Online Article

Comparison between CB-SEM and PLS-SEM

13th January 2023

Related topic: Quantitative research

Author: Ady Hameme N. A.

Structural equation modeling (SEM) is a powerful statistical technique that is used to analyze and understand complex relationships between variables. There are two main types of SEM: covariance-based SEM (CB-SEM) and partial least squares SEM (PLS-SEM). CB-SEM and PLS-SEM are similar in many ways, but they also have some important differences. In this essay, we will explore the main differences between CB-SEM and PLS-SEM based on the literature.

One of the main differences between CB-SEM and PLS-SEM is the type of data they can handle. CB-SEM is typically used for data that is normally distributed, while PLS-SEM is used for data that is not normally distributed. This is because PLS-SEM is based on the partial least squares (PLS) algorithm, which is a technique that can handle non-normally distributed data by reducing the dimensionality of the data. This means that PLS-SEM can be used for a wider range of data types, including data that is ordinal, binary, or categorical.

Another difference between CB-SEM and PLS-SEM is the way in which they estimate model parameters. CB-SEM uses maximum likelihood estimation (MLE) to estimate model parameters, while PLS-SEM uses a technique called PLS path modeling. MLE is a well-established and widely used technique that is based on the likelihood function, which measures the goodness of fit of a model. PLS path modeling, on the other hand, is based on the PLS algorithm, which uses the relationship between the independent and dependent variables to estimate model parameters. This means that PLS-SEM is less sensitive to outliers and measurement errors, which can be a problem in CB-SEM.

A third difference between CB-SEM and PLS-SEM is the way in which they test hypotheses. CB-SEM uses chi-square tests and other goodness-of-fit measures to test hypotheses, while PLS-SEM uses bootstrap methods and other resampling techniques to test hypotheses. Bootstrap methods are a powerful technique that allows researchers to estimate the uncertainty in the model parameters and to test hypotheses. This means that PLS-SEM is less sensitive to the assumptions of the model and can be used to test hypotheses with more certainty.

In conclusion, CB-SEM and PLS-SEM are both powerful techniques for analyzing and understanding complex relationships between variables. However, they have some important differences, including the type of data they can handle, the way in which they estimate model parameters, and the way in which they test hypotheses. Researchers should carefully consider these differences when choosing which technique to use in their research.

Cite this article: Ady Hameme, N. A. (2023, January 13). Comparison between CB-SEM and PLS-SEM. Retrieved <insert month> <insert date>, <insert year>, from https://www.myadvrc.com/publications/article-10

< previous next >