Suppose an economist has developed a model for forecasting annual consumption, yt, as function of total labor income, x1t , and total property income, x2t based on 20 years on annual data. The following regression model has been developed: t = 7.81 + 0.91x1t + 0.57x2t with the standard error = 1.29 and the Durbin-Watson d statistic = 2.09. Using an alpha = .05, which of the following conclusions should be reached?
A) Conclude that no positive autocorrelation exists
B) Conclude that significant positive autocorrelation exists
C) Uncertain whether positive or negative autocorrelation exists
D) No significant autocorrelation exists.

Question

Quizplus · Accepted Answer

DThe Durbin-Watson d statistic is used to detect the presence of autocorrelation in the residuals from a regression analysis. A d value close to 2 suggests no autocorrelation. Since the d statistic is 2.09, which is very close to 2, it indicates that there is no significant autocorrelation in the model. Therefore, both conclusions that no positive autocorrelation exists (A) and that no significant autocorrelation exists (D) are correct.

Suppose an Economist Has Developed a Model for Forecasting Annual