Introduction to Measurement Invariance

and Item Response Theory

Measurement invariance and item response theory (IRT) are a suite of methodological approaches to comparing the generalizability assessment tools and measures across time points, samples, and populations. This workshop is designed to teach researchers the fundamentals undergirding invariance and IRT approaches and their applications.

Register
Register

Instructor:
Craig Rodriguez-Seijas, Ph.D. (University of Michigan)

Workshop Format:
Two-Day Live Online Workshop

Workshop Dates and Times:
June 26–27, 2026
(9:30am to 4:30pm ET)

Video Availability:
New Workshop for 2026!
New Videos Available on June 28, 2026

Social and behavioral theories of attitudes, abilities, and psychopathology—and the measures created to index relevant constructs—are often assumed universal and generalizable across samples and populations. However, increasing attention is being paid to empirically testing this assumption. There are several approaches to examining potential measurement problems that arise in test development, equating test scores across groups, and identifying psychometric bias. Measurement invariance and item response theory (IRT) approaches are a suit of powerful statistical tools that can be used to examine and compare psychometric properties on social and behavioral assessment tools. These approaches may be used as steppingstones to support other substantive questions (e.g., the assumption of temporal measurement invariance undergirding latent growth models). They may also be a study’s main focal point (e.g., using IRT to examine psychometric bias of any specific instrument among different populations, or to qualify the trustworthiness of group score differences). There are multiple invariance and IRT models with differing assumptions and uses. For instance, we will cover IRT using both binary—appropriate for diagnostic categories—and polytomous (i.e., graded-response) indicators.

This 2-day course will introduce the fundamentals of measurement invariance and IRT theory and their various models. Published research across various social and behavioral research domains will be presented as examples and illustrations of different approaches to invariance (e.g., multi-group CFA models vs. Multiple Indicator, Multiple Cause models) and IRT (e.g., 1-PL vs. 2-PL models) analyses. In addition, example code and data will be provided throughout the workshop alongside the lecture for participants to gain hands-on experience with applying these models. Participants are also welcome to use their own data as well.

What you’ll learn

  • Fundamentals – Understand fundamentals of classical test theory vs. item response theory, levels of measurement invariance.

  • Flexibility – Invariance and IRT approaches can be applied in multiple settings and frameworks. Learn about how you might apply invariance and IRT approaches depending on your data characteristics.

  • Advances in Invariance and IRT – Learn about more novel approaches to invariance and IRT and their uses (e.g., working with smaller samples, continuous covariates).

  • Coding – We will conduct some analyses using example data. Participants are welcome to use their own data as well.

Syllabus

Day 1 – Measurement Invariance

1.     Measurement Invariance & Differential Item Functioning

2.     Multi-Group CFA Invariance Approach

3.     Multiple Indicators, Multiple Causes (MIMIC) Approach

4.     Moderated Nonlinear Factor Analysis (MNLFA) Approach

5.     Recent critiques of Measurement Invariance

Day 2 – Item Response Theory  

1.     Classical Test vs. Item Response Theory

2.     Rasch Model (1-PL Model)

3.     2-Parameter Logistic (2-PL) Model

4.     Extensions: 3-PL & Multivariate IRT

5.     Study Design Considerations (e.g., IRT with smaller sample sizes)

6.     Applications (e.g., shortening/expanding measures, detecting psychometric bias)

Registration Options

Introduction to Measurement Invariance and Item Response Theory

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

 FAQs

  • This workshop is specially designed for researchers in the behavioral and health sciences who are interested in learning strategies for comparing psychometric properties of measures across groups, conditions, time-points, etc.  

  • Intermediate

  • Though we will go over some basics for latent variable modeling, learners should already have some understanding of factor analytic approaches. Advanced experience is not necessary.

  • Code will be provided in R. Some examples of previous analyses might be through MPlus.

  • Slides, code, and appropriate readings will be provided.