I am struggling with ATE. I have two treatment group in a cohort (binary 0=treatment A; 1=treatment B). I want to know the odds of variable Y (binary 0=no event; 1=event) b/w two treatment groups while controlling independent variables. I am using teffects ra command with logit function. However, the output is in Coef.. I assume these are exponentiated coefficients. What i understood is i have to convert these Coefficients to exponentiated to get Odds ratios. As i cant interpret nominal variable like contrinoues by using these. Am i missing something to get odds ratios ?

Thank you in advance

teffects ra (WeightbearingastoleratedWBA ib(0).Ageofpatientwithpatientsove ib(0).Gender ib(0).BMIBinned2 ib(0).Steroiduseforchronicconditio ib(0).Congestiveheart
> failureCHFi ib(0).PostopDiagnosisText2 ib(0).ASAclassification ib(0).Principalanesthesiatechnique2 ib(0).Totaloperationtime2, logit) (CPT)
note: 0.PostopDiagnosisText2 identifies no observations in the sample
note: 1.PostopDiagnosisText2 omitted because of collinearity

Iteration 0: EE criterion = 5.007e-07
Iteration 1: EE criterion = 5.430e-08
Iteration 2: EE criterion = 1.598e-09

Treatment-effects estimation Number of obs = 9,181
Estimator : regression adjustment
Outcome model : logit
Treatment model: none
| Robust
WeightbearingastoleratedWBA | Coef. Std. Err. z P>|z| [95% Conf. Interval]
(1 vs 0) | .0061516 .0126761 0.49 0.627 -.0186931 .0309964
POmean |
0 | .7271077 .0116412 62.46 0.000 .7042914 .749924
  • $\begingroup$ You are misunderstanding what Stata is doing. The ATE of 0.6 percentage points already takes the logit functional form into the account, so you don't need to do anything. See here for more details. You should also peruse the pdf manual for teffects ra. $\endgroup$ – Dimitriy V. Masterov May 26 '20 at 18:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.