TOPICS
Search

Durbin-Watson Statistic


The Durbin-Watson statistic is a statistic for detecting first-order serial correlation in the residuals e_1, ..., e_n of a regression. It is defined by

 d=(sum_(t=2)^(n)(e_t-e_(t-1))^2)/(sum_(t=1)^(n)e_t^2).

It satisfies 0<=d<=4, with values near 2 indicating little first-order autocorrelation. If rho^^_1 is the lag-one residual autocorrelation, then d=2(1-rho^^_1).

The null distribution depends on the regression design, and the usual Durbin-Watson inference is not generally valid when a lagged dependent variable is included among the regressors.


See also

Autocorrelation, Ljung-Box Test, Null Distribution, Residual

Explore with Wolfram|Alpha

References

Durbin, J. and Watson, G. S. "Testing for Serial Correlation in Least Squares Regression. I." Biometrika 37, 409-428, 1950. https://doi.org/10.1093/biomet/37.3-4.409.

Cite this as:

Weisstein, Eric W. "Durbin-Watson Statistic." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/Durbin-WatsonStatistic.html

Subject classifications