Fits and interprets mixed-effects models for clustered,
longitudinal and heterogeneous biological data. Provides variance
partitioning, intraclass correlation, penalized likelihood summaries,
a heterogeneous-data information criterion, model comparison,
diagnostics, and ensemble-style summaries for multilevel data. The
package is designed as a complementary, interpretable workflow around
established mixed-model methods. Methods for intraclass correlation
and variance partitioning are informed by Nakagawa and Schielzeth
(2010)
BioMixModel provides an interpretable workflow for mixed-effects analysis of clustered, longitudinal and heterogeneous biological data. Version 1.0.0 focuses on model fitting, variance partitioning, ICC, heterogeneity-aware AIC (HAIC), comparison and diagnostics, while using established nlme machinery for random-effects estimation.