Structural Equation Modelling (SEM) is actually a standard statistical modelling strategy to ascertain associations amid variables.  A important aspect of SEM is observed variables are understood to depict a small number of "latent constructs" that can't be right measured, only inferred through the observed measured variables.So I searched online,