第三讲 非线性模型

Dependent Variable: CONS Method: Least Squares Date: 06/22/11 Time: 17:01 Sample: 1978 1998 Included observations: 21

C GDP

R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic)

Coefficient 620.6414 0.573004

Std. Error 94.48131 0.002688

t-Statistic 6.568933 213.1681

Prob. 0.0000 0.0000 14984.05 14470.05 14.35909 14.45857 14.38068 0.805802

0.999582 Mean dependent var 0.999560 S.D. dependent var 303.5092 Akaike info criterion 1750239. Schwarz criterion -148.7705 Hannan-Quinn criter. 45440.65 Durbin-Watson stat 0.000000

cons=620.64+0.57*GDP

(64.481) (0.003) R^2过高，伪回归 TSLS

Dependent Variable: CONS Method: Two-Stage Least Squares Date: 06/22/11 Time: 17:06 Sample: 1978 1998 Included observations: 21 Instrument list: C CAPI

C GDP

R-squared Adjusted R-squared S.E. of regression F-statistic Prob(F-statistic)

Coefficient 630.2961 0.572619

Std. Error 94.61792 0.002694

t-Statistic 6.661488 212.5318

Prob. 0.0000 0.0000 14984.05 14470.05 1752130. 0.804143 22208974

0.999582 Mean dependent var 0.999560 S.D. dependent var 303.6732 Sum squared resid 45169.75 Durbin-Watson stat 0.000000 Second-Stage SSR

Cons=630.30+0.57*GDP GDP=645.63+2.53*CAPI