Here also the etregress is used with a log transformed dependent variable and a treatment component. Qui svy: etregress loglead i.female i.diabetes, treat(diabetes = weight age height i.female) // coefl I can provide reproducible example of the problem with the reference given by others: webuse nhanes2f, clear When I try the predict residual or predict rstudent nothing is reported making plotting not possible again. When I have this model simple assumptions related to a linear model like: Check linearity or assumption of independence and the homoscedasticity, normality, or goodness of fit diagnostics do not give output.Ī residuals versus predicted values plot could have been a rvfplot but this gives the error:Īnd the same for the estat hettest help etregress postestimationĭoes not discuss model assumption tests or goodness of fit tests which we normally see with regress or log-linear model in Stata. svy: etregress logwage i.race gender, treat(training = i.education gender) Complex strings may be very long and may contain binary information. which are your outcome and predictor variables).A regression makes sense only if there is a sound theory behind it. A complex string is a string that contains more than one piece of information. Before running a regression it is recommended to have a clear idea of what you are trying to estimate (i.e. When using survey data and etregress with an endogenous treatment effect in Stata number of diagnostics and post estimate parts stop being available for the use. A variable might contain strings because the data simply could not be coerced into being stored numerically. Can we do regression analysis with non normal data distribution I used a 710 sample size and got a z-score of some skewness between 3 and 7 and Kurtosis between 6 and 8.8.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |