A Review of Eight Software Packages for Structural Equation Modeling

作者：Narayanan A^*

来源：American Statistician, 2012, 66(2): 129-138.

DOI：10.1080/00031305.2012.708641

• Analysis of moment structures
• Confirmatory factor analysis
• Covariance structure analysis
• Latent variable modeling
• Linear structural relations
• Path analysis

This article reviews eight different software packages for linear structural equation modeling. The eight packages Amos, SAS PROC CALIS, R packages sem, lavaan, OpenMx, LISREL, EQS, and Mplus-can help users estimate parameters for a model where the structure is well specified. Capabilities for handling single group, multiple group, nonnormal variables, and missing data are considered and the eight packages are Compared across a variety of criteria from documentation to parameter estimation. The main difference between the packages is the presence of a graphical interface for model specification and presentation of results. Each package differs in terms of strengths, areas of improvement, and unique features that may dictate the choice of selection. Some suggestions on areas of improvement for all software packages are made. This article has supplementary materials online.

• 出版日期2012-5