OpenMx

OpenMx is a free, open-source software package for structural equation modeling (SEM) and related multivariate modeling. It runs within the R programming environment and provides a flexible framework for specifying and estimating latent variable models, path analysis, confirmatory factor analysis, and growth models.

Technically, OpenMx is built around a matrix-based modeling language. Models are defined with objects such as MxMatrix and MxModel, and estimation proceeds via likelihood-based methods with definitions of expected covariance matrices and means. It supports multiple data types (raw data, covariance matrices, or a mix), algebraic constraints, and equality and inequality constraints. It also offers facilities for multigroup and multilevel models, and handles missing data using full-information maximum likelihood.

OpenMx emphasizes transparency and extensibility, with a design that favors model specification over bespoke programming. It integrates with R for data handling, results summarization, and graphics, and provides extensive documentation, tutorials, and example models to aid learning and replication.

Originating as an open-source successor to the commercial Mx software, OpenMx has been developed by researchers and institutions including UCLA and collaborating universities. It has a worldwide user base in psychology, education, and social sciences, where it is used for research, pedagogy, and methodological work.

OpenMx runs on Windows, macOS, and Linux and is released under an open-source license, inviting contributions from the academic community.