Can psychometric measurement models inform behavior genetic models? A bayesian model comparison approach

Abstract

As methodologists have increasingly noted, the role of psychometrics in operationalizing a construct is often overlooked when evaluating research claims (Borsboom 2006). In a related vein, others have noted that psychological research appears to move away from assessment and interpretation of a single a priori statistical model to a more nuanced comparison of models which assess the trade-off between a model’s parsimony and complexity in explaining behavior (e.g., Rodgers 2010). The genetic factor model is one such statistical model often used to estimate the relative contributions of genetic and environmental components of observed behavior in genetically informative designs (Heath, Neale, Hewitt, Eaves, & Fulker 1989; Martin & Eaves 1977; Neale & Cardon 1992). Mathematically, the genetic factor model decomposes observed phenotypic variability into additive genetic (A), common (C), and unique (E) environmental components and is, for that reason, often referred to as the ACE model.