Forum: help
Monitor Forum | 
Start New Thread
| RE: Bivariate probit model with binary outcome [ Reply ] By: Arne Henningsen on 2018-11-17 12:32 | [forum:46426] |
| It is not clear to me what you have done and what did not work. Bivariate probit models can be estimated by the "mvProbit" package (and some other R packages). However, as far as I know, there is no function that uses a 2-step procedure for estimating bivariate probit models. I suggest that you post a minimal reproducible example. | |
| Bivariate probit model with binary outcome [ Reply ] By: rogelio andrade on 2018-11-01 10:19 | [forum:46396] |
|
Hello, I am fitting a bivariate probit model with binary outcome, using both the ML and 2step approach. The problem is that the correlation coefficient is close to the boundary, and hence I dont obtain p value on that coefficient. I have tried the approach suggested in the paper linked in the webpage for the library. Unfortunately, that doesnt solve the issue. Is it any other work around to obtain the std. errors and its p-value? |
|
