Accessing coefficients with _b[var] after mi estimate

[Jeremy Reynolds]

Dear Stata Forum,

I would like to access the coefficients from a regression estimated with multiply imputed data. After a normal regression with normal data, I can do this with the with _b[var] construction. However, as shown below, this does not seem to work if the regression has been estimated with the "mi estimate" prefix. Is there an easy way to retrieve individual coefficients after mi estimate?

Thanks,

Jeremy

. use http://www.stata-press.com/data/r13/mheart1s20
(Fictional heart attack data; bmi missing)

.
. mi estimate: logit attack smokes age bmi hsgrad female

Multiple-imputation estimates                   Imputations       =         20
Logistic regression                             Number of obs     =        154
                                                Average RVI       =     0.0312
                                                Largest FMI       =     0.1355
DF adjustment:   Large sample                   DF:     min       =   1,060.38
                                                        avg       = 223,362.56
                                                        max       = 493,335.88
Model F test:       Equal FMI                   F(   5,71379.3)   =       3.59
Within VCE type:          OIM                   Prob > F          =     0.0030

------------------------------------------------------------------------------
      attack |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      smokes |   1.198595   .3578195     3.35   0.001     .4972789    1.899911
         age |   .0360159   .0154399     2.33   0.020     .0057541    .0662776
         bmi |   .1039416   .0476136     2.18   0.029      .010514    .1973692
      hsgrad |   .1578992   .4049257     0.39   0.697    -.6357464    .9515449
      female |  -.1067433   .4164735    -0.26   0.798    -.9230191    .7095326
       _cons |  -5.478143   1.685075    -3.25   0.001    -8.782394   -2.173892
------------------------------------------------------------------------------

. di _b[female]
no variables defined
r(111);

end of do-file

r(111);

[Carlo Lazzaro]

Jeremy:
thanks for providing a reproducible example.
You may want to access the coefficients you're interested in via -mat list-, as in the following example:

. use http://www.stata-press.com/data/r13/mheart1s20
(Fictional heart attack data; bmi missing)

. mi estimate: logit attack smokes age bmi hsgrad female

Multiple-imputation estimates                     Imputations     =         20
Logistic regression                               Number of obs   =        154
                                                  Average RVI     =     0.0312
                                                  Largest FMI     =     0.1355
DF adjustment:   Large sample                     DF:     min     =    1060.38
                                                          avg     =  223362.56
                                                          max     =  493335.88
Model F test:       Equal FMI                     F(   5,71379.3) =       3.59
Within VCE type:          OIM                     Prob > F        =     0.0030

------------------------------------------------------------------------------
      attack |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      smokes |   1.198595   .3578195     3.35   0.001     .4972789    1.899911
         age |   .0360159   .0154399     2.33   0.020     .0057541    .0662776
         bmi |   .1039416   .0476136     2.18   0.029      .010514    .1973692
      hsgrad |   .1578992   .4049257     0.39   0.697    -.6357464    .9515449
      female |  -.1067433   .4164735    -0.26   0.798    -.9230191    .7095326
       _cons |  -5.478143   1.685075    -3.25   0.001    -8.782394   -2.173892
------------------------------------------------------------------------------



. mat list e(b_mi)

e(b_mi)[1,6]
        attack:     attack:     attack:     attack:     attack:     attack:
        smokes         age         bmi      hsgrad      female       _cons
r1   1.1985949   .03601589   .10394161   .15789924  -.10674329  -5.4781425

[daniel klein]

Another option is specifying the post option with mi estimate. This has potentially serious consequence for anything other than obtaining individual coefficients.

Best
Daniel

[Jeremy Reynolds]

Thank you both for your suggestions. The post option suggested by Daniel makes the _b[var] approach work just fine--thank you for the warning about the potential hazards of using the post option. Accessing the matrix as suggested by Carlo works well too.

Jeremy