female#c.exp |
1  |     -.0047   .0134151    -0.35   0.726    -.0309931    .0215931
|
female#c.wage |
1  |  -4.51e-06   3.73e-06    -1.21   0.227    -.0000118    2.81e-06
|
female#edu |
1 2  |   1.671398     .30319     5.51   0.000     1.077157     2.26564
1 3  |   2.788041   .3342213     8.34   0.000      2.13298    3.443103
1 4  |   4.461757   .5745605     7.77   0.000     3.335639    5.587875
|
_cons |   1.049029   .3234171     3.24   0.001     .4151428    1.682915
-------------------------------------------------------------------------------

.
.
. // refine models after reviewing results
. mi impute chained (logit) urban (mlogit) race (ologit) edu (pmm) exp wage, dryrun by(female)

Performing setup for each by() group:

-> female = 0
Conditional models:
exp: pmm exp i.urban i.race wage i.edu
urban: logit urban exp i.race wage i.edu
race: mlogit race exp i.urban wage i.edu
wage: pmm wage exp i.urban i.race i.edu
edu: ologit edu exp i.urban i.race wage

-> female = 1
Conditional models:
urban: logit urban wage i.race i.edu exp
wage: pmm wage i.urban i.race i.edu exp
race: mlogit race i.urban wage i.edu exp
edu: ologit edu i.urban wage i.race exp
exp: pmm exp i.urban wage i.race i.edu
.
. // test new models for convergence
. bysort female: reg exp i.urban i.race wage i.edu

----------------------------------------------------------------------------------------------------------------------------------
-> female = 0

Source |       SS       df       MS              Number of obs =     892
-------------+------------------------------           F(  7,   884) =   52.98
Model |   24670.002     7    3524.286           Prob > F      =  0.0000
Residual |  58807.4441   884   66.524258           R-squared     =  0.2955
Total |  83477.4461   891   93.689614           Root MSE      =  8.1562

------------------------------------------------------------------------------
exp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
1.urban |  -.5563624   .5784369    -0.96   0.336    -1.691632    .5789075
|
race |
1  |   1.799778   .6750693     2.67   0.008     .4748526    3.124704
2  |   .8400697   .7025656     1.20   0.232    -.5388214    2.218961
|
wage |   .0001445   7.61e-06    18.99   0.000     .0001295    .0001594
|
edu |
2  |  -2.074689   .7327518    -2.83   0.005    -3.512825   -.6365529
3  |  -5.124519    .816809    -6.27   0.000     -6.72763   -3.521407
4  |  -8.313709   1.337503    -6.22   0.000    -10.93876   -5.688656
|
_cons |   8.402518   .9403253     8.94   0.000     6.556987    10.24805
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------
-> female = 1

Source |       SS       df       MS              Number of obs =     887
-------------+------------------------------           F(  7,   879) =   29.74
Model |  13874.8765     7  1982.12521           Prob > F      =  0.0000
Residual |  58577.2689   879  66.6408065           R-squared     =  0.1915
Total |  72452.1454   886  81.7744305           Root MSE      =  8.1634

------------------------------------------------------------------------------
exp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
1.urban |  -.6693439   .6690433    -1.00   0.317    -1.982453    .6437649
|
race |
1  |   .9598328   .6860308     1.40   0.162    -.3866169    2.306282
2  |   1.164061   .7025743     1.66   0.098     -.214858     2.54298
|
wage |    .000121   8.71e-06    13.89   0.000     .0001039    .0001381
|
edu |
2  |  -1.867343   .8689937    -2.15   0.032    -3.572888   -.1617985
3  |   -4.84498   .9371085    -5.17   0.000    -6.684211   -3.005748
4  |  -7.366648   1.147116    -6.42   0.000    -9.618054   -5.115242
|
_cons |   8.896763   .8998265     9.89   0.000     7.130704    10.66282
------------------------------------------------------------------------------

. by female: logit urban exp i.race wage i.edu

----------------------------------------------------------------------------------------------------------------------------------
-> female = 0

