HELP: Random Effect problem: GLS weight??
Posted: Tue Oct 01, 2013 8:43 pm
HI,
My name is Intan, I am using Eviews 8.
I have tested using Hausman test, and my model calls for Random Effect.
I have also checked and found that my model has heteroschedasticity problem, so I need to apply GLS weight to correct it (is this right?)
1. But when i try to estimate my model using cross-section random effect, the GLS weight button greyed out, hence I cannot choose any GLS weight..
Why is it so? Does it mean that the cross-section random effect will automatically correct my heteroschedasticity without the need to put any GLS weight?
If it doesnt automatically correct, what can i do to correct the heteroschedasticity? Is it by putting some coef covariance method?
2. What does it mean by panel EGLS method? this appears on the estiamtion output when i run using cross sectn random effect
Looking forward to ur reply...
-Intan
My name is Intan, I am using Eviews 8.
I have tested using Hausman test, and my model calls for Random Effect.
I have also checked and found that my model has heteroschedasticity problem, so I need to apply GLS weight to correct it (is this right?)
1. But when i try to estimate my model using cross-section random effect, the GLS weight button greyed out, hence I cannot choose any GLS weight..
Why is it so? Does it mean that the cross-section random effect will automatically correct my heteroschedasticity without the need to put any GLS weight?
If it doesnt automatically correct, what can i do to correct the heteroschedasticity? Is it by putting some coef covariance method?
2. What does it mean by panel EGLS method? this appears on the estiamtion output when i run using cross sectn random effect
Looking forward to ur reply...
-Intan