Print Print

Introduction to Structural Equation Modeling (SEM)

Faculty
PhD, Post-Doctorial Research Fellow Sabina Nielsen (Department of International Economics and Management, CBS), Associate Professor Bo Bernhard Nielsen (Center for Strategy and Globalisation, CBS) and Dr. G. Thomas M. Hult (The Eli Broad Graduate School of Management, Michigan State University)
Course Coordinator
Bo Bernhard Nielsen
Prerequisite
PhD enrolment – open to all faculty
Prerequisite/progression of the course
The course requires a basic understanding of statistics, but does not require knowledge of structural equation modeling or any particular SEM software. We will provide an overview of different SEM software packages and depending on the preferences of participants, student (demo) versions will be made available for downloading and used during lab sessions.
Aim of the course
This course is designed to help participants understand the theoretical basis and practical application of latent variable structural equation models. More specifically:
  1. What is latent variable modeling and what types of research questions can it help answer?
  2. Theoretical understanding of measurement models and structural models
  3. An overview of state-of-the-art SEM software packages and their relative strengths and limitations
  4. Conducting basic SEM analysis
  5. Reading SEM output and providing guidelines for reporting SEM results
  6. Providing an overview of more advanced analyses.
Course content, structure and teaching
09.10.2008: Foundations (13:00-18:00)
  • Welcome: purpose of the course
  • Introduction to latent variables and their use in social science research
  • Steps of SEM – an overview
  • Measurement models and confirmatory factor analysis (CFA)
    • Factors and indicators
    • Identification of CFA models
    • Interpretation of estimates
    • Model evaluation and modification
10.10.2008: Structural models (09:30-16:00)
  • Structural modeling
    • Casual models
    • Interpretation of estimates
    • Model evaluation
    • Equivalent models
  • Overview of SEM softwares (LISREL, EQS, Mplus, AMOS)
  • Reporting SEM results in article form.
11.10.2008: Practical application (lab session) (09:30-15:00)
  • Conducting SEM analysis with your preferred software
    • Measurement modeling
    • Structural modeling
    • Equivalent models
  • Overview of advanced SEM techniques
    • Multigroup (sample) analysis
    • Latent growth modeling (testing for change over time)
    • Interaction and curvilinear effects
    • Multi-level SEM.
Teaching methods
Lectures and interactive lab-sessions.
Course literature
Main texts:
  • Rex B. Kline: Principles and practices of structural equation modeling, 2nd edition 2005: New York: Guilford Press
  • Research articles TBA.
Choose one of the following:
  • Barbara M. Byrne: Structural equation modeling with EQS: Basic concepts, applications, and programming, 2nd edition 2006: New Jersey: Lawrence Erlbaum Associates
  • Barbara M. Byrne: Structural equation modeling with Lisrel, Prelis, and Simplis: Basic concepts, applications, and programming, 1998: Multivariate Application Book Series: New Jersey: Lawrence Erlbaum Associates
  • Barbara M. Byrne: Structural equation modeling with AMOS: Basic concepts, applications, and programming, 2006: New Jersey: Lawrence Erlbaum Associates.
Enrolment
Frederik Hoff fh.smg@cbs.dk, Bo Bernhard Nielsen bn.smg@cbs.dk

Sidst opdateret af Frederik Hoff 04.03.2009