Model baseline as:

Use the radio buttons to specify how you want to use the baseline measurements as the standard of comparison, in a repeated-measures ANOVA model.

Your options include Covariate or Response.

Selection	Description
Covariate	• When you select covariate, the first observation for each subject is removed. The deleted values are then averaged. This average value is entered as the first fixed effect and used as a covariate in the repeated-measures anova model. • Inference is conditional on the covariate values.
Response	• When you select response, the baseline measurements are treated the same as the trial measurements in the repeated-measures anova model. • Inference is unconditional. • You must account for no treatment effect at the baseline, often by means of an effect with two values indicating whether the measurement is baseline or not, and forming interactions of this effect with other effects in the model.

For Additional Information

Refer to SAS for Mixed Models¹ for additional details about Repeated Measures anova.

---

1

Littell, R.C., G.A. Milliken, W.W.Stroup, R.D. Wolfinger, and O. Schabenberger. 2006. SAS for Mixed Models, Second Edition Cary, NC: SAS Institute Inc..