[Gretl-users] Re: OLS Results mis-match
Please give us the details of the gretl version you're using. I'm not
seeing the problem you describe here.
Model 1: OLS, using observations 1-90
Dependent variable: C
coefficient std. error t-ratio p-value
------------------------------------------------------------
const 1.15856e+06 360593 3.213 0.0018 ***
Q 2.02611e+06 61806.9 32.78 2.25e-50 ***
PF 1.22535 0.103722 11.81 1.07e-19 ***
LF −3.06575e+06 696327 −4.403 3.06e-05 ***
Mean dependent var 1122524 S.D. dependent var 1192075
Sum squared resid 6.82e+12 S.E. of regression 281559.5
R-squared 0.946093 Adjusted R-squared 0.944213
F(3, 86) 503.1176 P-value(F) 2.10e-54
Log-likelihood −1254.988 Akaike criterion 2517.975
Schwarz criterion 2527.974 Hannan-Quinn 2522.007
Allin Cottrell
On Mon, Apr 6, 2026 at 5:25 AM Ramki S <ramakrishnasalagrama(a)gmail.com> wrote:
>
> I agree that se's are large but can we have an option to agree with results of
other softwares? stata, r, spss are all producing similar results.