A test of the null hypothesis of no serial correlation against the alternative of first-order serial correlation in the error term in a linear regression. The test statistic is based on the residuals from the least squares regression, and under the null hypothesis has a non-standard distribution that depends on the values of the explanatory variables. The upper bound and the lower bound on the critical values, that do not depend on the explanatory variables and only depend on the sample size and the number of regressors, are tabulated using Monte Carlo simulations; the Durbin–Watson bounds test uses these bounds rather than the critical values. This test is not valid if the regression contains the lagged dependent variable among explanatory variables or there is no intercept; it cannot be used to test for higher-order serial correlation; it is inconclusive when the value of the test statistic falls between the upper bound and the lower bound of critical values.
Subjects: Economics — Probability and Statistics.