Skip to content
Analysing within treatment arms is nonsense!
Branch: master
Clone or download


I perform a simulation of a randomised control trial demonstrating why it is wrong to analyse arms of a trial separately looking to identify responders and non responders. There are two examples the main difference being one simulation is based on parameters I made up whilst the other used the example in Stephen Senn's BMJ paper (see the reference in my example)

Please refer to the free published paper in this repository

You can’t perform that action at this time.