自回归系数的Bootstrap检验

Bootstrap test of autocorrelation

在自回归模型中,德宾-沃森(D-W)检验是一种使用非常广泛的自回归系数检验。但D-W检验有一个很大的限制:当检验的统计量落入了所谓"不确定域"则无法进行检验。本文利用Bootstrap重复抽样方法对自回归系数进行检验。蒙特卡罗研究表明,该方法消除了不确定域,改善了检验效果。

The Durbin-Watson（D-W） test is one of the most widely used tests for autocorrelation in regression models.The D-W test has,however,an important limitation：the test is inconclusive when the test statistic falls into the so-called ＂indeterminate range＂.The paper proposes a bootstrap test for autocorrelation.Monte Carlo study shows that the indeterminate range is eliminated with the bootstrap method and the power of the test is impproved.