异方差分析报告

on:DependentVariable:RESID^2Method:LeastSquaresDate:03/30/15Time:12:44Sample:121Includedobservations:21VariableCoefficientStd.Errort-StatisticProb.  C-66423.22189316.3-0.3508580.7296X1043.122418.31222.4936460.0220R-squared0.246578    Meandependentvar351198.3AdjustedR-squared0.206924    S.D.dependentvar454261.0S.E.ofregression404541.2    Akaikeinfocriterion28.74929Sumsquaredresid3.11E+12    Schwarzcriterion28.84877Loglikelihood-299.8675    Hannan-Quinncriter.28.77088F-statistic6.218270    Durbin-Watsonstat1.756144Prob(F-statistic)0.022033拒绝原假设,不拒绝备择假设,表明模型存在异方差。从表5.5可以看出,,由White检验知,在下,查分布表,得临界值(在(5.34)式中只有两项含有解释变量,故自由度为2),比较计算的统计量与临界值,因为>,所以拒绝原假设,不拒绝备择假设,表明模型存在异方差。个人收集整理勿做商业用途表5.5四、异方差性的修正(一)加权最小二乘法(WLS)在运用WLS法估计过程中,我们分别选用了权数。权数的生成过程如下,由图5.4,在对话框中的EnterQuation处,按如下格式分别键入:;;,个人收集整理勿做商业用途W1W2