Partial Least Squares - Structural Equation Modelling

SEM is a statistical approach to analyze a hypothetical framework simultaneously, allowing researcher to test and develop theories and test the relationships between latent variables on the conceptual model. The two types of SEM are: Covariance-based technique (CB-SEM) and Variance-based or Partial Least Squares SEM (PLS-SEM). CB-SEM is confirmatory in nature and should meet some assumptions and requirements such as; the normality of data, and a minimum sample size based on the complexity of model, whereas PLS-SEM can handle theory development and no need to meet the model’s assumption and requirement such as normality of the data, and large sample size. The PLS-SEM is a variance-based approach and aims to maximize explained variance of dependent latent constructs; whereas the CB-SEM aims to reproduce the theoretical covariance matrix. Therefore the PLS-SEM is preferred when the researcher wants to predict construct and identify relationship between them. 

Two common and powerful software programs will be employed for PLS-SEM analysis namely: WarpPLS 6.0 and SmartPLS 3.0.