Steps in Forecasting and Other Issues
Assuming critical values of 1.39 and 1.51, is there _most likely_ serial correlation of the residuals for the slope coefficient?
Incorrect.
If the actual Durbin–Watson were outside of the critical value ranges, serial correlation _would_ be present.
Precisely.
The Durbin–Watson statistic is used to test for serial correlation. Using the chart, the Durbin–Watson critical values for one independent variable with 35 degrees of freedom are 1.39 and 1.51. Since the actual Durbin–Watson statistic is 0.5522 and is below the lower critical value, there's evidence of the residuals being positively serially correlated.
Incorrect.
If the Durbin–Watson statistic were inside the critical ranges, there would be no indication of serial correlation.
Yes, because the Durbin–Watson statistic is within its critical values
Yes, because the Durbin–Watson statistic is outside its critical values
No, because the actual Durbin–Watson is outside the critical value ranges.