Iteration 0:   log likelihood = -576.14858
Iteration 1:   log likelihood = -568.10266
Iteration 2:   log likelihood = -568.08745
Iteration 3:   log likelihood = -568.08745

Logistic regression                               Number of obs   =        892
LR chi2(7)      =      16.12
Prob > chi2     =     0.0240
Log likelihood = -568.08745                       Pseudo R2       =     0.0140

------------------------------------------------------------------------------
urban |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |  -.0083494    .008748    -0.95   0.340    -.0254951    .0087964
|
race |
1  |  -.0250462   .1793684    -0.14   0.889    -.3766018    .3265095
2  |  -.3721612   .1813618    -2.05   0.040    -.7276238   -.0166986
|
wage |   6.78e-06   2.36e-06     2.88   0.004     2.16e-06    .0000114
|
edu |
2  |  -.1155276   .1933929    -0.60   0.550    -.4945708    .2635156
3  |  -.2803377   .2178799    -1.29   0.198    -.7073745    .1466991
4  |  -.3283938   .3534695    -0.93   0.353    -1.021181    .3643936
|
_cons |   .5155881   .2385894     2.16   0.031     .0479615    .9832148
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------
-> female = 1

Iteration 0:   log likelihood = -566.19162
Iteration 1:   log likelihood = -450.72498
Iteration 2:   log likelihood = -445.73919
Iteration 3:   log likelihood = -445.57881
Iteration 4:   log likelihood = -445.57813
Iteration 5:   log likelihood = -445.57813

Logistic regression                               Number of obs   =        887
LR chi2(7)      =     241.23
Prob > chi2     =     0.0000
Log likelihood = -445.57813                       Pseudo R2       =     0.2130

------------------------------------------------------------------------------
urban |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |  -.0107927    .010163    -1.06   0.288    -.0307119    .0091265
|
race |
1  |  -.6841661   .2108969    -3.24   0.001    -1.097516   -.2708159
2  |  -1.259647   .2157675    -5.84   0.000    -1.682544   -.8367507
|
wage |   1.35e-06   2.93e-06     0.46   0.645    -4.39e-06    7.09e-06
|
edu |
2  |   1.619022   .2371928     6.83   0.000     1.154132    2.083911
3  |   2.681609   .2672965    10.03   0.000     2.157717      3.2055
4  |    4.43645   .4644515     9.55   0.000     3.526142    5.346758
|
_cons |   -.531204   .2617202    -2.03   0.042    -1.044166   -.0182419
------------------------------------------------------------------------------

. by female: mlogit race exp i.urban wage i.edu

----------------------------------------------------------------------------------------------------------------------------------
-> female = 0

Iteration 0:   log likelihood = -979.43224
Iteration 1:   log likelihood = -935.80472
Iteration 2:   log likelihood = -934.98446
Iteration 3:   log likelihood = -934.97944
Iteration 4:   log likelihood = -934.97944

Multinomial logistic regression                   Number of obs   =        892
LR chi2(12)     =      88.91
Prob > chi2     =     0.0000
Log likelihood = -934.97944                       Pseudo R2       =     0.0454

------------------------------------------------------------------------------
race |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
0            |
exp |  -.0274735   .0102909    -2.67   0.008    -.0476434   -.0073037
1.urban |   .0262569   .1794679     0.15   0.884    -.3254937    .3780074
wage |   6.28e-06   2.77e-06     2.27   0.023     8.60e-07    .0000117
|
edu |
2  |  -.4080685    .206086    -1.98   0.048    -.8119896   -.0041474
3  |  -.5017744   .2464439    -2.04   0.042    -.9847955   -.0187533
4  |  -1.492324   .5722452    -2.61   0.009    -2.613904   -.3707442
|
_cons |   .2462701   .2715063     0.91   0.364    -.2858725    .7784127
-------------+----------------------------------------------------------------
1            |  (base outcome)
-------------+----------------------------------------------------------------
2            |
exp |  -.0143098   .0102621    -1.39   0.163    -.0344231    .0058036
1.urban |  -.3510997   .1744316    -2.01   0.044    -.6929795     -.00922
wage |   9.35e-07   2.75e-06     0.34   0.734    -4.46e-06    6.33e-06
|
edu |
2  |   .3857799   .2450301     1.57   0.115    -.0944702      .86603
3  |   1.092693   .2658459     4.11   0.000     .5716446    1.613741
4  |   1.673159   .4094672     4.09   0.000     .8706185      2.4757
|
_cons |   -.254461   .2948447    -0.86   0.388     -.832346    .3234241
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------
-> female = 1

