Introduction to Class

Structural equation models for the social sciences, including specification, estimation, and testing. Topics include path analysis, confirmatory factor analysis, linear models with latent variables, MIMIC models, non-recursive models, models for nested data. Emphasizes applications to substantive problems in the social sciences. Prerequisite: SOC 424, SOC 425, SOC 426 or equivalent; recommended: CS&SS 505 and CS&SS 506, or equivalent. Offered: jointly with CS&SS 526.

Class Notes

• Lecture Notes: Introduction to Structural Equation Models
  
• Lecture 1: A Bivariate Structural Equation Model
  
• Lecture 2: Recursive Model in Standardized Form
  
• Lecture 3: Decomposing Effects in Recursive Models
  
• Lecture 3A Instrumenal Variables
  
• Lecture 3B: A Non-Recursive Model
  
• Lecture 4: Causality and Random Measurement Error
  
• Lecture 5: A Walking Dog Model
  
• Lecture 6: The General LISREL Model
  
• Lecture 7: The LISREL 8 and PRELIS 2 Programs
  
• Lecture 8: Simple Rules for Identification
  
• Lecture 9: MIMIC Models
  
• Lecture 10: Confirmatory Factor Models
  
• Lecture 11: Reliability and Validity in Measurement Models
  
• Lecture 12: Maximum Likelihood Estimation
  
• Lecture 13: Test Statistics and Multiple Group Models
  
• Lecture 14 WLS and Models for Ordinal and Limited Variables
  

Files For Download

• Status attainment annotated LISREL output
  
• Walking Dog LISREL Annotated Output
  
• PRELIS annotated output continuous variables
  

Homework Assignments

• Seminar Paper Assignment Format
  

Readings

• Matsueda (2012) Key Advances in the History of Structural Equation Modeling
  
• Duncan, Chapters 1-2
  
• Duncan, Chapters 3-4
  
• Duncan, Chapters 5-7
  
  Back