这个结论需要记忆么？

问题如下：

Question Which of the following statements about a linear trend model and the Durbin–Watson statistic is most accurate?

选项：

A.A.A Durbin–Watson statistic close to 2.0 indicates that the errors in a time-series model are serially uncorrelated B.B.A statistically significantly low value of the Durbin–Watson statistic indicates that the errors in a time-series model are negatively serially correlated C.C.A statistically significantly high value of the Durbin–Watson statistic indicates that the errors in a time-series model are positively serially correlated

解释：

• A is Correct because, if a linear trend model is correctly specified, the Durbin–Watson statistic should not differ significantly from 2.0. Otherwise, the errors in the model are either positively or negatively serially correlated, and that correlation can be used to build a better forecasting model for those time series.

• B is Incorrect because significantly small values of the Durbin–Watson statistic indicate positive serial correlation; significantly large values point to negative serial correlation. So, a significantly low value of the Durbin–Watson statistic indicates that the model errors are positively, not negatively, serially correlated.

• C is Incorrect because significantly small values of the Durbin–Watson statistic indicate positive serial correlation; significantly large values point to negative serial correlation. So, a significantly high value of the Durbin–Watson statistic indicates that the model errors are negatively, not positively, serially correlated.