Iteration 0:   log likelihood = -973.97087
Iteration 1:   log likelihood = -934.18474
Iteration 2:   log likelihood = -933.62658
Iteration 3:   log likelihood = -933.62647
Iteration 4:   log likelihood = -933.62647

Multinomial logistic regression                   Number of obs   =        887
LR chi2(12)     =      80.69
Prob > chi2     =     0.0000
Log likelihood = -933.62647                       Pseudo R2       =     0.0414

------------------------------------------------------------------------------
race |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
0            |
exp |  -.0180992   .0106928    -1.69   0.091    -.0390567    .0028584
1.urban |     1.2596   .2155262     5.84   0.000     .8371769    1.682024
wage |   6.92e-06   3.05e-06     2.27   0.023     9.36e-07    .0000129
|
edu |
2  |  -1.015529   .2869699    -3.54   0.000     -1.57798   -.4530781
3  |  -1.873085   .3132415    -5.98   0.000    -2.487027   -1.259143
4  |  -2.917669   .3964615    -7.36   0.000    -3.694719   -2.140619
|
_cons |   .3468113    .282675     1.23   0.220    -.2072216    .9008442
-------------+----------------------------------------------------------------
1            |
exp |  -.0030983   .0100496    -0.31   0.758    -.0227951    .0165985
1.urban |   .5720785   .1999865     2.86   0.004     .1801122    .9640448
wage |   2.49e-06   2.89e-06     0.86   0.388    -3.17e-06    8.16e-06
|
edu |
2  |  -.7094723   .2814518    -2.52   0.012    -1.261108    -.157837
3  |  -1.229685   .3028501    -4.06   0.000    -1.823261     -.63611
4  |  -1.319989   .3548621    -3.72   0.000    -2.015506   -.6244719
|
_cons |   .4289424    .276908     1.55   0.121    -.1137873     .971672
-------------+----------------------------------------------------------------
2            |  (base outcome)
------------------------------------------------------------------------------

. by female: reg wage exp i.urban i.race i.edu

----------------------------------------------------------------------------------------------------------------------------------
-> female = 0

Source |       SS       df       MS              Number of obs =     892
-------------+------------------------------           F(  7,   884) =   74.03
Model |  4.7872e+11     7  6.8388e+10           Prob > F      =  0.0000
Residual |  8.1660e+11   884   923758881           R-squared     =  0.3696
Total |  1.2953e+12   891  1.4538e+09           Root MSE      =   30393

------------------------------------------------------------------------------
wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |   2005.904   105.6246    18.99   0.000       1798.6    2213.208
1.urban |   6217.463   2146.452     2.90   0.004     2004.727     10430.2
|
race |
1  |  -5654.544     2518.5    -2.25   0.025    -10597.48   -711.6073
2  |  -4775.141   2615.229    -1.83   0.068    -9907.923    357.6405
|
edu |
2  |   10720.55   2719.075     3.94   0.000     5383.955    16057.15
3  |   28231.17   2962.326     9.53   0.000     22417.16    34045.18
4  |    50933.9   4794.995    10.62   0.000        41523     60344.8
|
_cons |   30542.89   3511.689     8.70   0.000     23650.67    37435.11
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------
-> female = 1

Source |       SS       df       MS              Number of obs =     887
-------------+------------------------------           F(  7,   879) =   54.13
Model |  3.1047e+11     7  4.4353e+10           Prob > F      =  0.0000
Residual |  7.2028e+11   879   819436657           R-squared     =  0.3012
Total |  1.0308e+12   886  1.1634e+09           Root MSE      =   28626

