Chartered Financial Analyst (CFA) Practice Exam Level 2

How is the Durbin-Watson statistic interpreted in terms of correlation?

1. A value exactly equal to 2 indicates no correlation

2. A value below 2 indicates negative correlation

3. A value above 2 indicates positive correlation

4. Values near 2 indicate reduced predictive accuracy

The correct answer is: A value exactly equal to 2 indicates no correlation

The Durbin-Watson statistic is a test used to detect the presence of autocorrelation in the residuals from a regression analysis. Its values range from 0 to 4, where a value around 2 suggests that there is no autocorrelation. Specifically, a value exactly equal to 2 indicates that the residuals are uncorrelated, which implies that the errors from one time period to the next are not correlated with each other. This is a crucial indicator in regression analysis as it confirms the validity of the model's assumptions. When interpreting the Durbin-Watson statistic, values less than 2 indicate positive autocorrelation, while values greater than 2 suggest negative autocorrelation. Therefore, the assertion that a value exactly equal to 2 indicates no correlation is accurate and reflects the core purpose of the statistic in regression analysis. Understanding this helps analysts ensure that their regression model meets the assumption of independence of errors, which is vital for reliable inference and forecasting based on the model.