Share
Explore BrainMass

Serial correlation using Durbin Watson

Using the Durbin-Watson test for first-order serial correlation. Determining whether the results are significant. Detecting impure serial correlation.

Attachments

Solution Preview

a. The Durbin-Watson test is a test for first-order serial correlation in the residuals of a time series regression. A value of 2.0 for the Durbin-Watson statistic indicates that there is no serial correlation. Using the book Statistics for Environmental Science and Management by Manly, I found the necessary critical values for the DW test at the 5% level. For n = 40, p = 3, the critical values of the DW test statistic are 1.25 for d1 and 1.57 for d2. If V is less than 2, it is definitely significant if V < d1, might be significant if d1< V< d2, and is not significant if V>d2. So in your case, where V=0.85 is less than 1.25, it is definitely significant. This indicates a tendency for observations which are close in time to be similar.

b. We want to know if 20 and 2600 are appropriate values for the coefficients of L and P. The null ...

Solution Summary

Using the Durbin-Watson test for first-order serial correlation. Determining whether the results are significant. Detecting impure serial correlation.

$2.19