How to write CONTRAST and ESTIMATE statements in SAS regression procedures

1

I got several positive comments about a recent tip, "How to fit a variety of logistic regression models in SAS." A reader asked if I knew any other similar resources about statistical analysis in SAS.

Absolutely! One gem that comes to mind is "Examples of writing CONTRAST and ESTIMATE statements." SAS statistical programmers often ask how to write CONTRAST and ESTIMATE statements on discussion forums such as the SAS Support Community for Statistical Procedures.

How to write CONTRAST and ESTIMATE statements in #SAS regression procedures. #Statistics Click To Tweet

The Knowledge Base article features regression models that you might encounter in PROC GLM, PROC LOGISTIC, and PROC GENMOD. The article includes the following topics:

  • How to express certain hypotheses as linear combinations of regression coefficients.
  • Why you must know the order of parameters for classification variables to properly write CONTRAST and ESTIMATE statements.
  • How to write CONTRAST and ESTIMATE statements for interaction effects.
  • How to specify linear combinations that include fractions like 1/3 or 1/6 that cannot be expressed as a terminating decimal value.
  • How to estimate or test contrasts of log odds in logistic models that use either GLM or EFFECT (deviation from the mean) encodings.
  • How to use the CONTRAST statement to compare nested models.

The article is written for a technical audience, and the examples are complex. However, if you are statistically sophisticated analyst, then "Examples of Writing CONTRAST and ESTIMATE Statements" is an excellent tutorial. Bookmark this page in case you need it!

And if you still have questions after reading the article, remember that the SAS Support Community is just a click away.

Share

About Author

Rick Wicklin

Distinguished Researcher in Computational Statistics

Rick Wicklin, PhD, is a distinguished researcher in computational statistics at SAS and is a principal developer of PROC IML and SAS/IML Studio. His areas of expertise include computational statistics, simulation, statistical graphics, and modern methods in statistical data analysis. Rick is author of the books Statistical Programming with SAS/IML Software and Simulating Data with SAS.

1 Comment

  1. Michelle Homes

    Good tip Rick!

    A few years ago Chris Daman wrote a series of blog posts that your readers may also find useful:
    1) The magical ESTIMATE (and CONTRAST) statements - http://blogs.sas.com/content/sastraining/2012/04/23/the-magical-estimate-and-contrast-statements/
    2) "Easy button" for ESTIMATE statements - http://blogs.sas.com/content/sastraining/2012/04/25/easy-button-for-estimate-statements/
    3) ESTIMATE Statements - the final installment - http://blogs.sas.com/content/sastraining/2012/05/02/estimate-statements-the-final-installment/

Leave A Reply

Back to Top