Structural Equation Modeling (SEM) is also known as Graphical or Path Modeling, allows the estimation of a causal theoretical network of relationships linking latent complex concepts, by means of a number of observable indicators. The major technical approaches include Covariance Structure Analysis (CSA) and PLS path modeling. CSA is a traditional method of SEM, and called hard modeling, as it makes heavy distributional assumptions and requires hundreds of cases. While the PLS approach is called soft modeling, as it makes looser distributional assumptions and requires only a few cases. PyStat supports CSA for now.

The model consists of three parts: Measurement Modeling, Structural Modeling, and Rsidual Correlation. In each case, when you select a variable on the left side of the dialog box, an expression that specifies the link between the variables is displayed on the right side.

Click the Run button and the network diagram will appear.

Index for model quality are also provided.

The results can be copied and pasted into other software.