I think one interesting issue in this particular case is that age is negatively related to hours in the selection model but positively related to hours in the linear model. That is, older people are less likely to exercise at all, but among people who do exercise, older people exercise more hours. This is something unique that you can only get from hurdle models.
I think one interesting issue in this particular case is that age is negatively related to hours in the selection model but positively related to hours in the linear model. That is, older people are less likely to exercise at all, but among people who do exercise, older people exercise more hours. This is something unique that you can only get from hurdle models.
Hi! Do you know what the difference is between using the churdle and craggit estimator? Thank you, Elina
i am getting intial values not feasible error while running this codes. what may be the reason?
how to interpret the marginal effect on a log transformed variable? percent-wise as usually? what about the probability of overcoming the hurdle?
did you get the answer?
@@annewambui68 no
@@RPDBY argh okay
I tried to do hurdle regression but it was end up with convergence not achieved. What is the matter.?
how can you use this for willingness to pay, i.e what variables for select?
please tell me about double hurdle regression command
Thanks a lot!
Hi! Is it impossible to put the same variables (all the independent variables) in both stages?
Yes
What is the alternative to this command in Stata 12?
it is available only starting from version 15
truncreg
thanks a lot for this. I have a small doubt. If the dependant variable is ordinal and not continuous, can we still use the churdle ?
You can, but we won't use the linear version as it's done here.