Question

After a linear forecasting model is found for a time series, if the Durbin-Watson statistic is less than dL this means that:

a. no positive autocorrelation exists, a nonlinear model should be tried.

b. positive autocorrelation exists, a nonlinear model should be tried.

c. no positive autocorrelation exists, the linear model is adequate.

d. positive autocorrelation exists, the linear model is adequate.

explain please

Answer 1

Solution :

Null hypothesis : A positive autocorrelation do not exist between the error terms.

Alternative hypothesis : A positive autocorrelation do not exist between the error terms.

If Durbin Watson test statistic is less than dL, then we reject the null hypothesis and conclude that positive autocorrelation exists.

Also in a linear model, error terms should not have autocorrelation.

For the given scenario a positive autocorrelation exists because Durbin Watson test statistic is less than dL. And hence, linear model is not adequate. A non linear model should be tried.

Hence, option (b) will be correct.

Please rate the answer. Thank you.