02 February 2018 3 8K Report

Hello,

When analysing scores between 2 treatment groups, I am often confronted to violation of odds ratio proportionality assumption.

One recommendation is the use of partial proportional odds (PPO) for the analysis.

J. Whitehead. "Sample size calculation for ordered categorical data", 1994. presented a method for sample-size calculation when the assumption of proportionional odds is met.

But how can we calculate the required sample-size when this assumption is not met? (under SAS or R)

Thank you in advance,

Similar questions and discussions