DATA Step, Macro, Functions and more

Correct for Heteroskedasticity with PROC REG

Reply
Contributor
Posts: 23

Correct for Heteroskedasticity with PROC REG

Hi all SAS users,

 

I'm running a linear regression model by using PROC REG (v. SAS 9.4).

 

The model did not overcome the test about heteroschedasticity and I want to correct for the presence of heteroschedastic errors.

 

Browsing on the internet I didn't find anything useful to do that.

 

Can you provide a way, like, for instance, a statement, to do it?

 

Thanks for the help!

Regular Contributor
Posts: 249

Re: Correct for Heteroskedasticity with PROC REG

Posted in reply to Quantopic

Maybe try to log-transform your outcome variable if that helps?

Contributor
Posts: 23

Re: Correct for Heteroskedasticity with PROC REG

[ Edited ]

Thanks for replying @wong, but I cannot do that!

 

I need to correct it by applying the heteroscedasticity-consistent parameter estimates.

 

I can doit by using another PROC too.

Ask a Question
Discussion stats
  • 2 replies
  • 91 views
  • 0 likes
  • 2 in conversation