Р进行广义差分得:РDependent Variable: SER01-0.52595*SER01(-1)РMethod: Least SquaresРDate: 06/11/11 Time: 21:23РSample (adjusted): 1979 2009РIncluded observations: 31 after adjustmentsРVariableРCoefficientРStd. ErrorРt-StatisticРProb. РCР-790935.0Р298581.0Р-2.648979Р0.0133РSER02-0.52595*SER02(-1)Р0.591267Р0.122193Р4.838809Р0.0000РSER03-0.52595*SER03(-1)Р43475451Р5798252.Р7.498027Р0.0000РSER04-0.52595*SER04(-1)Р0.481626Р0.087227Р5.521523Р0.0000РR-squaredР0.988094Р Mean dependent varР5992200.РAdjusted R-squaredР0.986771Р S.D. dependent varР6651978.РS.E. of regressionР765087.7Р Akaike info criterionР30.05328РSum squared residР1.58E+13Р Schwarz criterionР30.23831РLog likelihoodР-461.8259Р F-statisticР746.9257РDurbin-Watson statР1.549372Р Prob(F-statistic)Р0.000000Р由DW值可以判断,模型依然存在自相关。