Common patterns of maltreatment exposure: A replication study using confirmatory latent class analysis

Abstract

Background

Child maltreatment (CM) is a multidimensional construct that requires complex multivariate approaches to properly operationalize the heterogeneity inherent in the exposure. Person-centered approaches, such as Latent Class Analysis (LCA), have emerged as promising methods, with a proliferation of studies in the past decade applying LCA to capture patterns of CM exposure across multiple dimensions. This line of research has been highly generative and also led to disparate findings about the conclusions regarding typical patterning of CM across studies, potentially due to differences in samples, measurements, and indicator selection. Moreover, it is yet unclear whether latent classes are a product of specific samples, or if they can be reproduced across studies that use similar methods.

Objective

The current study addresses this gap by replicating a previous LCA of CM exposure using Confirmatory LCA (CLCA)—a first in the literature.

Participants

The sample comprises a racially-ethnically diverse cohort of 566 low-income children aged 8–10 (Mean = 9.42, SD = 0.88). Half of the cohort was recruited based on documented histories of CM exposure and the other half contained demographically-matched, nonmaltreated children.

Results

Results provide evidence for the replication of four unique latent classes of CM exposure: a) nonmaltreated, b) Episodic, Single-Type (Neglect), c) Episodic, Single-Type (Abuse), d) Chronic, Multi-Type.

Conclusion

Reproducing latent classes of CM exposure in different samples of CM survivors contributes to the robustness of these specific CM LCAs and serves as an important step toward addressing replication concerns.