Dear all,
I am estimating a fixed effects model and am wondering if inputting the common AR(1) term into my list of common coefficients will correct for first order autocorrelation within cross-sections or between cross-sections given that it is highly significant. I'm not sure how it works in Eviews.
Thanks
Rhea
Common AR(1) term to list of coefficients
Moderators: EViews Gareth, EViews Jason, EViews Steve, EViews Moderator
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Common AR(1) term to list of coefficients
If you have a panel workfile, AR(1) will work within cross-sections.
Re: Common AR(1) term to list of coefficients
Thank you very much! So am I correct in saying that the interpretation is then that of a first order autoregressive fixed effects panel estimation?
Also could I ask - is this valid considering I have already used cluster-robust standard errors - clustered by cross-section? I have come to understand that using this estimation ensures robustness for within cluster heteroskedasticity and serial correlation so I'm not sure what to make of this common AR(1) term.
Also could I ask - is this valid considering I have already used cluster-robust standard errors - clustered by cross-section? I have come to understand that using this estimation ensures robustness for within cluster heteroskedasticity and serial correlation so I'm not sure what to make of this common AR(1) term.
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Common AR(1) term to list of coefficients
Using AR(1) corrects the coefficient estimates and standard errors for serial correlation. Clustering by cross-section fixes the standard errors for serial correlation, but does not affect the coefficient estimates.
Re: Common AR(1) term to list of coefficients
Thank you very much! Really appreciate the help!
Re: Common AR(1) term to list of coefficients
Hello e-views forum,
hope everybody is doing well. I would really appreciate some help.
While estimating a fixed effects model, if I choose as a coefficient covariance method--> White cross section, then it means that my standard errors are robust for heteroskedasticity and first order serial correlation?
Thank you very much in advance.
hope everybody is doing well. I would really appreciate some help.
While estimating a fixed effects model, if I choose as a coefficient covariance method--> White cross section, then it means that my standard errors are robust for heteroskedasticity and first order serial correlation?
Thank you very much in advance.
Re: Common AR(1) term to list of coefficients
Hello again to everybody,
I am back with more information cause it is complicated for me and I need some help,
I have panel (balanced data) and I estimated a fixed effects model. I have 14 countries and 11 years.
When I estimate, the DW statistic is 1, hence suggesting that I have positive autocorrelation.
so,
1) If I suppose that the autocorrelation is within the country, e-views help suggests : "we may adopt the Arellano (1987) approach of computing White coefficient covariance estimates that are robust to arbitrary within cross-section residual correlation (clustering by cross-section). Select the Options page and choose White period as the coefficient covariance method."
However when I use it, I get the warning regarding the number of cross sections. So it is not good to use it right?
2) this may sound very starters question, but when talking about an AR(1), you mean to include in the regression a lagged dependent variable.? Is that correct?
In that case the DW stat increases to 1.30
3) If I would like also to control for heteroskedasticity, which method to use?
Thank you very much,
Hercules
I am back with more information cause it is complicated for me and I need some help,
I have panel (balanced data) and I estimated a fixed effects model. I have 14 countries and 11 years.
When I estimate, the DW statistic is 1, hence suggesting that I have positive autocorrelation.
so,
1) If I suppose that the autocorrelation is within the country, e-views help suggests : "we may adopt the Arellano (1987) approach of computing White coefficient covariance estimates that are robust to arbitrary within cross-section residual correlation (clustering by cross-section). Select the Options page and choose White period as the coefficient covariance method."
However when I use it, I get the warning regarding the number of cross sections. So it is not good to use it right?
2) this may sound very starters question, but when talking about an AR(1), you mean to include in the regression a lagged dependent variable.? Is that correct?
In that case the DW stat increases to 1.30
3) If I would like also to control for heteroskedasticity, which method to use?
Thank you very much,
Hercules
Who is online
Users browsing this forum: No registered users and 16 guests