DAY 1: Structural Equation Modeling

Introduction to Structural Equation Modeling (SEM)

• Welcome
• Things You Should Know Before the analysis
• What is SEM?
• Types of research questions addressed by SEM
• Strengths and limitations of SEM
• Assumptions relating to SEM
• Introduction to factor analysis
• Software programs
• Preparing your data for analysis

Things to consider before trying SEM- The use of SEM can be impacted by:

• The research hypothesis of interest
• Sample size
• Variable distributions
• Outliers
• Missing data
• Complexities of Salivary Research

Terminology and type of variables that occur in SEM,- Intro papers, structure in Statistical Horizons

• Exogenous variables
• Endogenous variables\manifest variables
• Latent variables
• Covariance and correlation
• Path Models
  • Recursive
  • Non-recursive
• Direct and indirect effect
• Structural Model
• Measurement model

DAY 2: Structural Equation Modeling

Steps for developing the Model hypothesis and Model specifications

• Model specification
• Specify a model
• Visual representation of model
• Estimation, fixing, and constraining variables
• Model Identification
• Preparing a statistical program to analyze your model
  • Example syntax
• Model estimation
  • Preliminary descriptive statistical analysis
  • Estimate parameters in the model
• Model testing
  • Assess model fit
    • Comparative fit indices
    • Variance accounting indices
    • Parameter based indices
• Model modification
  • Modifying a model to improve fit
  • Chi-square difference test
  • The Lagrange-multiplier test
  • The Wald test

Interpreting and reporting the results

• How to interpret you model parameters
• Compare output from different statistical packages
• What is important to report in methods and result sections

Advanced Techniques and Best practices

• Mean structure and latent growth models
• Multilevel SEM
• Full Information Maximum Likelihood

DAY 3: Structural Equation Modeling

Review basic syntax for estimating a SEM

• Compare available options between SEM analysis in R, SAS, and Stata

Assistance with preparing data for analysis

Assistance with adapting and applying basic SEM syntax to estimate model using personal data sets

• Trouble shooting errors

Assistance with initial interpretation of output