The University of Tennessee Health Science Center

Simpson’s Paradox: Can a new Therapy be ‘better’ for both High Risk and Low Risk patients but ‘worse’ for everyone?

by Mehmet Kocak, PhD, UTHSC Dept. of Preventive Medicine

Abstract:

Simpson’s Paradox occurs when an observed association between two factors of interest changes its direction when a third factor of interest is used. In this talk, we will give examples of Simpson’s paradox in categorical, count, and survival data analysis,  and discuss ways to avoid its occurrence.

Everybody with an interest in the statistical issues we are discussing is welcome!