七、模型缺陷分析:
1、平稳性检验:
ADF Test Statistic 5.715142 1% Critical Value* -3.7497
5% Critical Value -2.9969
10% Critical Value -2.6381
*MacKinnon critical values for rejection of hypothesis of a unit root.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(X2)
Method: Least Squares
Date: 06/03/05 Time: 07:56
Sample(adjusted): 1981 2003
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
X2(-1) 0.163568 0.028620 5.715142 0.0000
C 2.705680 4.359016 0.620709 0.5415
R-squared 0.608668 Mean dependent var 19.37000
Adjusted R-squared 0.590033 S.D. dependent var 24.26967
S.E. of regression 15.53955 Akaike info criterion 8.407595
Sum squared resid 5071.027 Schwarz criterion 8.506333
Log likelihood -94.68734 F-statistic 32.66285
Durbin-Watson stat 1.535714 Prob(F-statistic) 0.000011
(表13)
X2通过平稳性检验。
ADF Test Statistic 0.783554 1% Critical Value* -3.7497
5% Critical Value -2.9969
10% Critical Value -2.6381
*MacKinnon critical values for rejection of hypothesis of a unit root.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(X3)
Method: Least Squares
Date: 06/03/05 Time: 08:04
Sample(adjusted): 1981 2003
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
X3(-1) 0.093081 0.118794 0.783554 0.4421
C 0.870734 1.435891 0.606407 0.5507
R-squared 0.028406 Mean dependent var 1.670870
Adjusted R-squared -0.017861 S.D. dependent var 4.798540
S.E. of regression 4.841204 Akaike info criterion 6.075145
Sum squared resid 492.1823 Schwarz criterion 6.173884
Log likelihood -67.86417 F-statistic 0.613956
Durbin-Watson stat 1.694248 Prob(F-statistic) 0.442050
(表14)
X3未通过平稳性检验。
ADF Test Statistic 6.958324 1% Critical Value* -3.7497
5% Critical Value -2.9969
10% Critical Value -2.6381
*MacKinnon critical values for rejection of hypothesis of a unit root.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(X4)
Method: Least Squares
Date: 06/03/05 Time: 08:05
Sample(adjusted): 1981 2003
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
X4(-1) 0.282094 0.040541 6.958324 0.0000
C -1.221729 2.736698 -0.446425 0.6599
R-squared 0.697486 Mean dependent var 11.14000
Adjusted R-squared 0.683081 S.D. dependent var 17.73394
S.E. of regression 9.983426 Akaike info criterion 7.522671
Sum squared resid 2093.045 Schwarz criterion 7.621410
Log likelihood -84.51072 F-statistic 48.41827
Durbin-Watson stat 1.871356 Prob(F-statistic) 0.000001
(表15)
X4通过平稳性检验。
很遗憾,由于所学知识有限,所有没有能力对非平稳数据进行调整,有可能导致模型的准确性不足。因此,待掌握足够的知识后再进行调整,提高模型的精确度,使其更具有说服力。