I would really appreciate help!
I have eviews6 and I'm a beginner struggling with some basic problems..
My data is copy pasted from exel in the same form as in the 2nd example on this page. Data attached here: As You can see, the data has 218 observations and 10 independent variables.
Cross sections are 27 countries and for every country there are 8 years.
The data is about FDI flows to each country and I would like to use random effects, as my resources.
I should do a test about which are significant variables, when determining the FDI flows to each country, so I need coefficence, SE, t-test.
First of all is this a test where I need random effects?
I have tried to use "unstructured/undated" workfile structure, but it does not have random effects.
Also I have tried to input only one country to the panel data- structure but then eviews won't accept any cross-section or data-series ID (in the Proc-Structure/..-workfile structure).
So as far as I understand I should make a "balanced panel" workfile, which I have succeeded to do but when I run the regression I get the answer for a group of these countries.
I guess I should do like in the 1st sample here: http://forums.eviews.com/viewtopic.php? ... dom#p12822
Now my question is, how should I reformulate my data to be able to run the regression only for one country (or is there an easier way to do it)?
And is this easily done somewhere in eviews or should I rather do it on exel?
Can I do it trough the panel data -mode?
This is propably really simple, but I would really appreciate help
Also would it be possible to get a little hint about random effects in panel data (when running the regression for a group of countries):
What is the difference if the cross-section is random or the period is random?
Thanks a lot in advance!!