problem with estat hettest

[Philip Gigliotti]

Hello,

I'm trying to test my model for heteroskedasticity, but I'm having problems with the command.

I'm entering the line:

estat hettest

and getting the following error:

subcommand estat hettest is unrecognized

Any idea what's going on?

[Clyde Schechter]

-estat hettest- is, as far as I know, only usable after -regress-. If you have run some other kind of model, you will get that error message.

[Philip Gigliotti]

my model is a fixed effects model and I have calculated it with and without heteroskedasticity robust standard errors. I want to test the model without for heteroskedasticity, to see whether the robust errors are actually necessary. Is there a command to test fixed effects models for heteroskedasticity?

[Carlo Lazzaro]

Philip may find the following thread useful: http://www.statalist.org/forums/foru...dasticity-test