What indicates no evidence of autocorrelation in the Durbin Watson statistic?

• A. Value of 0
• B. Value of 1
• C. Value of 2
• D. Value of 3

Answer: (C)

The Durbin-Watson statistic is a test statistic used to detect the presence of autocorrelation in the residuals from a regression analysis. The statistic ranges from 0 to 4, where a value around 2 suggests that there is no autocorrelation.

A value of 2 indicates that there is no first-order autocorrelation present in the residuals. This means that the residual values are not correlated with each other and that the assumption of independence between residuals holds true. When the Durbin-Watson statistic is significantly below 2, it suggests positive autocorrelation, while a value significantly above 2 suggests negative autocorrelation.

Thus, a value of 2 serves as the benchmark indicating no evidence of autocorrelation in the data, making it the correct choice for this question.