[Gretl-users] Heteroskedasticity / correction

Ana Amaro ISG anaamaro at isg.pt
Fri Jan 10 13:01:35 EST 2014



Sent from my iPhone
Apologize for the brevity/grammar/spelling 

No dia 10/01/2014, às 17:48, Sven Schreiber <svetosch at gmx.net> escreveu:

> Am 10.01.2014 18:38, schrieb Ana Amaro ISG:
>> Hi everyone
>> I need some help on this topic:
>> 1- simple model one regressor (x)
>> 2- error is heteroskedastic (and residual analysis shows that the error variance increases with x variable)
>> 3- rerun the analysis dividing both model members by the sqrt(x) - no constant
> 
> why no constant here? usually you need to have good reasons for that -- 
> if you had a constant in step 1, dividing everything by some number does 
> not eliminate the constant.
When dividing the original constant by sqrt(x) the constant disapears:(
> 
>> 
>> Eviews runs the white test with a constant
>> Gretl runs the white test with no constant!
>> The decision is, of course, different.
>> 
>> Which software is doing it the right way? Eviews, right?
> 
> I'm trying to avoid the question by suggesting to include the constant 
> in the estimated equation :-)
> 
> However, I tend to agree that the variance test equation should always 
> have a constant term, otherwise there is almost no chance to accept the 
> null, even without heteroscedasticity. So it looks as if gretl may need 
> to change here -- take it with a grain of salt though, as I haven't 
> thought much about it.
> 
> -sven
> _______________________________________________
> Gretl-users mailing list
> Gretl-users at lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users



More information about the Gretl-users mailing list