Skip to Main Content

Methods Matter: Improving causal Inference in Educational and Social Science Research by Richard J. Murnane and John B. Willett Chapter 11: Using IVE to Recover the Treatment Effect in a Quasi-Experiment | Stata Textbook Examples

use https://stats.idre.ucla.edu/stat/stata/examples/methods_matter/chapter11/colombia_voucher, clear 

Descriptive statistics for Table 11.1 on page 270.

sum finish8th use_fin_aid won_lottry base_age male 

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
   finish8th |      1171    .6814688    .4661058          0          1
 use_fin_aid |      1171    .5815542    .4935148          0          1
  won_lottry |      1171    .5055508    .5001828          0          1
    base_age |      1171    12.00427    1.347038          7         17
        male |      1171    .5046968    .5001916          0          1


sort won_lottry 
by won_lottry: sum finish8th use_fin_aid won_lottry base_age male 

-----------------------------------------------------------------------------------------
-> won_lottry = 0

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
   finish8th |       579    .6252159    .4844857          0          1
 use_fin_aid |       579    .2400691    .4274945          0          1
  won_lottry |       579           0           0          0          0
    base_age |       579    12.03627    1.351814          7         16
        male |       579    .5043178    .5004137          0          1

-----------------------------------------------------------------------------------------
-> won_lottry = 1

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
   finish8th |       592    .7364865     .440911          0          1
 use_fin_aid |       592    .9155405    .2783108          0          1
  won_lottry |       592           1           0          1          1
    base_age |       592    11.97297    1.342755          9         17
        male |       592    .5050676    .5003971          0          1
        	

Significance tests for Table 11.1 on page 270.

reg finish8th won_lottry

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  1,  1169) =   16.91
       Model |  3.62413371     1  3.62413371           Prob > F      =  0.0000
    Residual |   250.56374  1169  .214340239           R-squared     =  0.0143
-------------+------------------------------           Adj R-squared =  0.0134
       Total |  254.187874  1170  .217254593           Root MSE      =  .46297

------------------------------------------------------------------------------
   finish8th |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  won_lottry |   .1112706   .0270601     4.11   0.000     .0581787    .1643625
       _cons |   .6252159   .0192403    32.50   0.000     .5874664    .6629653
------------------------------------------------------------------------------

reg use_fin_aid won_lottry

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  1,  1169) = 1031.16
       Model |  133.554147     1  133.554147           Prob > F      =  0.0000
    Residual |  151.407424  1169  .129518755           R-squared     =  0.4687
-------------+------------------------------           Adj R-squared =  0.4682
       Total |  284.961571  1170  .243556899           Root MSE      =  .35989

------------------------------------------------------------------------------
 use_fin_aid |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  won_lottry |   .6754715   .0210351    32.11   0.000     .6342007    .7167422
       _cons |   .2400691   .0149564    16.05   0.000     .2107247    .2694135
------------------------------------------------------------------------------

reg base_age won_lottry

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  1,  1169) =    0.65
       Model |  1.17274119     1  1.17274119           Prob > F      =  0.4217
    Residual |  2121.80591  1169  1.81506066           R-squared     =  0.0006
-------------+------------------------------           Adj R-squared = -0.0003
       Total |  2122.97865  1170  1.81451167           Root MSE      =  1.3472

------------------------------------------------------------------------------
    base_age |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  won_lottry |  -.0632965   .0787452    -0.80   0.422    -.2177941    .0912012
       _cons |   12.03627   .0559895   214.97   0.000     11.92642    12.14612
------------------------------------------------------------------------------

reg male won_lottry

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  1,  1169) =    0.00
       Model |  .000164554     1  .000164554           Prob > F      =  0.9796
    Residual |  292.724003  1169  .250405477           R-squared     =  0.0000
-------------+------------------------------           Adj R-squared = -0.0009
       Total |  292.724167  1170  .250191596           Root MSE      =  .50041

------------------------------------------------------------------------------
        male |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  won_lottry |   .0007498   .0292483     0.03   0.980    -.0566352    .0581348
       _cons |   .5043178   .0207961    24.25   0.000     .4635159    .5451197
------------------------------------------------------------------------------
	

Naive OLS estimate of the model shown in Table 11.2 on page 273.

reg finish8th use_fin_aid base_age male

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  3,  1167) =   26.53
       Model |   16.230726     3  5.41024199           Prob > F      =  0.0000
    Residual |  237.957148  1167  .203905011           R-squared     =  0.0639
-------------+------------------------------           Adj R-squared =  0.0614
       Total |  254.187874  1170  .217254593           Root MSE      =  .45156

------------------------------------------------------------------------------
   finish8th |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
 use_fin_aid |   .1209084   .0268012     4.51   0.000     .0683244    .1734924
    base_age |   -.062961   .0098456    -6.39   0.000    -.0822779    -.043644
        male |  -.0858505   .0264754    -3.24   0.001    -.1377952   -.0339057
       _cons |   1.410283   .1206031    11.69   0.000      1.17366    1.646906
------------------------------------------------------------------------------
	

Instrumental-variables (2SLS) from Table 11.2 on page 273.

ivreg finish8th base_age male (use_fin_aid=won_lottry), first

First-stage regressions
-----------------------

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  3,  1167) =  346.26
       Model |  134.199092     3  44.7330305           Prob > F      =  0.0000
    Residual |   150.76248  1167  .129188072           R-squared     =  0.4709
-------------+------------------------------           Adj R-squared =  0.4696
       Total |  284.961571  1170  .243556899           Root MSE      =  .35943

------------------------------------------------------------------------------
 use_fin_aid |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
    base_age |  -.0151604    .007826    -1.94   0.053    -.0305151    .0001943
        male |  -.0202571     .02107    -0.96   0.337    -.0615965    .0210823
  won_lottry |    .674527   .0210141    32.10   0.000     .6332974    .7157567
       _cons |   .4327598   .0951594     4.55   0.000     .2460573    .6194624
------------------------------------------------------------------------------


Instrumental variables (2SLS) regression

      Source |       SS       df       MS              Number of obs =    1171
-------------+------------------------------           F(  3,  1167) =   25.21
       Model |  15.8188369     3  5.27294563           Prob > F      =  0.0000
    Residual |  238.369037  1167  .204257958           R-squared     =  0.0622
-------------+------------------------------           Adj R-squared =  0.0598
       Total |  254.187874  1170  .217254593           Root MSE      =  .45195

------------------------------------------------------------------------------
   finish8th |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
 use_fin_aid |   .1590002   .0391733     4.06   0.000     .0821423    .2358581
    base_age |  -.0621574   .0098725    -6.30   0.000    -.0815271   -.0427876
        male |  -.0851448   .0265036    -3.21   0.001    -.1371449   -.0331448
       _cons |   1.378128   .1230896    11.20   0.000     1.136626    1.619629
------------------------------------------------------------------------------
Instrumented:  use_fin_aid
Instruments:   base_age male won_lottry
------------------------------------------------------------------------------