Re: [R] question about 2SLS

Thanks. Estimation of the same model with the same dataset gives different results when "tsls" (from package sem) is used as opposed to "ivreg()" (from package AER); both parameter estimates and standard errors are different. This is intriguing. Can anyone throw some light on this? Is there any re

Re: [R] question about 2SLS

Please ignore my previous message about the difference between "tsls" and "ivreg" results; that was my mistake. Deepankar On Sun, May 2, 2010 at 5:10 PM, Dipankar Basu wrote: > Thanks. > > Estimation of the same model with the same dataset gives different results > when "tsls" (from package sem

Re: [R] question about 2SLS

On Sun, 2 May 2010, Dipankar Basu wrote: Hi All, I am using R 2.11.0 on a Ubuntu machine. I estimated a model using "tsls" from the package "sem". Is there a way to get Newey West standard errors for the parameter estimates? When estimating the model by OLS, I used "NeweyWest" from the package

[R] question about 2SLS

Hi All, I am using R 2.11.0 on a Ubuntu machine. I estimated a model using "tsls" from the package "sem". Is there a way to get Newey West standard errors for the parameter estimates? When estimating the model by OLS, I used "NeweyWest" from the package "sandwich" to get HAC standard errors. But,