I am trying to find out the treatment effect using propensity score matching. I have a query on the choice of covariates required for determining the propensity scores. As mentioned in the previous literature, I understood that only the background variables are to be included which is not affected by treatment. In the pscore command in stata, I have incorporated such controls but none of them are significant. In this case, will this approach work?
Also, I have another query regarding the balancing property. According to the pscore command, the results show that the final number of blocks is 5. However, the detailed output only shows two blocks - 5 and 4
It would be a great help if I could get some clarifications on whether my procedure is correct.
Also, I have another query regarding the balancing property. According to the pscore command, the results show that the final number of blocks is 5. However, the detailed output only shows two blocks - 5 and 4
Code:
****************************************************
Algorithm to estimate the propensity score
****************************************************
The treatment is dumtui
(firstnm) |
dumtui | Freq. Percent Cum.
------------+-----------------------------------
0 | 40 7.29 7.29
1 | 509 92.71 100.00
------------+-----------------------------------
Total | 549 100.00
Estimation of the propensity score
Iteration 0: log likelihood = -143.27483
Iteration 1: log likelihood = -140.05137
Iteration 2: log likelihood = -140.01568
Iteration 3: log likelihood = -140.01567
Probit regression Number of obs = 549
LR chi2(7) = 6.52
Prob > chi2 = 0.4807
Log likelihood = -140.01567 Pseudo R2 = 0.0227
------------------------------------------------------------------------------
dumtui | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
logfeediff | -.4351585 .3457078 -1.26 0.208 -1.112733 .2424163
meaninc | .0000241 .000032 0.75 0.452 -.0000386 .0000868
engdiff | .7431845 .4173667 1.78 0.075 -.0748392 1.561208
profdiff | .020041 .5687919 0.04 0.972 -1.094771 1.134853
govprop | -.0285287 .5192992 -0.05 0.956 -1.046337 .9892791
propunemp | -.2523839 .6641851 -0.38 0.704 -1.554163 1.049395
propmaleat~d | .0205683 .5769323 0.04 0.972 -1.110198 1.151335
_cons | 4.965179 2.654825 1.87 0.061 -.238183 10.16854
------------------------------------------------------------------------------
Note: the common support option has been selected
The region of common support is [.79588694, .99533472]
Description of the estimated propensity score
in region of common support
Estimated propensity score
-------------------------------------------------------------
Percentiles Smallest
1% .8346569 .7958869
5% .8778466 .8202823
10% .8917124 .8244704 Obs 549
25% .910725 .830139 Sum of Wgt. 549
50% .9283149 Mean .9271435
Largest Std. Dev. .0288954
75% .9503249 .9760316
90% .9618948 .9777905 Variance .0008349
95% .9658854 .9830645 Skewness -.8160764
99% .9709594 .9953347 Kurtosis 4.133895
******************************************************
Step 1: Identification of the optimal number of blocks
Use option detail if you want more detailed output
******************************************************
The final number of blocks is 5
This number of blocks ensures that the mean propensity score
is not different for treated and controls in each blocks
**********************************************************
Step 2: Test of balancing property of the propensity score
Use option detail if you want more detailed output
**********************************************************
The balancing property is satisfied
This table shows the inferior bound, the number of treated
and the number of controls for each block
Inferior |
of block | (firstnm) dumtui
of pscore | 0 1 | Total
-----------+----------------------+----------
.6 | 0 1 | 1
.8 | 40 508 | 548
-----------+----------------------+----------
Total | 40 509 | 549
Note: the common support option has been selected
*******************************************
End of the algorithm to estimate the pscore
*******************************************
It would be a great help if I could get some clarifications on whether my procedure is correct.
