“…As of this writing, this is the only model so far developed explicitly for analyses of evolutionary integration. Consequently, nearly all empirical studies of evolutionary integration and modularity assume that phenotypic evolution can be modeled as Brownian motion, applying methods commonly used to study integration and modularity to the Brownian rate matrix ( R ), such as PLS (Alvarez et al, 2015; Bright et al, 2016; Hedrick et al, 2019; Klingenberg & Marugan‐Lobon, 2013; Randau & Goswami, 2017); RV coefficient (Alvarez et al, 2015; Klingenberg & Marugan‐Lobon, 2013); Covariance Ratio (Dellinger et al, 2019; Du et al, 2019; Perez et al, 2020), and EMMLi (Evans et al, 2019; Felice et al, 2019; Hedrick et al, 2019). If the Brownian model is not true, the parameter estimates, that is, the evolutionary covariances among traits and evolutionary rates, are not correctly estimated.…”