Poster #52 - A Person-Centered Approach to Violence Exposure and Behavioral Outcomes Among Preschoolers

• In Event: 2-055 - Poster Session 06
  In Poster Session: PS 06 Section - Developmental Psychopathology

Integrative Statement

Approximately one in three children from low-income families is exposed to violence before the age of 5, putting them at greater risk for maladaptive developmental outcomes (Briggs-Gowan, Ford, Fraleigh, McCarthy, & Carter, 2010). Unfortunately, violent events are rarely experienced in isolation; rather, youth who are exposed to one event are more likely to experience additional exposures (Finklehor, Ormrod, & Turner, 2007). Nonetheless, most studies of early violence examine one specific type of exposure or the sum number of exposures. These variable-centered approaches may not be representative of children’s actual experiences.

Specifically, examinations of outcomes associated with a specific type of violence exposure do not account for the cumulative effects of multiple exposures, whereas studies that utilize sum scores of exposure events implicitly assume that the impacts of different events are comparable. Thus, a person-centered analytic approach that examines the participant rather than the exposure variable as the unit of analysis would complement existing research by identifying constellations of violence exposure that are likely to co-occur and differentially relate to behavioral outcomes. Grasso et al. (2016) conducted an LCA of family violence exposure among preschoolers, yielding three profiles that were differentially related to internalizing and externalizing symptoms. The present study seeks to extend this research by examining community violence in addition to family violence (i.e., partner-directed and child-directed) among 137 low-income, predominantly African American (81%) mothers from an urban southern city and their preschool-aged children (M=51.49 months; SD=9.15; Range=3-5 years).

Mothers completed interviews and questionnaires assessing children’s partner-directed (Conflict Tactics Scales – Revised), child-directed (Conflict Tactics Scale – Parent-Child), and community violence exposure (adapted Exposure to Violence Interview), as well as children’s internalizing and externalizing behavior (Child Behavior Checklist 1.5-5). A 3-step approach identified latent profiles based on dichotomous violence indicators and associations with child outcomes. First, we compared two- and three-class solutions for exploratory LCAs with 16 partner-directed, 8 child-directed, and 5 community violence indicators; fit indices supported a 2-class solution: a polyvictimized class exposed to community and family violence (33.6%) and a child-directed violence class (66.4%). However, an examination of conditional response probabilities revealed that community and child-directed violence did not differentiate between classes (CRP differences > .34). Six partner-directed items also failed to differentiate. Thus, 10 partner-directed violence items were retained in the refined model, which resulted in a 2-class solution: high (32.8%) and low (67.2%) partner-directed violence exposure. Children in the high exposure class exhibited higher externalizing symptoms than low-exposure children, F(1,125)=4.84, p = .030. Further, children in the high exposure class also exhibited higher internalizing symptoms, albeit at trend level, F(1,125)=3.855, p = .052. Finally, classes did not differ on child sex, maternal race or ethnicity, or maternal or child age.

Our results indicate that community and child-directed violence did not provide information above and beyond partner-directed violence about how violence is contextually experienced among preschoolers from low-income, diverse families. Findings highlight the importance of targeted assessment and interventions to decrease exposure to partner-directed violence, particularly when working with children exhibiting externalizing behavior.

Authors

• Erin Glackin, Tulane University
  Presenting Author

• Virginia Hatch, Tulane University
  Non-Presenting Author

• Elsia Obus, Tulane University
  Non-Presenting Author

• Sarah A Gray, Tulane University
  Non-Presenting Author