E. Chasseloup

Extension of individual model averaging assessments to unbalanced designs and dose-response

This study evaluates two NLME model approaches for drug effect assessment in unbalanced designs and dose-response studies, finding that IMA and sIMA control type I error and bias, while STD does not.

Assessment of non-linear mixed effects model-based approaches to test for drug effect using simulated data: type I error and power properties

This paper evaluates IMA and STD methods for drug effect testing using simulated data, finding IMA more robust to model misspecification with controlled type I error and unbiased estimates.