Latent Class Analysis With Distal Outcomes: A Flexible Model-Based Approach

By Stephanie T. Lanza, Xianming Tan, B. Bray. Structural equation modeling : a multidisciplinary journal, 2013

697 citations recorded by the source corpus.

Although prediction of class membership from observed variables in latent class analysis is well understood, predicting an observed distal outcome from latent class membership is more complicated. A flexible model-based approach is proposed to empirically derive and summarize the class-dependent density functions of distal outcomes with categorical, continuous, or count distributions. A Monte Carlo simulation study is conducted to compare the performance of the new technique to 2 commonly used classify-analyze techniques: maximum-probability assignment and multiple pseudoclass draws. Simulation results show that the model-based approach produces substantially less biased estimates of the effect compared to either classify-analyze technique, particularly when the association between the latent class variable and the distal outcome is strong. In addition, we show that only the model-based approach is consistent. The approach is demonstrated empirically: latent classes of adolescent depression are used to predict smoking, grades, and delinquency. SAS syntax for implementing this approach using PROC LCA and a corresponding macro are provided.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/ec03e92cff09c119c37dd1b0d5a39c761c68bdcf

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

A Comparison of Three-Step Approaches for Auxiliary Variables in Latent Class and Latent Profile Analysis (2017, methodology): “In the next section, the PC method, Vermunt’s (2010) method, and the Lanza, Tan, and Bray (2013) method are presented.”
Robustness of Stepwise Latent Class Modeling With Continuous Distal Outcomes (2016, methodology): “It should, however, be mentioned that it can also be used with distal outcomes that are not continuous (Lanza et al, 2013).”
Latent profile analysis: A review and “how to” guide of its application within vocational behavior research (2020, background): “Such research could examine if the substantive nature of profiles changes across time, and apply latent transition analysis to examine how individuals transition from one profile to another over time, and which factors might explain such shifts (Collins & Lanza, 2013).”
Patterns of psychological adaptation to divorce after a long-term marriage (2015, methodology): “Following Asparouhov and Muthen (2013) and Lanza, Tan, and Bray (2013), a two-step procedure was used to investigate individual differences in reacting to marital breakup.”
Mixture Modeling for Lifespan Developmental Research (2019, methodology): “…approach shows that they all tend to perform well, although they all present limitations in specific conditions (Asparouhov & Muthén, 2014; Bakk, Oberski, & Vermunt, 2016; Bakk & Vermunt, 2016; Bakk et al., 2013; Collier & Leite, 2017; Lanza et al., 2013; McLarnon & O’Neill, 2018; Vermunt, 2010).”
Prediction from Latent Classes: A Demonstration of Different Approaches to Include Distal Outcomes in Mixture Models (2019, methodology): “…distribution of the distal outcome, and the estimated marginal distribution of the latent class variable, this approach computes the conditional distribution of the distal outcome given latent class membership, usually in the form of class-specific means of the distal outcomes (Lanza et al., 2013).”
Fuzzy Profiles (2018, methodology): “In conducting our analyses, we followed the previously described automatic three-step approach for LPA (e.g., Asparouhov & Muthén, 2014; Lanza et al., 2013; Vermunt, 2010).”
Technical Guide for Latent GOLD 5.1: Basic, Advanced, and Syntax 1 (2016, methodology): “Lenk and DeSarbo (2000) proposed using random effects in FMs of generalized linear models, Böckenholt (2001) proposed using random effects in LC Regression models for ranking data, and Muthén (2004) and Vermunt (2007) proposed including random effects in LC growth models.”
Dual commitment to organization and supervisor: A person-centered approach (2015, methodology): “The retained latent profiles were then contrasted on the basis of the distal outcome (turnover) using a method recently proposed by Lanza, Tan, and Bray (2013) and implemented in Mplus through the Auxiliary (DCAT) function (Asparouhov & Muthén, 2013).”