Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • missing wald-test after mi:estimate heckprob

    Dear all!

    I am running a heckprob model, with the mi: estimate command with the cmdok option:

    Code:
    mi estimate, cmdok: heckprobit nonviolent_success lagfpe lagpolyarchy lnlaggdp lnlagpop lnlagmilper i.region lagtimesv lagtimesv2 lagtimesv3 lagtimesnv lagtimesnv2 lagtimesnv3 coldwar, select(nonongoing=lagfpe lagpolyarchy lnlaggdp lnlagpop lnlagmilper i.region lagtimesv lagtimesv2 lagtimesv3 lagtimesnv lagtimesnv2 lagtimesnv3 coldwar lagelection) vce(cluster country_name)
    Which produces the following results:

    Code:
    Multiple-imputation estimates                   Imputations       =         10
    Probit model with sample selection              Number of obs     =     11,973
                                                    Average RVI       =     0.2167
                                                    Largest FMI       =     0.4430
                                                    DF:     min       =      50.67
                                                            avg       =  10,593.00
    DF adjustment:   Large sample                           max       = 178,342.58
                                                    F(  16,      .)   =          .
    Within VCE type:       Robust                   Prob > F          =          .
    
                               (Within VCE adjusted for 195 clusters in country_name)
    ---------------------------------------------------------------------------------
                    | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
    ----------------+----------------------------------------------------------------
    violent_success |
             lagfpe |   -.792135    .470048    -1.69   0.092    -1.713801    .1295309
       lagpolyarchy |   .1639627   .3715389     0.44   0.661      -.58205    .9099753
           lnlaggdp |  -.1644244    .071781    -2.29   0.025    -.3073087     -.02154
           lnlagpop |   -.144808   .1143908    -1.27   0.207    -.3705178    .0809018
        lnlagmilper |   .1246189    .082104     1.52   0.130    -.0366877    .2859256
                    |
             region |
          Americas  |   .4410626   .2548345     1.73   0.084    -.0584481    .9405733
              Asia  |  -.1329461   .1615759    -0.82   0.411    -.4497744    .1838822
            Europe  |  -.1340773   .3896208    -0.34   0.731    -.8978504    .6296959
           Oceania  |   .5885227    .822563     0.72   0.474    -1.023821    2.200867
                    |
          lagtimesv |   .1544413   .1543353     1.00   0.317    -.1483351    .4572176
         lagtimesv2 |  -.0057165    .005885    -0.97   0.331    -.0172572    .0058242
         lagtimesv3 |   .0000654   .0000632     1.04   0.301    -.0000585    .0001894
         lagtimesnv |  -.0090928   .0162873    -0.56   0.577    -.0410298    .0228442
        lagtimesnv2 |   .0001368   .0003064     0.45   0.655     -.000465    .0007385
        lagtimesnv3 |  -1.16e-06   1.70e-06    -0.68   0.498    -4.52e-06    2.21e-06
            coldwar |  -.5870651   .1419212    -4.14   0.000    -.8653189   -.3088113
              _cons |   4.443759   1.877056     2.37   0.020     .7269174    8.160602
    ----------------+----------------------------------------------------------------
    vongoing        |
             lagfpe |  -.1599916   .2543776    -0.63   0.529    -.6588976    .3389143
       lagpolyarchy |  -.0677608    .131671    -0.51   0.607    -.3261963    .1906747
           lnlaggdp |   .0064057   .0306551     0.21   0.835    -.0540294    .0668408
           lnlagpop |   .2379262   .0434689     5.47   0.000     .1523928    .3234595
        lnlagmilper |   -.108601   .0426838    -2.54   0.012    -.1929865   -.0242155
                    |
             region |
          Americas  |  -.1054436   .1397666    -0.75   0.451    -.3793829    .1684957
              Asia  |   .1732049   .1042689     1.66   0.097    -.0311775    .3775873
            Europe  |  -.1549118   .1630714    -0.95   0.342    -.4745931    .1647695
           Oceania  |  -.3968091   .3216496    -1.23   0.217    -1.027238    .2336195
                    |
          lagtimesv |  -.4439047   .0209216   -21.22   0.000    -.4850612   -.4027482
         lagtimesv2 |    .015747   .0010678    14.75   0.000     .0136345    .0178595
         lagtimesv3 |  -.0001607   .0000149   -10.80   0.000    -.0001902   -.0001311
         lagtimesnv |   .0283246   .0066913     4.23   0.000     .0151758    .0414735
        lagtimesnv2 |  -.0002824   .0001476    -1.91   0.057    -.0005738    9.09e-06
        lagtimesnv3 |   1.15e-06   8.00e-07     1.44   0.154    -4.41e-07    2.74e-06
            coldwar |   .4390622   .0854265     5.14   0.000     .2715571    .6065673
        lagelection |   .0455585   .0636749     0.72   0.474     -.079255    .1703719
              _cons |  -3.481097   .6556771    -5.31   0.000    -4.776122   -2.186072
    ----------------+----------------------------------------------------------------
            /athrho |   -.355242   .4811313    -0.74   0.461    -1.299587    .5891031
    ----------------+----------------------------------------------------------------
                rho |  -.3410162   .4251796                     -.8616169    .5292503
    However, no Wald test is reported for the finished output; this is only reported on each M regression. What is the best option here? Is there any way to report a Wald test for the final output, or should I just compare the Wald test for each of the imputed datasets/ rely on the rho only?

    All suggestions are appreciated!
Working...
X