Hi, I am doing my analysis based on time series data. My regression results reveal a very low DW statistic. Can anybody let me know the adjustment that needs to be done to the regression equation to account for autocorrelation?
Regards,
Sewwandi.
The Durbin-Watson test is used to determine if the residuals from your model have significant autocorrelation. So you look at the p-value for the test and conclude that there is autocorrelation if the p-value is small (usually taken as less than 0.05). If you do have a p-value less than 0.05, a way to account for this would be to construct a model that includes residual correlation structure such as an autoregressive model for the residuals.
Dear Thushari,
To do it by hand, you need to: 1. Save the residual (u) from the first run. 2. Estimate the autocorrelation (p) with the residual . A first order estimate will be u(t) =p.u(t-1)+ error(v(t)). 3. Use the estimate of p in the following procedure from Damordar Gujarati book "Essentials of Econometrics." The same procedure can be found in Johnson's Book on Econometrics. Notice that this way you will loose one observation. This is a little more detail than Andrew Pegman has suggested.
Thank you Dr. Lall Ramrattan!
What you have suggested and the document attached are very much helpful!
Thank you once again!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics