pool or panel setup
Posted: Mon Dec 09, 2019 6:44 pm
Using v9.5.
Sorry to bother but having trouble setting data structure, even after reading EViews illustrated. Don't know if pooled or panel structure is best, and would assume fixed effects.
I have 20 indexed observations with rank 1 = 100 and rank 20 declining toward zero. I have such a series of 20 observations for each of 39 years.
Trying to find variables to estimate the index rank so the numbers 1 to 20 are on left column, and dates on top row shown as ser01, ser02...ser39.
So there are 780 pieces of data in all.
What is the best way or the steps I can perform to set up file structure and make an estimate all at once from the entire data (if possible). I've managed to estimate a coefficient for each series in an equation such as obsrank c ser01, which correctly returns a negative coefficient.
Thanks.
Sorry to bother but having trouble setting data structure, even after reading EViews illustrated. Don't know if pooled or panel structure is best, and would assume fixed effects.
I have 20 indexed observations with rank 1 = 100 and rank 20 declining toward zero. I have such a series of 20 observations for each of 39 years.
Trying to find variables to estimate the index rank so the numbers 1 to 20 are on left column, and dates on top row shown as ser01, ser02...ser39.
So there are 780 pieces of data in all.
What is the best way or the steps I can perform to set up file structure and make an estimate all at once from the entire data (if possible). I've managed to estimate a coefficient for each series in an equation such as obsrank c ser01, which correctly returns a negative coefficient.
Thanks.