Structural Equation Modeling

Structural Equation Modeling (SEM) is a sophisticated statistical approach that enables researchers to explore and analyze the relationships between observed variables and underlying latent constructs. It effectively combines principles from factor analysis, which identifies underlying factors from observed variables, and multiple regression analysis, which assesses how one set of variables predicts another.

In SEM, we distinguish between two types of variables:

  • Endogenous variables, which are analogous to dependent variables in traditional analyses, representing outcomes or effects.
  • Exogenous variables, similar to independent variables, are considered the causes or predictors that influence endogenous variables.

Theoretical Framework of SEM

SEM operates on two foundational models:

  • Measurement Model: This component of SEM delineates how observed variables relate to their respective latent constructs, essentially defining the operationalization of theoretical concepts.
  • Structural Model: This model outlines the hypothesized relationships among latent constructs themselves, providing a conceptual map of how different constructs influence one another.

SEM is particularly valued for its ability to estimate complex relationships involving multiple dependent and independent variables within a unified analytical framework. This capacity makes it an ideal tool for testing theoretical models that propose causal pathways and interdependencies among variables.

One of the critical aspects of SEM is its reliance on certain statistical assumptions, including the need for multivariate normal distribution of the variables for the application of maximum likelihood estimation methods. Adherence to these assumptions is crucial for the validity of the chi-square test of model fit, an essential component of SEM that assesses how well the proposed model represents the data.

SEM is often referred to as causal modeling due to its utility in testing hypothesized causal relationships between variables. It provides a comprehensive method for researchers to test and refine theoretical models, making it an indispensable tool in the social sciences, psychology, education, and beyond. By offering insights into the direct and indirect relationships between variables, SEM facilitates a deeper understanding of the underlying mechanisms driving observed phenomena.

Need help with your research?

Schedule a time to speak with an expert using the calendar below.

User-friendly Software

Transform raw data to written interpreted APA results in seconds.

Essential Assumptions in SEM


SEM operates on the assumption of a linear relationship between endogenous (dependent) and exogenous (independent) variables. This linearity is crucial for the accurate estimation of the relationships among variables.

Outlier Management

Data utilized in SEM should be free of outliers, as outliers can significantly impact the model’s significance and distort the results.

Sequence and Causality

A clear cause-and-effect relationship between endogenous and exogenous variables is required. Importantly, the cause must precede the effect, establishing a temporal sequence that supports causality.

Non-spurious Relationships

The observed covariance between variables must reflect a true relationship and not be the result of spurious or confounding factors.

Model Identification

For a model to be considered viable, it must be properly identified. This means the number of equations must exceed or equal the number of estimated parameters, aiming for models that are over-identified or exactly identified to ensure solvability and meaningfulness.

Sample Size Considerations

A sample size ranging from 200 to 400, with 10 to 15 indicators per variable, is generally recommended. This guideline serves as a rough estimate, suggesting 10 to 20 times as many cases as variables to ensure sufficient data for reliable analysis.

Independence of Error Terms

Error terms within the model should not correlate with each other or with other variables’ error terms, maintaining the integrity of the model’s estimates.

Data Requirements

SEM typically requires interval-level data to accurately model the relationships between variables.

Steps in SEM

Defining Constructs

The initial step involves theoretically defining the constructs and conducting pretests to evaluate the items. Confirmatory Factor Analysis (CFA) is then used to confirm the measurement model.

Developing the Measurement Model

Also known as path analysis, this step establishes the relationships between exogenous and endogenous variables through directional arrows, adhering to the principle of unidimensionality.

Study Design for Empirical Results

This phase requires specifying the model and designing the study to minimize identification problems, utilizing order condition and rank condition methods to address potential issues.

Assessing Measurement Model Validity

Known as CFA, this stage involves comparing the theoretical measurement model against actual data to assess construct validity.

Specifying the Structural Model

Structural paths between constructs are delineated, ensuring no arrows enter an exogenous construct and that each hypothesized relationship accounts for one degree of freedom. The model may be recursive or non-recursive.

Examining Structural Model Validity

The final step involves validating the structural model. A good fit is indicated by an insignificant chi-square test value and satisfactory performance on incremental fit indexes (CFI, GFI, TLI, AGFI) and badness of fit indexes (RMR, RMSEA, SRMR).

These steps and assumptions guide researchers through the complex yet powerful process of SEM, allowing for a nuanced understanding of the relationships among variables in various fields of study.


Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411-423.

Bentler, P. M., & Chou, C. -P. (1987). Practical issues in structural modeling. Sociological Methods & Research, 16(1), 78-117.

Bollen, K. A. (1989). Structural equations with latent variables. New York: John Wiley & Sons. View

Bollen, K. A. (1990). Overall fit in covariance structure models: Two types of sample size effects. Psychological Bulletin, 107(2), 256-259.

Bollen, K., & Lennox, R. (1991). Conventional wisdom on measurement: A structural equation perspective. Psychological Bulletin, 110(2), 305-314.

Boomsma, A. (2000). Teacher’s corner: Reporting analyses of covariance structures. Structural Equation Modeling: A Multidisciplinary Journal, 7(3), 461-483.

Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum Associates. View

Byrne, B. M. (2001). Structural equation modeling with AMOS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum Associates.

