Panel Data OLS Estimation

[marcusj91]

Hi there

Just wondering if anyone has any idea how to run a ordinary least squares regression on eviews for panel data, dependent variable then 6 explanatory variables. Apologies if this is the simplest task ever, I am a complete novice when it comes to eviews. Thanks for any help!!!. Is it just a simple case of quick> estimate equation > then typing the variables in like you would on a cross section

[marcusj91]

Forgot, any idea to makes your standard errors robust in terms of heterskedasticity and serial correlation? Thanks

[EViews Gareth]

It works essentially the same was as in a cross-section (or a time-series) file. Quick->Estimate Equation, put your variables in, and hit ok. If you want robust standard errors, look at the other tabs in the estimate dialog. More details are in the manual in the Panel Estimation chapter.

[marcusj91]

Thanks for the advice!!. I am now trying to assess the level of accuracy of the model and efficiency of it by testing for endogeneity, collinearity, serial correlation etc. I have used the DW stat for serial correlation but does anyone have an idea how to test for collinearity and endogeneity, I was thinking of trying to produce a correlation matrix but when I do all it comes up with is N/A for all the values. Anyone know why this is or of alternative ways to test and interpret results?

Thanks