Identifying Patterns of Victimization Among Child Welfare Involved Youth: A Latent Class Analysis

BACKGROUND AND PURPOSE: Finkelhor et al. (2007a) coined the term polyvictimization to describe exposure to multiple forms of violence, crime, and abuse. Using a nationally representative sample, Finkelhor et al. (2007b) found that 22% of participants were polyvictims, and the polyvictims experienced 57% of the victimizations reported. Polyvictims reported experiencing the most serious forms of victimization (e.g., sexual victimization) and they manifested symptoms of depression, anxiety, and anger/aggression. Polyvictimization is a more powerful predictor of depression, anger/aggression (Finkelhor et al, 2007), and delinquency than any single type of victimization (Ford et al., 2010). Polyvictimization is pronounced among homeless and runaway youth, who experience additional victimizations on the street (Bender et al., 2014). Bender et al. (2014) found that 93% of study participants experienced at least one form of maltreatment before leaving home, and 30% reported experiencing physical, emotional, and sexual abuse before leaving home.  Participants who reported three forms of abuse had more than double the odds of post-traumatic stress disorder and increased odds of depression. The purpose of this study was to investigate the co-occurrence of victimization among child welfare involved-youth based on caseworker records. The study also examined characteristics of the latent classes to see what differences exist.

METHODS: The study used a subsample of 1,054 youths (ages 11-17) from the second National Survey of Child and Adolescent Well-Being (NSCAW II) who were investigated for maltreatment between February 2008 and April 2009.  The subset of youth (mean age = 13.65) includes: 45% boys; 38% White, 27% Black, 24% Hispanic and 10% other races. Latent class analysis (LCA) was used to investigate patterns of victimization during one point in time. LCA is a person-centered analytic approach that combs through a dataset to locate and group together individuals with similar patterns of survey question responses. LCA was conducted in MPLUS 7.1 and post-hoc analyses were conducted in Stata 13.1.

RESULTS: Five latent classes of victimization were identified: (1) neglect, which included general and physical neglect, and substance abusing parents (47.64%); (2) sexual abuse, which also included low rates of general neglect (11.32%); (3) physical abuse (20.31%); (4) emotional and physical abuse, which also included low rates of general neglect and sexual abuse; (9.69%) and (5) other abuse, which also included some general neglect (11.04%). The sexual abuse class included a greater proportion of females than the other classes (81.08%, p<.01). Runaway youth were more likely to be in the physical abuse class than the neglect class (21.14% vs. 10.13%, p<.05), but differences were not found across other classes. No differences were found across groups for depression.

IMPLICATIONS FOR PRACTICE: Based on previous literature we expected a group with four or more forms of victimization to emerge. Thus, results raise questions about whether caseworkers’ files adequately and comprehensively capture the victimization experiences of child welfare involved-youth beyond the initial report. If caseworkers fail to regularly ask about (and clearly document) victimization experiences among youth, there will be missed opportunities to initiate timely and effective trauma-specific interventions to improve child well being.