Hi all, I am running regression on time series data and getting Durbin-Watson value of "2.55". Is this value is fine to go ahead for interpretation or should I try to reduce the D-W value?
Durbin-Watson tests for autocorrelation in residuals from a regression analysis. The test statistic ranges in between 0 to 4. A value of 2 indicates that there is no autocorrelation. Value nearing 0 (i.e., below 2) indicates positive autocorrelation and value towards 4 (i.e., over 2) indicates negative autocorrelation. It looks like you have a negative autocorrelation case, but wait! Did you compare this value with critical value/p-values that are reported? If you have done this manually, google for Savin and White table that lists critical values for sample size and/or number of regressors. It would test whether the value 2.55 is significantly different from 2.00 (null hypothesis) given the sample size and/or the number of regressors. Hope it helps.
Durbin-Watson tests for autocorrelation in residuals from a regression analysis. The test statistic ranges in between 0 to 4. A value of 2 indicates that there is no autocorrelation. Value nearing 0 (i.e., below 2) indicates positive autocorrelation and value towards 4 (i.e., over 2) indicates negative autocorrelation. It looks like you have a negative autocorrelation case, but wait! Did you compare this value with critical value/p-values that are reported? If you have done this manually, google for Savin and White table that lists critical values for sample size and/or number of regressors. It would test whether the value 2.55 is significantly different from 2.00 (null hypothesis) given the sample size and/or the number of regressors. Hope it helps.
You need to test negative autocorrelation case, because of the test value is dw=2.55>2. Primarily, you must determine the tabulated critical values dL and dU which depend on the sample size and the number of explanatory variables. If dw=2.55>4-dL, you have negative autocorrelation problem in the data set. If dw=2.55
I suggest you follow the way that Pro TUgba has suggested for you. Basing on your sample, number of variables and then identifing the du and dl to test the first order autocorrelation
YES, it is. 0-2 stands for positive auto-correlation of the sample scores. And 2 means no auto-correlation. Auto-correlation means that errors in adjacent scores impact in others. It means that that the sample scores have inter-mutual influence on one another. Hence, significant findings are not strict predictors of behaviours from 0-2. In your own case, 2.55 shows negative auto-correlation indicating that the significant variables tested are good predictors of the behaviours of interests. Thanks.