------------------------------------------------------------------------------
wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |   1488.074   107.0972    13.89   0.000     1277.878     1698.27
1.urban |   1266.729   2347.021     0.54   0.590     -3339.69    5873.148
|
race |
1  |  -3442.619   2405.519    -1.43   0.153    -8163.851    1278.613
2  |  -5492.975   2460.533    -2.23   0.026    -10322.18   -663.7687
|
edu |
2  |   5988.045   3048.533     1.96   0.050     4.792275     11971.3
3  |   26068.51   3217.693     8.10   0.000     19753.25    32383.77
4  |   41106.03   3875.211    10.61   0.000     33500.29    48711.78
|
_cons |   25087.89   3216.737     7.80   0.000     18774.51    31401.27
------------------------------------------------------------------------------

. by female: ologit edu exp i.urban i.race wage

----------------------------------------------------------------------------------------------------------------------------------
-> female = 0

Iteration 0:   log likelihood = -1092.3176
Iteration 1:   log likelihood = -986.99706
Iteration 2:   log likelihood =  -984.5232
Iteration 3:   log likelihood = -984.51851
Iteration 4:   log likelihood = -984.51851

Ordered logistic regression                       Number of obs   =        892
LR chi2(5)      =     215.60
Prob > chi2     =     0.0000
Log likelihood = -984.51851                       Pseudo R2       =     0.0987

------------------------------------------------------------------------------
edu |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |   -.059575   .0078897    -7.55   0.000    -.0750385   -.0441115
1.urban |  -.1649053   .1336636    -1.23   0.217    -.4268811    .0970705
|
race |
1  |    .420148   .1571087     2.67   0.007     .1122206    .7280753
2  |   1.275139   .1623148     7.86   0.000     .9570078     1.59327
|
wage |   .0000242   2.09e-06    11.58   0.000     .0000201    .0000283
-------------+----------------------------------------------------------------
/cut1 |  -.1678223    .205886                     -.5713513    .2357068
/cut2 |   2.064867    .217243                      1.639079    2.490656
/cut3 |   4.553283   .2681615                      4.027696     5.07887
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------
-> female = 1

Iteration 0:   log likelihood = -1172.2304
Iteration 1:   log likelihood = -973.18412
Iteration 2:   log likelihood = -964.74324
Iteration 3:   log likelihood = -964.72408
Iteration 4:   log likelihood = -964.72408

Ordered logistic regression                       Number of obs   =        887
LR chi2(5)      =     415.01
Prob > chi2     =     0.0000
Log likelihood = -964.72408                       Pseudo R2       =     0.1770

------------------------------------------------------------------------------
edu |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
exp |  -.0555539   .0078704    -7.06   0.000    -.0709795   -.0401282
1.urban |   1.938634   .1527146    12.69   0.000     1.639319    2.237949
|
race |
1  |   .6760943   .1598405     4.23   0.000     .3628128    .9893759
2  |   1.299797   .1621729     8.01   0.000      .981944     1.61765
|
wage |   .0000265   2.22e-06    11.93   0.000     .0000221    .0000309
-------------+----------------------------------------------------------------
/cut1 |   .5531119   .1965451                      .1678906    .9383332
/cut2 |   2.799431   .2172816                      2.373567    3.225295
/cut3 |   5.078792   .2594613                      4.570257    5.587326
------------------------------------------------------------------------------

. // for real work you would explore misspecification of refined models as well
.
.
. // test convergence of imputation process
. // since by() and savetrace() don't get along right now, we'll remove by() then throw away these imputations and do them with by
> () but no savetrace().
. preserve

. mi impute chained (logit) urban (mlogit) race (ologit) edu (pmm) exp wage = female, add(5) rseed(88) savetrace(extrace, replace)
>  burnin(100)

Conditional models:
urban: logit urban i.race exp wage i.edu female
race: mlogit race i.urban exp wage i.edu female
exp: pmm exp i.urban i.race wage i.edu female
wage: pmm wage i.urban i.race exp i.edu female
edu: ologit edu i.urban i.race exp wage female

Performing chained iterations ...

Multivariate imputation                     Imputations =        5
Imputed: m=1 through m=5                        updated =        0

Initialization: monotone                     Iterations =      500
burn-in =      100

urban: logistic regression
race: multinomial logistic regression
edu: ordered logistic regression
exp: predictive mean matching
wage: predictive mean matching

------------------------------------------------------------------
|               Observations per m
|----------------------------------------------
Variable |   Complete   Incomplete   Imputed |     Total
-------------------+-----------------------------------+----------
urban |       2727          273       273 |      3000
race |       2707          293       293 |      3000
edu |       2681          319       319 |      3000
exp |       2707          293       293 |      3000
wage |       2701          299       299 |      3000
------------------------------------------------------------------
(complete + incomplete = total; imputed is the minimum across m
of the number of filled-in observations.)

.
. use extrace, replace
(Summaries of imputed values from -mi impute chained-)

. reshape wide *mean *sd, i(iter) j(m)
(note: j = 1 2 3 4 5)

Data                               long   ->   wide
-----------------------------------------------------------------------------
Number of obs.                      505   ->     101
Number of variables                  12   ->      51
j variable (5 values)                 m   ->   (dropped)
xij variables:
urban_mean   ->   urban_mean1 urban_mean2 ... urban_mean5
race_mean   ->   race_mean1 race_mean2 ... race_mean5
exp_mean   ->   exp_mean1 exp_mean2 ... exp_mean5
wage_mean   ->   wage_mean1 wage_mean2 ... wage_mean5
edu_mean   ->   edu_mean1 edu_mean2 ... edu_mean5
urban_sd   ->   urban_sd1 urban_sd2 ... urban_sd5
race_sd   ->   race_sd1 race_sd2 ... race_sd5
exp_sd   ->   exp_sd1 exp_sd2 ... exp_sd5
wage_sd   ->   wage_sd1 wage_sd2 ... wage_sd5
edu_sd   ->   edu_sd1 edu_sd2 ... edu_sd5
-----------------------------------------------------------------------------

. tsset iter
time variable:  iter, 0 to 100
delta:  1 unit

. tsline exp_mean*, title("Mean of Imputed Values of Experience") note("Each line is for one imputation") legend(off)

. graph export conv1.png, replace
(file conv1.png written in PNG format)

. tsline exp_sd*, title("Standard Deviation of Imputed Values of Experience") note("Each line is for one imputation") legend(off)

. graph export conv2.png, replace
(file conv2.png written in PNG format)

. restore

.
.
. // "real" imputation
. mi impute chained (logit) urban (mlogit) race (ologit) edu (pmm) exp wage = i.female, add(5) rseed(88) by(female)

Performing setup for each by() group:

-> female = 0
Conditional models:
exp: pmm exp i.urban i.race wage i.edu i.female
urban: logit urban exp i.race wage i.edu i.female
race: mlogit race exp i.urban wage i.edu i.female
wage: pmm wage exp i.urban i.race i.edu i.female
edu: ologit edu exp i.urban i.race wage i.female

-> female = 1
Conditional models:
urban: logit urban wage i.race i.edu exp i.female
wage: pmm wage i.urban i.race i.edu exp i.female
race: mlogit race i.urban wage i.edu exp i.female
edu: ologit edu i.urban wage i.race exp i.female
exp: pmm exp i.urban wage i.race i.edu i.female

Performing imputation for each by() group:

-> female = 0
Performing chained iterations ...

-> female = 1
Performing chained iterations ...

Multivariate imputation                     Imputations =        5
Imputed: m=1 through m=5                        updated =        0

Initialization: monotone                     Iterations =       50
burn-in =       10

urban: logistic regression
race: multinomial logistic regression
edu: ordered logistic regression
exp: predictive mean matching
wage: predictive mean matching

------------------------------------------------------------------
|               Observations per m
by()               |----------------------------------------------
Variable |   Complete   Incomplete   Imputed |     Total
-------------------+-----------------------------------+----------
female = 0         |                                   |
urban |       1369          143       143 |      1512
race |       1364          148       148 |      1512
edu |       1345          167       167 |      1512
exp |       1372          140       140 |      1512
wage |       1347          165       165 |      1512
|                                   |
female = 1         |                                   |
urban |       1358          130       130 |      1488
race |       1343          145       145 |      1488
edu |       1336          152       152 |      1488
exp |       1335          153       153 |      1488
wage |       1354          134       134 |      1488
|                                   |
-------------------+-----------------------------------+----------
Overall            |                                   |
urban |       2727          273       273 |      3000
race |       2707          293       293 |      3000
edu |       2681          319       319 |      3000
exp |       2707          293       293 |      3000
wage |       2701          299       299 |      3000
------------------------------------------------------------------
(complete + incomplete = total; imputed is the minimum across m
of the number of filled-in observations.)

