Fitted values of a panel regression with 95% CI

[Carl Franklin]

Hi,

I am a new member, so apologies in advance if I am inadvertently not adhering to certain instructions for posting questions on the forum.

I am trying to do a very simple exercise, but for some reason I cannot find an answer immediately related to this question anywhere on the forum. Suppose I have an N by T panel with i=1,...,N, t=1,...,T. All I am attempting to do is after using, say, the fixed-effects estimator obtain the fitted values, as well as the 95CI for each individual i in N. How can this be accomplished? Perhaps there is an easier way to accomplish this with LSDV?

Thank you in advance.

[Clyde Schechter]

The -predict- command will do this for you, whether after one of the -xt- commands, or after -regress- with indicator ("dummy") variables. For that matter, it can be used after any Stata estimation command.

For details, run -help xtreg postestimation- and then click on the blue -predict- link to see the options that will give you the predicted values and their standard errors. You can then calculate the confidence intervals from there by the standard statistical formulas.