Byrne, B. M. (2004). Testing for multigroup invariance using AMOS Graphics: A road less traveled. Structural Equation Modeling, 11(2), 272-300.

Chen, F., Bollen, K. A., Paxton, P., Curran, P. J., & Kirby, J. B. (2001). Improper solutions in structural equation models: Causes, consequences, and strategies. Sociological Methods and Research, 29(4), 468-508.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9(2), 233-255.

Curran, P. J., Bollen, K. A., Paxton, P., Kirby, J., & Chen, F. (2002). The noncentral chi-square distribution in misspecified structural equation models: Finite sample results from a Monte Carlo simulation. Multivariate Behavioral Research, 37(1), 1-36.

Fan, X., Thompson, B., & Wang, L. (1999). Effects of sample size, estimation method, and model specification on structural equation modeling fit indexes. Structural Equation Modeling, 6(1), 56-83.

Hatcher, L. (1994). A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, NC: SAS Institute.

Hipp, J. R., & Bollen, K. A. (2003). Model fit in structural equation models with censored, ordinal, and dichotomous variables: Testing vanishing tetrads. Sociological Methodology, 33, 267-305.

Hoyle, R. H. (Ed.). (1995). Structural equation modeling: Concepts, issues, and applications. Thousand Oaks, CA: Sage Publications.

Jöreskog, K. G. (1970). A general method for estimating a linear structural equation system (Report No. RB-70-54). Princeton, NJ: Educational Testing Service.

Jöreskog, K. G., & Yang, F. (1996). Non-linear structural equation models: The Kenny-Judd model with interaction effects. In G. A. Marcoulides & R. E. Schumacker (Eds.), Advanced structural equation modeling (pp. 57-88), Mahwah, NJ: Lawrence Erlbaum Associates.

Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). New York: Guilford Press. View

Lee, S., & Hershberger, S. (1990). A simple rule for generating equivalent models in covariance structure modeling. Multivariate Behavioral Research, 25(3), 313-334.

Lee, S. (2007). Structural equation modeling: A Bayesian approach. New York: John Wiley & Sons. View

Maruyama, G. M. (1998). Basics of structural equation modeling. Thousand Oaks, CA: Sage Publications. View

McDonald, R. P., & Ho, M. -H. R. (2002). Principles and practice in reporting structural equation analyses. Psychological Methods, 7(1), 64-82.

Mueller, R. O. (1996). Basic Principles of structural equation modeling: An introduction to LISREL and EQS. New York: Springer-Verlag. View

Mulaik, S. A., & Millsap, R. E. (2000). Doing the four-step right. Structural Equation Modeling, 7(1), 36-73.

Olsson, U. H., Foss, T., Troye, S. V., & Howell, R. D. (2000). The performance of ML, GLS, and WLS estimation in structural equation modeling under conditions of misspecification and nonnormality. Structural Equation Modeling, 7(4), 557-595.

Raykov, T. (2000). On the large-sample bias, variance, and mean squared error of the conventional noncentrality parameter estimator of covariance structure models. Structural Equation Modeling, 7(3), 431-441.

Raykov, T. (2005). Bias-corrected estimation of noncentrality parameters of covariance structure models. Structural Equation Modeling, 12(1), 120-129.

Raykov, T., & Marcoulides, G. A. (2006). A first course in structural equation modeling (2nd ed.). New York: Lawrence Erlbaum Associates. View

Raykov, T., Tomer, A., & Nesselroade, J. R. (1991). Reporting structural equation modeling results in Psychology and Aging: Some proposed guidelines. Psychology and Aging, 6(4), 499-503.

Schreiber, J. B. (2008). Core reporting practices in structural equation modeling. Research in Social & Administrative Pharmacy, 4(2), 83-97.

Schumacker, R. E. (2002). Latent variable interaction modeling. Structural Equation Modeling, 9(1), 40-54.

Schumacker, R. E., & Lomax, R. G. (2004). A beginner’s guide to structural equation modeling (2nd ed.). London: Routledge. View

Shipley, B. (2000). Cause and correlation in Biology: A user’s guide to path analysis, structural equations and causal inference. Cambridge, UK: Cambridge University Press. View

Spirtes, P., Richardson, T., Meek, C., Scheines, R., & Glymour, C. (1998). Using path diagrams as a structural equation modeling tool. Sociological Methods & Research, 27(2), 182-225.

Suyapa, E., Silva, M., & MacCallum, R. C. (1988). Some factors affecting the success of specification searches in covariance structure modeling. Multivariate Behavioral Research, 23(3), 297-326.

Thompson, B. (2000). Ten commandments of structural equation modeling. In L. Grimm & P. Yarnell (Eds.), Reading and understanding more multivariate statistics (pp. 261-284). Washington, DC: American Psychological Association.

Ullman, J. B. (2001). Structural equation modeling. In B. G. Tabachnick & L. S. Fidell (Eds.), Using Multivariate Statistics (4th ed.) (pp. 653-771). Needham Heights, MA: Allyn & Bacon.

Vermunt, J. K., & Magidson, J. (2005). Structural equation models: Mixture models. In Encyclopedia of statistics in behavioral science (pp. 1922-1927). Chichester, UK: John Wiley & Sons.

Related Pages:

Path Analysis

Conduct and Interpret a Factor Analysis