Additive structural mean model

Read in binary outcome data; y outcome, x exposure, w covariate, z* instrumental variables (genotypes).

use ivbinoutdata, clear

Fit the model with a single instrumental variable.

ivregress 2sls y (x = z1)
Instrumental variables 2SLS regression            Number of obs   =      2,500
                                                  Wald chi2(1)    =     100.20
                                                  Prob > chi2     =     0.0000
                                                  R-squared       =     0.4338
                                                  Root MSE        =     .37484

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
           x |   .1433391   .0143193    10.01   0.000     .1152738    .1714043
       _cons |   .2817003   .0271777    10.37   0.000      .228433    .3349676
------------------------------------------------------------------------------
Endogenous: x
Exogenous:  z1
ivreg2 y (x = z1)
IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only

                                                      Number of obs =     2500
                                                      F(  1,  2498) =   100.12
                                                      Prob > F      =   0.0000
Total (centered) SS     =     620.3344                Centered R2   =   0.4338
Total (uncentered) SS   =         1358                Uncentered R2 =   0.7413
Residual SS             =   351.260716                Root MSE      =    .3748

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
           x |   .1433391   .0143193    10.01   0.000     .1152738    .1714043
       _cons |   .2817003   .0271777    10.37   0.000      .228433    .3349676
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic):         171.159
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              183.591
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments):           0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         x
Excluded instruments: z1
------------------------------------------------------------------------------

Fit the model with multiple instruments.

ivregress 2sls y (x = z1 z2 z3)
Instrumental variables 2SLS regression            Number of obs   =      2,500
                                                  Wald chi2(1)    =     323.89
                                                  Prob > chi2     =     0.0000
                                                  R-squared       =     0.4223
                                                  Root MSE        =     .37863

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
           x |   .1307856   .0072671    18.00   0.000     .1165423    .1450289
       _cons |   .3046021    .015268    19.95   0.000     .2746774    .3345267
------------------------------------------------------------------------------
Endogenous: x
Exogenous:  z1 z2 z3
ivreg2 y (x = z1 z2 z3)
IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only

                                                      Number of obs =     2500
                                                      F(  1,  2498) =   323.63
                                                      Prob > F      =   0.0000
Total (centered) SS     =     620.3344                Centered R2   =   0.4223
Total (uncentered) SS   =         1358                Uncentered R2 =   0.7361
Residual SS             =  358.3930269                Root MSE      =    .3786

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
           x |   .1307856   .0072671    18.00   0.000     .1165423    .1450289
       _cons |   .3046021    .015268    19.95   0.000     .2746774    .3345267
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic):         678.024
                                                   Chi-sq(3) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              309.618
Stock-Yogo weak ID test critical values:  5% maximal IV relative bias    13.91
                                         10% maximal IV relative bias     9.08
                                         20% maximal IV relative bias     6.46
                                         30% maximal IV relative bias     5.39
                                         10% maximal IV size             22.30
                                         15% maximal IV size             12.83
                                         20% maximal IV size              9.54
                                         25% maximal IV size              7.80
Source: Stock-Yogo (2005).  Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments):           1.662
                                                   Chi-sq(2) P-val =    0.4357
------------------------------------------------------------------------------
Instrumented:         x
Excluded instruments: z1 z2 z3
------------------------------------------------------------------------------

Fit the model with multiple exposures, and instruments, and adjusting for w.

ivregress 2sls y w (x1 x2 = z1 z2 z3)
Instrumental variables 2SLS regression            Number of obs   =      2,500
                                                  Wald chi2(3)    =    1010.76
                                                  Prob > chi2     =     0.0000
                                                  R-squared       =     0.4481
                                                  Root MSE        =     .37006

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
          x1 |   .1212489   .0108116    11.21   0.000     .1000585    .1424392
          x2 |   .0142141   .0135397     1.05   0.294    -.0123231    .0407513
           w |   .0721392   .0129687     5.56   0.000     .0467211    .0975573
       _cons |   .3056947    .014944    20.46   0.000     .2764051    .3349843
------------------------------------------------------------------------------
Endogenous: x1 x2
Exogenous:  w z1 z2 z3
ivreg2 y w (x1 x2 = z1 z2 z3)
IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only

                                                      Number of obs =     2500
                                                      F(  3,  2496) =   336.38
                                                      Prob > F      =   0.0000
Total (centered) SS     =     620.3344                Centered R2   =   0.4481
Total (uncentered) SS   =         1358                Uncentered R2 =   0.7479
Residual SS             =  342.3610763                Root MSE      =    .3701

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
          x1 |   .1212489   .0108116    11.21   0.000     .1000585    .1424392
          x2 |   .0142141   .0135397     1.05   0.294    -.0123231    .0407513
           w |   .0721392   .0129687     5.56   0.000     .0467211    .0975573
       _cons |   .3056947    .014944    20.46   0.000     .2764051    .3349843
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic):         413.984
                                                   Chi-sq(2) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              165.050
Stock-Yogo weak ID test critical values: 10% maximal IV size             13.43
                                         15% maximal IV size              8.18
                                         20% maximal IV size              6.40
                                         25% maximal IV size              5.45
Source: Stock-Yogo (2005).  Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments):           0.523
                                                   Chi-sq(1) P-val =    0.4696
------------------------------------------------------------------------------
Instrumented:         x1 x2
Included instruments: w
Excluded instruments: z1 z2 z3
------------------------------------------------------------------------------