# Tip of the Day

## Contents## sasCommunity Tip of the DayAutocorrelation of the error function is something that needs to be addressed in linear regression. If strong autocorrelation exists, then autoregressive models will be more appropriate as opposed to linear regression for the independence assumption of linear regression is severely violated. If no such strong autocorrelation exist, for beginnners, we have no additional reasons to reject linear regression model as a suitable model. To test for serial autocorrelation in linear regression, one should use the DW option in PROC REG. PROC REG; MODEL Y = X/DW; RUN;QUIT; It should be noted that the Durbin-Watson statistics are not relevant in certain scenarios where normality assumptions are violated or when variables which are extremely time-dependent (lag variables) are used. In these cases, the more relevant serial correlation test which is the Breusch-Godfrey test will be more relevant. Submitted by Murphy Choy. Contact me at my Discussion Page. ....see also ....see also ....see also
Feel free to comment on this tip. |

Prior tip - Next tip -- Random Tip

## Tips Index by StatusCurrently in Use | Yet to be Scheduled | Still Under Review | Tip Awaiting Refinement |
## Search Tips: |

### The Tip of the Day Project

A new technical **Tip of the Day** related to SAS software will be shown on this page and the sasCommunity.org **main page** every day. Before getting started, you can **Browse and Search** the current list of tips contributed by users.

Anyone can participate. We hope you do! Here's how:

**Submit a Tip**by posting a SAS tip that you think others might find useful**Review a Tip**prior to publishing

**tell us what you think**