SEMinR: Domain-specific language for building, estimating, and visualizing structural equation models in R

Abstract

SEMinR allows researchers to easily create, estimate, and visualize structural equation models (SEMs) for multiple estimation methods. SEMs are popular modeling techniques in social sciences and the life sciences, and can estimate relationships between concepts that need to be measured by multiple items. SEMinR can estimate SEMs using either covariance-based SEM (CBSEM, such as found in LISREL and Lavaan), or Partial Least Squares Path Modeling (PLS-PM, such as found in SmartPLS, semPLS, plspm, and csem). Moreover, SEMinR implements several advances in SEM methodologies not found elsewhere. And it also allows for visualization of all kinds of SEM models. SEMinR’s model description syntax is plain-old-R-functions (PORF!), which allows users to extend and compose syntax in novel ways. Thus, SEMinR is a one-stop-shop for both SEM practitioners seeking to analyze empirical models and SEM methodologists seeking to automate and extend SEM methods. SEMinR is increasingly being used in universities, for both research and teaching needs, and companies world-wide.