In what circumstances would you use the Durbin h-statistic when testing for autocorrelation?
The Durbin Watson-D statistic can used to determine the autocorrelation at lag in the residuals from the regression analysis. This test the value is ranges between zero and four. The value is ranges between 0 to 2 is positive autocorrelation and value between 2 to 4 is negative autocorrelation. The common Durbin-Watson statistic cannot used for testing autocorrelation in regression models containing lagged values in dependent variable as independent variable. In such cases the Durbin- Watson H statistic will use.
Get Answers For Free
Most questions answered within 1 hours.