Forum: help
Monitor Forum | Start New ThreadRE: Clustering errors for a multivariate probit [ Reply ] By: Arne Henningsen on 2015-01-25 20:24 | [forum:41838] |
Hi Jacquelyn Thank you for the clarification. I am not sure, how one could implement clustered standard errors in a sample selection model with a binary dependent variable of the outcome equation. Sorry! However, if a user (e.g. you) wants to implement this feature, I would be happy to support him/her. Best wishes, Arne |
RE: Clustering errors for a multivariate probit [ Reply ] By: Jacquelyn Pless on 2015-01-25 19:22 | [forum:41836] |
Hi Arne, Thanks again for your follow-up. I misspoke - it is vcovHC that allows you to cluster (on group, for instance): http://www.inside-r.org/packages/cran/plm/docs/vcovHC. This only works, however, for objects of the class plm or pgmm, and I am trying to cluster my errors for a selection model. Any thoughts or insights would be greatly appreciated. Thank you! Jacquelyn |
RE: Clustering errors for a multivariate probit [ Reply ] By: Arne Henningsen on 2015-01-25 13:07 | [forum:41831] |
Dear Jacquelyn As far as I know, it is currently not possible to cluster errors based on specific variables. However, if a user (e.g. you) wants to implement this feature, I would be happy to support him/her. Which function/method do you mean by "the standard coef() function for doing so"? Can you please give me a hint, example, and/or URL to its documentation? Best regards, Arne |
Clustering errors for a multivariate probit [ Reply ] By: Jacquelyn Pless on 2015-01-21 22:54 | [forum:41821] |
Hi all, I am currently estimating a bivariate probit model with selection. I am doing this with the selection() function from the sampleSelection package and specifying the method as binaryOutcome. Is it possible to cluster my errors based on a specific variable? For instance, I'd like to cluster errors by postal zip code, and the standard coef() function for doing so doesn't work for the selection model object. Thanks in advance for your help! Jacquelyn |