Massimiliano wrote
"the Prob > F is usually smaller than 0.05 (actually being equal to
0.00 several times)"
No; this is a misconception. If Stata reports a P-value as 0 to the
number of decimal places given, that is what is meant. 0.000, for
example, means only <0.0005. No P-value from a test of this kind is
exactly zero.
On a point of English usage: I advise against expressions such as "a
panel data". It's true that many English dictionaries and style guides
condone "data" as a singular noun, but that usage, which is very
recent, still strikes me as unidiomatic English, awkward even
informally. In contrast, "a panel dataset", "a panel data set", "some
panel data" are all alternatives that appear much more widely
acceptable.
I know that languages change -- but that's a good reason to support
what is also StataCorp house style, namely that "dataset" is a good
word that should sound fine to English learners and users at all
levels.
Nick
njcoxstata@gmail.com
On 8 April 2013 09:45, Massimiliano Sassone
<massimilianosassone@gmail.com> wrote:
I am using Stata 11 to analyze a panel data composed of 279
observations, derived from 31 regions over a 9-year period.
In order to check for autocorrelation on several models, I ran the
Wooldridge test by inputting the -xtserial- command. Since the Prob >
F is usually smaller than 0.05 (actually being equal to 0.00 several
times), I understood that we fail to reject the null hypothesis and
that, therefore, the data is strongly autocorrelated.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/
