R2: a useful measure of model performance when predicting a dichotomous outcome

✍ Scribed by Arlene Ash; Michael Shwartz

• Publisher: John Wiley and Sons
• Year: 1999
• Tongue: English
• Weight: 107 KB
• Volume: 18
• Category: Article
• ISSN: 0277-6715
• DOI: 10.1002/(sici)1097-0258(19990228)18:4<375::aid-sim20>3.0.co;2-j

✦ Synopsis

R has been criticized as a measure of model performance when predicting a dichotomous outcome, both because its value is often low and because it is sensitive to the prevalence of the event of interest. The C statistic is more widely used to measure model performance in a 0/1 setting. We use a simple parametric family of models to illustrate the potential usefulness of models with low R values, to clarify the effect of prevalence on both C and R, and to demonstrate how R captures information not picked up by C. We also show that C is subject to a 'random mixing' problem that does not affect R. Finally, we report both R and C values for different risk-adjustment models in situations with different prevalences and show the relationship between the measures and decile death rates, thereby providing a context for interpreting R values in a 0/1 setting.