Dear Statalist members,
I'm using the user written command "sensatt" (https://www.stata-journal.com/sjpdf....iclenum=st0130) to perform sensitivity analysis after propensity matching.
My outcome variable is an efficient score and the treatment regards the transition between organizational settings, where the explanatory variables relates with external factors such as GDP per capita, if organizations are in urban areas, unemployment rate, etc.
Code is reflected on below.
The final message is that "Both the outcome and the selection effect are odds ratios from logit estimations"
My questions are: 1) What workaround should be followed to understand and solve this problem?
2) It seems that this command does not allow dummy variables (in my case B variable), but I'm not sure if it should be put in other way.
Thank you in advance.
I'm using the user written command "sensatt" (https://www.stata-journal.com/sjpdf....iclenum=st0130) to perform sensitivity analysis after propensity matching.
My outcome variable is an efficient score and the treatment regards the transition between organizational settings, where the explanatory variables relates with external factors such as GDP per capita, if organizations are in urban areas, unemployment rate, etc.
Code is reflected on below.
The final message is that "Both the outcome and the selection effect are odds ratios from logit estimations"
My questions are: 1) What workaround should be followed to understand and solve this problem?
2) It seems that this command does not allow dummy variables (in my case B variable), but I'm not sure if it should be put in other way.
Thank you in advance.
Code:
teffects psmatch (outcome) ( treatment A i.B C D, probit ) if type=="SME", vce(robust) atet
Treatment-effects estimation Number of obs = 930
Estimator : propensity-score matching Matches: requested = 1
Outcome model : matching min = 1
Treatment model: probit max = 8
--------------------------------------------------------------------------------
| AI Robust
VRS_TE_1_1 | Coef. Std. Err. z P>|z| [95% Conf. Interval]
---------------+----------------------------------------------------------------
ATET |
d_transit_ever |
(1 vs 0) | .0173109 .0080928 2.14 0.032 .0014493 .0331725
sensatt outcome treatment (A C D) if type=="SME", r(100)
*** THIS IS THE BASELINE ATT ESTIMATION (WITH NO SIMULATED CONFOUNDER).
The program is searching the nearest neighbor of each treated unit.
This operation may take a while.
ATT estimation with Nearest Neighbor Matching method
(random draw version)
Analytical standard errors
---------------------------------------------------------
n. treat. n. contr. ATT Std. Err. t
---------------------------------------------------------
181 218 0.020 0.009 2.302
---------------------------------------------------------
Note: the numbers of treated and controls refer to actual
nearest neighbour matches
*** THIS IS THE SIMULATED ATT ESTIMATION (WITH THE CONFOUNDER U).
The probability of having U=1 if T=1 and Y=1 (p11) is equal to: 0.00
The probability of having U=1 if T=1 and Y=0 (p10) is equal to: 0.00
The probability of having U=1 if T=0 and Y=1 (p01) is equal to: 0.00
The probability of having U=1 if T=0 and Y=0 (p00) is equal to: 0.00
The probability of having U=1 if T=1 (p1.) is equal to: 0.00
The probability of having U=1 if T=0 (p0.) is equal to: 0.00
The program is iterating the ATT estimation with simulated confounder.
You have chosen to perform 100 iterations. This step may take a while.
ATT estimation with simulated confounder
General multiple-imputation standard errors
-----------------------------------------------
ATT Std. Err. Out. Eff. Sel. Eff.
-----------------------------------------------
0.020 0.009 . .
-----------------------------------------------
Note: Both the outcome and the selection effect
are odds ratios from logit estimations.
