Title: Tree Models for Assessing Covariate-Dependent Method Agreement

Authors: Siranush Karapetyan, Alexander Hapfelmeier, Achim Zeileis

Abstract:
Method comparison studies explore the agreement of measurements made by two or
more techniques, devices, or methods. With continuously scaled measurements,
agreement is usually evaluated by the well-established Bland-Altman analysis.
However, the underlying assumption is that differences between methods are
identically distributed for all observational units or subjects and in all
application settings. Therefore, we propose the concept of conditional method
agreement that employs differences in covariates to explain differences in the
agreement between two methods. More specifically, a model-based recursive
partitioning approach is developed, called conditional method agreement trees
(COAT). It is accompanied by an R implementation in package "coat", currently
available from https://github.com/Hapfelmeier/coat. Both the method and its
implementation are discussed, illustrated with accelerometer data, and explored
further in a simulation study.