Check for outliers, equal variance, normality, and model validity

E. Finish the analysis.

If you identified problems in Steps A through D, the SAS analysis needs to be rerun. Choose one of the red arrows below, which will lead you back through your ANOVA module to re-analyze your data, and then back to this page.

If you did not identify diagnostics problems in Steps A through D, go to You are done!

Choose this (Step F) if your diagnostics showed unequal variance or non-normality: Transform your data.

If you have transformed your data and you still have diagnostic problems, go through the transformations step again and choose a different transformation.

 
Choose this arrow if/when you have equal variance and normality but you removed outliers in Step A: Return to your analysis module.
       To do this, use your browser's Back button, or click on the ANOVA or Choose Design tabs to start over. In SAS, you simply click the Submit Icon again and the new dataset without the outlier(s) will be analysed.

After you re-analyze, you will be led back to this page to ensure that all problems have been corrected.

 next >> ( You are done! ).