Using Structural Equation Models to Uncover Relationships between Observed and Latent Variables

Presenters: Kemal Oflus

Using Structural Equation Models 

See how to:

  • Understand the basics of SEM
    • Confirmatory Factor Analysis (CFA) – a multivariate statistical procedure to test how well measured variables represent the number of constructs and confirm or reject a theory
    • Path Analysis - an extension of the regression model built from the correlation matrix that compares two or more casual models and uses a diagram flow chart to model path and causation
    • Structural Equation Modeling (SEM) – Multivariate statistical analysis technique that combines factor analysis with multiple regression analysis to analyze structural relationship between measured variables and latent constructs, estimating multiple and interrelated dependence in a single analysis
  • Select models and handle measurement areas using an industrial case study
    • Navigate the 3-panel SEM window (From List, To List, Diagram)
    • Use SEM to see if there is any built-in error in your measurements
    • Understand how to examine and compare restricted and independent model results
    • Build multiple models iteratively and interpret requirement Rule Status before running models
    • Locate and interpret model statistics and Report Warnings, and then modify or rerun models if necessary
    • Use Local Data filter to filter by variable
    • Save factor scores and apply latent variables to observed variables
  • Use SEM to test a hypothesis using a sports case study

Note: Q&A included at time 36:43.

Resources

Back to Top