Longitudinal Structural Equation Modeling

Longitudinal structural equation modeling offers a flexible and powerful framework for interrogating longitudinal data. This course emphasizes the importance of harmonizing theory, data collection, and statistical modeling to get longitudinal data analysis right in a way that is accessible to applied researchers.

Instructor:
Aidan Wright, PhD (University of Michigan)

Workshop Dates and Times:
Tuesday, May 28th, 10:00am – 3:30pm ET
Wednesday, May 29th, 10:00am – 3:30pm ET
Thursday, May 30th, 10:00am – 3:30pm ET

Workshop Format:
Three-Day Synchronous Online Workshop

Behavioral and health sciences journals are increasingly expecting longitudinal research designs because they allow investigators to model processes that (ideally) map on to mechanisms. However, to appropriately model longitudinal data requires additional considerations beyond cross-sectional techniques. This 3-day workshop will offer a focused treatment of structural equation modeling (SEM) of longitudinal data, with an emphasis on the applied use of these techniques. Longitudinal data offer insights into compelling questions related to stability, change, development, and timing of events. SEM offers one of the most flexible and powerful ways to study longitudinal questions. This workshop starts with a conceptual introduction to the five major types of stability/change over time (structural, normative, individual, relative, and ipsative), four of which attendees will learn how to model using SEM. Specific techniques that will be covered include longitudinal measurement invariance, autoregressive and latent difference score models, linear and non-linear latent growth curve models, advanced growth curve models, cross-lagged panel models, random-intercept cross-lagged panel models, and latent curve models with structured residuals. These represent the current state of the field’s most useful and popular modeling tools. Longitudinal SEM is ideal for panel data (2 to about 8 or so waves).

What you’ll learn

  • Harmonization – Understand how to match theoretical model of change with the correct statistical model.

  • Precision – Isolate individual trajectories of change from the normative trend with latent growth curves.

  • Interpretation – Learn the correct way do interpret longitudinal models and avoid common misinterpretations in longitudinal SEMs.

  • Unlock new insights – Develop the skills to estimate models that reveal different perspectives on stability and change.

Syllabus

Day 1 – Linear Change

  1. Perspectives on Stability and Change

  2. A (very brief) review of SEM basics

  3. Longitudinal measurement invariance

  4. Univariate auto-regressive models (e.g., latent difference score models)

  5. Univariate linear latent growth curve models (LGMs)

Day 2 – Non-linear Change

  1. Modeling of non-linear patterns of change and growth in LGMs

  2. Between-person predictors and outcomes of growth in LGMs

  3. Time-varying or within-person predictors in LGMs

Day 3 – Multivariate Change

  1. Modeling multivariate change over time using parallel process LGMs

  2. Cross-lagged panel models (CLPMs)

  3. Random-intercept cross-lagged panel models (RI-CLPMs)

  4. LGMs with structured residuals

Registration Options

Longitudinal Structural Equation Modeling

  • Professional
  • $824
  • Baseline Price for Faculty,
    Staff, and Other Professionals
  • Click Register Below
  • Trainee
  • $824 $549
  • 33% Discount for
    Students and Postdocs
  • Use code "TRAINEE" at Checkout

Combo 1: Longitudinal Structural Equation Modeling + Introduction to Structural Equation Modeling

  • Professional
  • $1274 $1019
  • Baseline Price for Faculty,
    Staff, and Other Professionals
  • 20% Combination Discount
  • Click Register Below
  • Trainee
  • $1019 $679
  • 33% Discount for
    Students and Postdocs
  • 20% Combination Discount
  • Use code "TRAINEE" at Checkout

 FAQs

  • Investigators interested in estimating sophisticated models of change in multi-wave data.

  • Intermediate to Advanced

  • Prior knowledge of or experience with structural equation models is necessary and expected. Learners should know the basics of structural equation modeling specification, estimation, and interpretation.

  • Mplus and the lavaan package in R

  • Participants will be provided with annotated example code, practice data, and example write-up.