A Introduction to Conducting Latent Class Analysis

The purpose of this workshop is to provide social work researchers with an introduction to conducting Latent Class Analysis (LCA) with Mplus.

LCA is a statistical modeling procedure used to identify a typology; stated differently, it is used to assess whether subgroups exist within a sample. LCA can inform social work research and practice because it allows for the identification of subgroups from a seemingly homogenous group. For example, researchers can use LCA to assess whether African American males report different patterns of exposure to risk factors linked to delinquency. By identifying potential patterns of risk, researchers can provide practitioners with guidance on which interventions to select for whom. LCA can be used by social work researchers interested in answering questions such as “Are there different patterns of behaviors among individuals in my sample?”, “Do patterns of behavior differ by gender?”, or “Do these patterns explain different outcomes?”

This workshop is targeted towards researchers who have had exposure to introductory statistics. Doctoral students are welcome. No experience to Mplus is necessary. Audience members are encouraged to bring their laptops and dataset. Although brief mention will be given to using LCA with longitudinal data, this workshop will focus primarily on using LCA with cross-sectional data.

By the end of this workshop, participants will have answers to the following questions: (1) When should I use LCA? (2) How important is theory in LCA? (3) How do I use Mplus to conduct LCA? (4) How do I include covariates in my statistical model? (5) How do I select the best class-solution? (6) What are the most common challenges to using Mplus, and how do I overcome these issues? (7) How can I use my results to inform social work practice? (8) Where can I get more help with conducting LCA?