Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: st: panel regression analysis interpretation
From
sun samn <[email protected]>
To
<[email protected]>
Subject
RE: st: panel regression analysis interpretation
Date
Thu, 22 Apr 2010 15:05:04 +0800
Hi, basically the assumed model there is:
yit=xit*beta+ui+eit
as we know, the fixed effect estimates are the same as Pooled OLS with ui as parameters.
For F test, since the P value is almost zero, then the null hypothesis is rejected, which means all uis cannot be zero, so the composite error terms(ui+eit) are correlated; so the iid condition is violated
For rho, it is defined as rho=sigma_u^2/(sigma_u^2+sigma_e^2)
corr(u_i, Xb) = -0.6067, means the correlation between ui and fitted values(xit*betahat) is -0.60
Please read page 273 of Wooldridge's book titled' econometrics analysis of cross section panel data'
I hope this will help a little bit.
----------------------------------------
> Date: Wed, 21 Apr 2010 20:05:00 -0400
> Subject: st: panel regression analysis interpretation
> From: [email protected]
> To: [email protected]
>
> Hi all,
>
> This is my first time on the listserve, so I apologize if my post
> isn't done correctly - please let me know. I am running an xtreg
> regression for a fixed-effects model on panel data. I would like to
> check whether all the OLS assumptions are ok and don't know which
> commands to employ to test for what potential problems I could have in
> my regression. I understand basic output results in terms of the box
> but don't know which other results correspond to which OLS
> assumptions. Thanks for your time. See results below:
>
>
> Fixed-effects (within) regression Number of obs = 11385
> Group variable: cntry_sector
> Number of groups = 1265
>
> R-sq: within = 0.0088 Obs per group: min = 9
> between = 0.0422 avg = 9.0
> overall = 0.0058 max = 9
>
> F(18,10102) = 4.99
> corr(u_i, Xb) = -0.6067 Prob> F = 0.0000
>
> ------------------------------------------------------------------------------
> invest | Coef. Std. Err. t P>|t| [95% Conf. Interval]
> -------------+----------------------------------------------------------------
> bilateral | -2115.566 1969.677 -1.07 0.283 -5976.524 1745.392
> braindrain | 1151.711 806.5183 1.43 0.153 -429.2253 2732.647
> regulation | 995.9853 532.6012 1.87 0.062 -48.0191 2039.99
> xrchange | 6.013753 13.17432 0.46 0.648 -19.81054 31.83805
> civil | 1234.246 590.0619 2.09 0.036 77.60741 2390.885
> environment | -610.7428 449.0376 -1.36 0.174 -1490.946 269.4602
> xr | .1900757 .6089638 0.31 0.755 -1.003614 1.383766
> corruption | 18.7143 25.00766 0.75 0.454 -30.30569 67.7343
> gdp | -7.588705 2.03874 -3.72 0.000 -11.58504 -3.59237
> gdpgrowth | 137.2847 76.59318 1.79 0.073 -12.85314 287.4226
> gdppercapita | -.2519762 .0719778 -3.50 0.000 -.3930671 -.1108854
> inflation | -10.41207 10.72849 -0.97 0.332 -31.44204 10.61789
> ipp | 309.6962 781.7363 0.40 0.692 -1222.662 1842.055
> internet | -6.65948 31.54257 -0.21 0.833 -68.48919 55.17023
> research | 603.5726 692.835 0.87 0.384 -754.5217 1961.667
> lendingrate | 70.42904 52.53728 1.34 0.180 -32.55447 173.4125
> phoneavg | 153.1085 31.87462 4.80 0.000 90.62792 215.5891
> university | -2173.322 663.4034 -3.28 0.001 -3473.725 -872.9195
> _cons | -2377.802 4799.696 -0.50 0.620 -11786.16 7030.557
> -------------+----------------------------------------------------------------
> sigma_u | 14989.05
> sigma_e | 21959.4
> rho | .31783249 (fraction of variance due to u_i)
> ------------------------------------------------------------------------------
> F test that all u_i=0: F(1264, 10102) = 2.52 Prob> F = 0.0000
>
> What does corr(u_i, Xb) = -0.6067 mean?
> Does F test that all u_i=0: F(1264, 10102) = 2.52 Prob
>> F = 0.0000 mean iid?
> and what does
> rho | .31783249 (fraction of variance due to u_i)
> mean (interpretation wise) ?
>
> *
> * 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/
_________________________________________________________________
Hotmail: Powerful Free email with security by Microsoft.
https://signup.live.com/signup.aspx?id=60969
*
* 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/