Re: Covariance Analysis Dialogue Box

[emelia67]

What is the difference between ordinary and ordinary (uncentered) method in the covariance analysis dialog box. How does one determine which method to use as they give completely different results. The Eviews dialogue box does not explain this on the relevant page of the user guide for eviews 8.1

[EViews Gareth]

Page 504 of User Guide I.

[emelia67]

the page talks about the ordinary uncentered setting the mean estimates to 0. So which of the two - ie. the ordinary and the ordinary uncentered is more reliable if I am estimating the convergence in macroeconomic variables, GDP growth rate, amongst the EURO zone countries.

[emelia67]

I am estimating a model for which I have a panel data with 572 cross-sections and 12 time periods (annual). I want to use robust standard errors to allow possible heteroscedasticity and serial correlations. I have read the relevant page of the eviews manual but I am not too sure which of White's covariance method takes care of both - is it White's period or White's cross-section, judging from the nature of my data set? I will appreciate your help.

[EViews Glenn]

White period. Clustering by cross-section.