Heteroskedasticity
Posted: Tue May 10, 2011 1:30 am
Hi all,
I know the theory of behind heteroskedasticity but not so good when it comes to Eviews...
For a project Im doing, I need to calculate the variance of the error term (from a regression). This can be done by dividing the residual output by N-2, only if the variance of the residuals in constant (ie homoskedastic). The graph of the residuals looks ok, but Id like to do a formal test to make sure there is no heteroskedasticity....
Any help or ideas would be appreciated.
L
I know the theory of behind heteroskedasticity but not so good when it comes to Eviews...
For a project Im doing, I need to calculate the variance of the error term (from a regression). This can be done by dividing the residual output by N-2, only if the variance of the residuals in constant (ie homoskedastic). The graph of the residuals looks ok, but Id like to do a formal test to make sure there is no heteroskedasticity....
Any help or ideas would be appreciated.
L