Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Panel Data-Autocorrelation


From   luis verdeja <luiseja@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Panel Data-Autocorrelation
Date   Sun, 17 Apr 2011 15:35:55 +0000

Dear Statalisters,

I am new into panel data so let me apologize in advance if this question is a bit too basic. I want to perform an analysis of trade flows (the entire universe of products) between 4 reporting countries and 14 partner countries, with fixed effects for both. Hence idr, with r=1...4 and idp, p=1...14. I have 5 other dependant variables and over 1,000,000 observations. I have checked collinearity issues and that's fine. However, there is evidence of autocorrelation in my data. Stata offers a series of commands to adjust for autocorrelation, including xtgls, c(ar1), xtpcse or prais. The only one that I can use is prais, because the others either give me the error that I have multiple panels (which is unavoidable) or ask me to set matsize 100,000, which is impossible. As far as I understand, I have three options:

1) Use prais (although I have to run over 100 regressions and it takes FOREVER to run one)
2) Average my data into two periods (with the problem of losing information)
3) Ignore autocorrelation (here is my key question, I have read that for small T, autocorrelation is not such a big deal)

Can you please point me to this, or any other solution?

Thank you very much,

Luis 		 	   		  
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index