A Durbin-Watson statistic value of which number indicates the absence of autocorrelation?

The Durbin-Watson statistic is a test statistic used to detect the presence of autocorrelation in the residuals from a regression analysis. The value of the Durbin-Watson statistic ranges between 0 and 4. A value around 2 suggests that there is no autocorrelation present in the residuals; it indicates that the residuals are independent of one another.

When the Durbin-Watson statistic approaches 2, it implies that any correlation present among the residuals is negligible. Values significantly lower than 2 indicate positive autocorrelation, while values significantly higher than 2 suggest negative autocorrelation. Therefore, the presence of the value around 2 is a strong indicator of the absence of autocorrelation in the data being analyzed.

In summary, the reason the answer indicating two is correct is that this value effectively signals independence among residuals, which is a key assumption in regression analysis.