Announcement

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

  • Very high coefficient 2SLS

    Hi

    I have run a 2SLS regression with the dependent variable, self reported health (excellent, good, very good, good, fair, poor). My independent variables include exercise(ex1 below). The results show a very high coefficient of exercise. Is there something I've done wrong- - this appears somewhat strange to me?

    My results are shown below.

    Code:
    
    . ivregress 2sls  W8GENA (ex1 = W2ExPEYP W6FriendNumYP) W8DLOCUS sex  eth1 W8EVERMAR W8DIN
    > CW W8WRKHRSA   wksearly W8DDEGP  W8SOCIALMED W1tvYP W1fameatYP W8AUDIT2 W8SLEEP2 W5agebd
    > 10mum educp, first
    
    First-stage regressions
    -----------------------
    
                                                    Number of obs     =      1,608
                                                    F(  17,   1590)   =       8.68
                                                    Prob > F          =     0.0000
                                                    R-squared         =     0.0849
                                                    Adj R-squared     =     0.0751
                                                    Root MSE          =     0.5821
    
    -------------------------------------------------------------------------------
              ex1 |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
    --------------+----------------------------------------------------------------
         W8DLOCUS |  -.0293182    .008982    -3.26   0.001     -.046936   -.0117005
              sex |  -.1802063   .0322194    -5.59   0.000    -.2434034   -.1170093
             eth1 |  -.1202028   .0483257    -2.49   0.013    -.2149916    -.025414
        W8EVERMAR |  -.0049658    .049327    -0.10   0.920    -.1017187    .0917871
          W8DINCW |  -.0008338   .0003018    -2.76   0.006    -.0014258   -.0002418
        W8WRKHRSA |    .044626   .0484779     0.92   0.357    -.0504613    .1397134
         wksearly |   -.005298   .0081444    -0.65   0.515    -.0212728    .0106768
          W8DDEGP |  -.0153314   .0317504    -0.48   0.629    -.0776085    .0469456
      W8SOCIALMED |   .0034061   .0059422     0.57   0.567    -.0082493    .0150615
           W1tvYP |  -.0965746   .0269797    -3.58   0.000    -.1494941   -.0436551
       W1fameatYP |   .0016477   .0157384     0.10   0.917    -.0292224    .0325178
         W8AUDIT2 |   .0001801   .0139301     0.01   0.990    -.0271432    .0275035
         W8SLEEP2 |  -.0092335   .0141343    -0.65   0.514    -.0369573    .0184904
     W5agebd10mum |  -.0038787   .0256409    -0.15   0.880    -.0541722    .0464148
            educp |   .0032174   .0346187     0.09   0.926    -.0646858    .0711205
         W2ExPEYP |  -.1093906    .031162    -3.51   0.000    -.1705135   -.0482676
    W6FriendNumYP |   .0606124   .0139416     4.35   0.000     .0332665    .0879582
            _cons |   3.259722    .248623    13.11   0.000     2.772059    3.747386
    -------------------------------------------------------------------------------
    
    
    Instrumental variables (2SLS) regression          Number of obs   =      1,608
                                                      Wald chi2(16)   =     171.32
                                                      Prob > chi2     =     0.0000
                                                      R-squared       =          .
                                                      Root MSE        =     .92091
    
    ------------------------------------------------------------------------------
          W8GENA |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
    -------------+----------------------------------------------------------------
             ex1 |  -.9303824   .2708754    -3.43   0.001    -1.461288   -.3994764
        W8DLOCUS |   .0868544   .0164686     5.27   0.000     .0545766    .1191323
             sex |  -.0104595   .0771549    -0.14   0.892    -.1616802    .1407613
            eth1 |  -.1772566   .0830206    -2.14   0.033     -.339974   -.0145393
       W8EVERMAR |   .0303651    .078027     0.39   0.697     -.122565    .1832953
         W8DINCW |  -.0015747   .0005072    -3.10   0.002    -.0025687   -.0005806
       W8WRKHRSA |   .1748691   .0769484     2.27   0.023     .0240531    .3256851
        wksearly |  -.0181418   .0130488    -1.39   0.164    -.0437169    .0074333
         W8DDEGP |  -.2017756   .0504776    -4.00   0.000    -.3007099   -.1028413
     W8SOCIALMED |   .0044162   .0094189     0.47   0.639    -.0140446     .022877
          W1tvYP |  -.0358402   .0510921    -0.70   0.483    -.1359788    .0642984
      W1fameatYP |  -.0200314   .0248962    -0.80   0.421     -.068827    .0287641
        W8AUDIT2 |   .0464691   .0220342     2.11   0.035     .0032829    .0896553
        W8SLEEP2 |    -.10722   .0224168    -4.78   0.000    -.1511561   -.0632838
    W5agebd10mum |  -.0709148   .0405905    -1.75   0.081    -.1504707    .0086411
           educp |   .0155376   .0546157     0.28   0.776    -.0915071    .1225824
           _cons |   4.933059   .9799335     5.03   0.000     3.012425    6.853694
    ------------------------------------------------------------------------------
    Instrumented:  ex1
    Instruments:   W8DLOCUS sex eth1 W8EVERMAR W8DINCW W8WRKHRSA wksearly
                   W8DDEGP W8SOCIALMED W1tvYP W1fameatYP W8AUDIT2 W8SLEEP2
                   W5agebd10mum educp W2ExPEYP W6FriendNumYP
    
    . 
    end of do-file
    
    .

  • #2
    The size of the coefficient will depend partly on how you measure your variables.

    Comment

    Working...
    X