umx

umx facilitates structural equation modeling (SEM) in R, enabling specification, estimation, and comparison of path-based and matrix-based models for social sciences and behavior genetics, including multigroup twin ACE and gene × environment models.

Key Features:

• Path-Based Model Specification: Enables specification and estimation of path-based structural equation models for representing directed relationships among observed and latent variables.
• Matrix-Based Models: Supports matrix-based model specification and estimation, including handling of raw and covariance data inputs and joint ordinal datasets.
• Behavior Genetic Models: Implements standard multigroup twin models, including ACE (Additive genetic, Common environment, Unique environment) models, incorporation of covariates, and both common- and independent-pathway models.
• Gene × Environment Interaction Models: Provides functionality for modeling gene × environment interaction within SEM frameworks.
• Graphical and Tabular Outputs: Produces graphical and tabular results for interpretation and presentation of model estimates and comparisons.

Scientific Applications:

• Social Sciences: Specification and testing of theoretical models with latent variables, path dependencies, and complex relationships among observed variables using SEM.
• Behavior Genetics: Analysis of twin and multigroup genetic data to estimate heritability, shared and unique environmental influences, pathway structures, and gene × environment interactions.

Methodology:

Implements path-based and matrix-based SEM methods, including standard multigroup twin models (ACE and common/independent-pathway models), and supports analysis of raw, covariance, and joint ordinal datasets.