HOME Module List Site Index About This Site glossary" Resources Innovative Technology Center UT Statistics Courses
Using SAS Choose Design ANOVA Compare Means Regression Examples
ANOVA Diagnostics Start   

D1. Check covariate model (continued)

Three examples of covariate output are provided here. They differ in the number of treatment factors. Choose single if you only have one treatment factor, double if you have two (factorial, split-plot, strip-plot, repeated measures), and triple if you have a split-split-plot.

                               


If any of the three tests illustrated in the popups are significant, add the significant term to the %MMAOV statement in the FIXED= list and rerun the SAS analysis.

For example, the standard analysis:
fixed = age package age*package times
becomes
fixed = age package age*package times times*times*age
if times*times*age is significant.

Be aware that if interactions of covariate with treatment are included, you are switching to "dummy" regression, and you may want to study such models under the Regression tab.

If all tests are non-significant, then the standard covariate model is appropriate, and you can go back to Step D.


next >> ( return to Step D )

 


  H I N T S :
  Remember, data transformations can fix apparent outliers.
  Do not be concerned if you have to rerun your SAS analysis many times. Correction of diagnostics may take
      several tries.
  To see an example using two covariates, see the Examples resource.

Home | Contact us | Module list & summary | Glossary/Terms | About this site | Stats courses | Links | Index