In order to show the effect of treatment, the change between two repeated psychometric measurements at the individual level should be estimated. The simplest method is to calculate the absolute difference between two measurements. However, measurements obtained in a clinical setting are often influenced by other changes not related to the treatment. One of the typical sources of error is regression to the mean (RTM). Iachina and Bilenberg (2012) propose a new method to adjust to the RTM effect. In this work, we will evaluate the performance of absolute difference method and adjusted method in a simulation study.
|Journal||Communications in Statistics: Simulation and Computation|
|Publication status||Published - 2. Jun 2020|
- Regression towards the mean