The Durbin–Watson test introduces a statistic d that is used to test the autocorrelation of the residuals obtained from a linear regression model. This is a problem that often appears during the application of a linear model to a time series, when we want to test the independence of the residuals obtained in this way.
HISTORY
Durbin, J. and Watson, G.S. invented this test in 1950.
MATHEMATICAL ASPECTS
Consider the case of a multiple linear regression model containing \( { p-1 } \) independent variables. The model is written:
where
- Y t :
-
is the dependent variable,
- X jt ,:
-
with \( { j=1, \ldots, p-1 } \) are the independent variables
- β j ,:
-
with \( { j=1, \ldots, p-1 } \) are the parameters to be estimated,
- ε t :
-
with \( { t=1, \ldots, T } \) is an unobservable random error term.
In the matrix form, the model is written as:
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
REFERENCES
Bourbonnais, R.: Econométrie, manuel et exercices corrigés, 2nd edn. Dunod, Paris (1998)
Durbin, J.: Alternative method to d-test. Biometrika 56, 1–15 (1969)
Durbin, J., Watson, G.S.: Testing for serial correlation in least squares regression, I. Biometrika 37, 409–428 (1950)
Durbin, J., Watson, G.S.: Testing for serial correlation in least squares regression, II. Biometrika 38, 159–177 (1951)
Harvey, A.C.: The Econometric Analysis of Time Series. Philip Allan, Oxford (Wiley, New York) (1981)
Rights and permissions
Copyright information
© 2008 Springer-Verlag
About this entry
Cite this entry
(2008). Durbin–Watson Test. In: The Concise Encyclopedia of Statistics. Springer, New York, NY. https://doi.org/10.1007/978-0-387-32833-1_122
Download citation
DOI: https://doi.org/10.1007/978-0-387-32833-1_122
Publisher Name: Springer, New York, NY
Print ISBN: 978-0-387-31742-7
Online ISBN: 978-0-387-32833-1
eBook Packages: Mathematics and StatisticsReference Module Computer Science and Engineering