.
. // check if imputed values match observed values
. foreach var of varlist urban race edu {
2.         mi xeq 0: tab `var'
3.         mi xeq 1/5: tab `var' if miss_`var'
4. }

m=0 data:
-> tab urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        921       33.77       33.77
1 |      1,806       66.23      100.00
------------+-----------------------------------
Total |      2,727      100.00

m=1 data:
-> tab urban if miss_urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        102       37.36       37.36
1 |        171       62.64      100.00
------------+-----------------------------------
Total |        273      100.00

m=2 data:
-> tab urban if miss_urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |         97       35.53       35.53
1 |        176       64.47      100.00
------------+-----------------------------------
Total |        273      100.00

m=3 data:
-> tab urban if miss_urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        107       39.19       39.19
1 |        166       60.81      100.00
------------+-----------------------------------
Total |        273      100.00

m=4 data:
-> tab urban if miss_urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        102       37.36       37.36
1 |        171       62.64      100.00
------------+-----------------------------------
Total |        273      100.00

m=5 data:
-> tab urban if miss_urban

urban |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |         97       35.53       35.53
1 |        176       64.47      100.00
------------+-----------------------------------
Total |        273      100.00

m=0 data:
-> tab race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        864       31.92       31.92
1 |        929       34.32       66.24
2 |        914       33.76      100.00
------------+-----------------------------------
Total |      2,707      100.00

m=1 data:
-> tab race if miss_race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |         97       33.11       33.11
1 |        113       38.57       71.67
2 |         83       28.33      100.00
------------+-----------------------------------
Total |        293      100.00

m=2 data:
-> tab race if miss_race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        107       36.52       36.52
1 |         88       30.03       66.55
2 |         98       33.45      100.00
------------+-----------------------------------
Total |        293      100.00

m=3 data:
-> tab race if miss_race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        101       34.47       34.47
1 |         98       33.45       67.92
2 |         94       32.08      100.00
------------+-----------------------------------
Total |        293      100.00

m=4 data:
-> tab race if miss_race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |        119       40.61       40.61
1 |         77       26.28       66.89
2 |         97       33.11      100.00
------------+-----------------------------------
Total |        293      100.00

m=5 data:
-> tab race if miss_race

race |      Freq.     Percent        Cum.
------------+-----------------------------------
0 |         76       25.94       25.94
1 |        116       39.59       65.53
2 |        101       34.47      100.00
------------+-----------------------------------
Total |        293      100.00

m=0 data:
-> tab edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |        511       19.06       19.06
High School |        996       37.15       56.21
Bachelors |        878       32.75       88.96
Advanced Degree |        296       11.04      100.00
----------------+-----------------------------------
Total |      2,681      100.00

m=1 data:
-> tab edu if miss_edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |         50       15.67       15.67
High School |        135       42.32       57.99
Bachelors |         98       30.72       88.71
Advanced Degree |         36       11.29      100.00
----------------+-----------------------------------
Total |        319      100.00

m=2 data:
-> tab edu if miss_edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |         53       16.61       16.61
High School |        129       40.44       57.05
Bachelors |        109       34.17       91.22
Advanced Degree |         28        8.78      100.00
----------------+-----------------------------------
Total |        319      100.00

m=3 data:
-> tab edu if miss_edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |         60       18.81       18.81
High School |        124       38.87       57.68
Bachelors |        105       32.92       90.60
Advanced Degree |         30        9.40      100.00
----------------+-----------------------------------
Total |        319      100.00

m=4 data:
-> tab edu if miss_edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |         62       19.44       19.44
High School |        124       38.87       58.31
Bachelors |         93       29.15       87.46
Advanced Degree |         40       12.54      100.00
----------------+-----------------------------------
Total |        319      100.00

m=5 data:
-> tab edu if miss_edu

edu |      Freq.     Percent        Cum.
----------------+-----------------------------------
< High School |         55       17.24       17.24
High School |        138       43.26       60.50
Bachelors |         93       29.15       89.66
Advanced Degree |         33       10.34      100.00
----------------+-----------------------------------
Total |        319      100.00

