Mplus Trees: Structural Equation Model Trees Using Mplus. Issue 1 (2nd January 2021)

• Record Type: Journal Article
• Authors: Serang, Sarfaraz; Jacobucci, Ross; Stegmann, Gabriela; Brandmaier, Andreas M.; Culianos, Demi; Grimm, Kevin J.
• Is Part Of: Structural equation modeling. Volume 28:Issue 1(2021)
• Page Start: 127
• Page End: 137
• Publication Date: 2021-01-02
• Subjects: Structural equation modeling -- decision trees -- data mining -- machine learning; Multivariate analysis -- Periodicals; Social sciences -- Statistical methods -- Periodicals; 519.535
• DOI: 10.1080/10705511.2020.1726179
• Languages: English
• ISSNs: 1070-5511

Abstract

ABSTRACT: Structural equation model trees (SEM Trees) allow for the construction of decision trees with structural equation models fit in each of the nodes. Based on covariate information, SEM Trees can be used to create distinct subgroups containing individuals with similar parameter estimates. Currently, the structural equation modeling component of SEM Trees is implemented in the R packages OpenMx and lavaan. We extend SEM Trees so that the models can be fit in M plus, in the hopes that its efficiency and accessibility allow a broader group of researchers to fit a wider range of models. We discuss the M plus Trees algorithm, its implementation, and its position among the growing number of tree-based methods in psychological research. We also provide several examples using publicly available data to illustrate how M plus Trees can be implemented in practice with the R package MplusTrees.