R-squared 0.978378 Mean dependent var 8434.222 Adiusted R-squared 0.974965 S.D.dependent var 9025.326 S.E.of regression 1428.041 Akaike info criterion 17.52277 Sum squared resid 38746720 Schwarz criterion 17.72024 Log likelihood -197.5118 F-statistic 286.5846 Durbin-Watson stat 2.047965 Prob(F-statistic) 0.000000 可以认为，这时模型设定无变量设定误差 2、LM检验 按照M检验步骤，首先生成残差序列（用EE表示），用EE对全部解释变量（包括 遗漏变量)进行回归，有： Dependent Variable:EE Method:Least Squares Date:07108/05Time:15:45 Sample(adjusted):19812003 Included observations:23 after adjustments Variable Coefficient Std.Error t-Statistic Prob. c 448.1584 511.0396 0.8769540.3915 GDP 0.912201 0.148119 6.1585680.0000 GDP(-1) -0.8158420.1439285.6684200.0000 EXCHANGE2 -0.022569 0.0032916.857844 0.000052 R-squared 0.978378 Mean dependent var 8434.222 Adjusted R-squared 0.974965 S.D. dependent var 9025.326 S.E. of regression 1428.041 Akaike info criterion 17.52277 Sum squared resid 38746720 Schwarz criterion 17.72024 Log likelihood -197.5118 F-statistic 286.5846 Durbin-Watson stat 2.047965 Prob(F-statistic) 0.000000 可以认为，这时模型设定无变量设定误差。 2、LM 检验 按照 LM 检验步骤，首先生成残差序列 i e （用 EE 表示），用 EE 对全部解释变量（包括 遗漏变量）进行回归，有： Dependent Variable: EE Method: Least Squares Date: 07/08/05 Time: 15:45 Sample (adjusted): 1981 2003 Included observations: 23 after adjustments Variable Coefficient Std. Error t-Statistic Prob. C 448.1584 511.0396 0.876954 0.3915 GDP 0.912201 0.148119 6.158568 0.0000 GDP(-1) -0.815842 0.143928 -5.668420 0.0000 EXCHANGE^2 -0.022569 0.003291 -6.857844 0.0000