.
. foreach var of varlist wage exp {
2.         mi xeq 0: sum `var'
3.         mi xeq 1/5: sum `var' if miss_`var'
4.         mi xeq 0: kdensity `var'; graph export chk`var'0.png, replace
5.         forval i=1/5 {
6.                 mi xeq `i': kdensity `var' if miss_`var'; graph export chk`var'`i'.png, replace
7.         }
8. }

m=0 data:
-> sum wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |      2701    71493.95     38104.3          0   227465.2

m=1 data:
-> sum wage if miss_wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |       299    73701.88    38620.86          0   192810.8

m=2 data:
-> sum wage if miss_wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |       299    75122.22    38976.49          0   193577.9

m=3 data:
-> sum wage if miss_wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |       299    73354.54    40547.16          0   193577.9

m=4 data:
-> sum wage if miss_wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |       299    75166.36    40163.56          0   193577.9

m=5 data:
-> sum wage if miss_wage

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
wage |       299    75681.66    41793.81          0   198598.6

m=0 data:
-> kdensity wage
-> graph export chkwage0.png, replace
(file chkwage0.png written in PNG format)

m=1 data:
-> kdensity wage if miss_wage
-> graph export chkwage1.png, replace
(file chkwage1.png written in PNG format)

m=2 data:
-> kdensity wage if miss_wage
-> graph export chkwage2.png, replace
(file chkwage2.png written in PNG format)

m=3 data:
-> kdensity wage if miss_wage
-> graph export chkwage3.png, replace
(file chkwage3.png written in PNG format)

m=4 data:
-> kdensity wage if miss_wage
-> graph export chkwage4.png, replace
(file chkwage4.png written in PNG format)

m=5 data:
-> kdensity wage if miss_wage
-> graph export chkwage5.png, replace
(file chkwage5.png written in PNG format)

m=0 data:
-> sum exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |      2707    15.57284    9.656566          0    47.8623

m=1 data:
-> sum exp if miss_exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |       293    14.98541     10.0319          0   46.35374

m=2 data:
-> sum exp if miss_exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |       293    15.42685    10.09567          0   46.35374

m=3 data:
-> sum exp if miss_exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |       293    15.19209    9.870792          0   41.14571

m=4 data:
-> sum exp if miss_exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |       293    14.67198    10.40626          0    47.8623

m=5 data:
-> sum exp if miss_exp

Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
exp |       293    14.94231    9.530698          0   46.35374

m=0 data:
-> kdensity exp
-> graph export chkexp0.png, replace
(file chkexp0.png written in PNG format)

m=1 data:
-> kdensity exp if miss_exp
-> graph export chkexp1.png, replace
(file chkexp1.png written in PNG format)

m=2 data:
-> kdensity exp if miss_exp
-> graph export chkexp2.png, replace
(file chkexp2.png written in PNG format)

m=3 data:
-> kdensity exp if miss_exp
-> graph export chkexp3.png, replace
(file chkexp3.png written in PNG format)

m=4 data:
-> kdensity exp if miss_exp
-> graph export chkexp4.png, replace
(file chkexp4.png written in PNG format)

m=5 data:
-> kdensity exp if miss_exp
-> graph export chkexp5.png, replace
(file chkexp5.png written in PNG format)

.
. save mi1,replace
file mi1.dta saved

. log close
name:
log:  \sscc\pubs\mi\miex.log
log type:  text
closed on:  17 Aug 2012, 13:11:21
----------------------------------------------------------------------------------------------------------------------------------

```

Last Revised: 8/17/2012