st: re: tsset with non-integers

[Kit Baum <baum@bc.edu>]

<>
Frank said

Thank you for your responses. My goal is only to check the  
independence of residual terms following a regression run. My reading  
of the durbinh command lead me to believe that it would help me  
achieve my goal. Can you suggest an alternative option to check the  
independence of residual terms that are non-integers? Thank you.


Residuals are never integers (except by chance). To analyze serial  
correlation, the data must be on a time-series calendar with even  
increments of time between observations. Those dates are integers.

As I said previously, durbinh considers the null that there is no  
serial correlation against the alternative that there is AR(1). You  
can fail to reject in a situation where the errors are most decidedly  
not i.i.d. but rather AR(p), p > 1. That is quite common in real data.  
A test such as Breusch-Godfrey (of which Durbin's h is just  a special  
case) allows for p > 1, as does the 'Q' test applied to residuals.

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                              An Introduction to Stata Programming   
|   http://www.stata-press.com/books/isp.html
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html



*
*   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/