Correlogram of Residuals te:09506Tme:1428 Sample:19602001 Included obserations: 42 Autocorrelation Partial Correlation AC Pac @-Stat Prob 10.8680868 0000 20.6580387 0000 30.5050232 000 40.3710232727870000 50.2470071758350000 60.1410116768520000 700280126769040000 801340317778730000 903190164835570000 1004650.118960280000 (6)式残差序列的相关图、偏相关图) 在对数线性模型(6)的基础上再加入两个AR项。得估计结果 LnYr=-8.7350+1.7443 LnGDP+1.1840AR(1)-0.3511AR(2) (-13.6)(252) (-23) R2=0998,DW=1.64,T=40,(1962-2001) Dependent Variable: LOGY) Method: Least squares Date: 09/05/06 Time: 13: 46 Sample (adjusted): 1962 2001 Included observations: 40 after adjustments Convergence achieved after 8 iterations Variable Coefficient Std Error t-Statistic Prob C -8.7349500.54246513595000.0000 LOG(GDP) 1.7442810.06916025220850000 AR(1) 1.1839510150943784377200000 AR(2) -0.3511310.152700229948800274 R squar 0.998491 Mean dependent var 6.946438 Adjusted R-squared 0.998366 S.D. dependent var 2.529740 S.E. of regression 0.102272 Akaike info criterion-1.627715 Sum squared resid 0. 376547 Schwarz criterion 1.458827 Log likelihood 36.55431 F-statistic 941876 Durbin-Watson stat 1.641384 Prob(F-statistic Inverted ar root 59-03 59+03i Breusch-Godfrey Serial Correlation LM Test F-statistic 2091288Prob.F234) 0.139134 Obs"R-squared 4.381660 Prob Chi-Square (2) 0. 111824 (自相关检验结果)6 （（6）式残差序列的相关图、偏相关图） 在对数线性模型（6）的基础上再加入两个 AR 项。得估计结果， LnYt = -8.7350 +1.7443 LnGDPt + 1.1840AR(1) -0.3511AR(2) （11） (-13.6) (25.2) (7.8) (-2.3) R 2 = 0.998, DW=1.64, T = 40, (1962-2001) （自相关检验结果）