An Ecological Typology of Child Maltreatment
This paper addresses the following research question: do psychological and sociological factors associated with child maltreatment risk occur in patterns that differentially affect child well-being and predict protective services (CPS) investigation? This question is important because information gained from such a study can be used to theorize testable causal pathways for understanding the etiology of maltreatment and in turn lead to differential intervention models. Despite a call for ecological paradigms assuming continuous interaction of risk and resilience factors (Belsky, 1993), etiological studies have not taken into account the scope to which these factors are interrelated. This paper presents a novel application of pattern analysis to typologizing child maltreatment, and in so doing presents the first empirical study of the relationship of both the individual and his/her context for analysis.
Methods:
This study advances a meta-analysis of maltreatment risk factors (Stith et al, 2009) by using latent profile analysis (LPA; Gibson, 1959) to study their interrelatedness in 1,354 high-risk families in the Longitudinal Studies of Child Abuse and Neglect. Data were gathered on parenting attitudes, discipline, depression, alcohol abuse, social supports, ambient stressors, family cohesion, neighborhood quality, and prior CPS involvement at age 6 and analyzed using MPlus. Logistic regression was used to predict CPS investigation for each profile and OLS regression assessed differences in child well-being by profile membership.
Results:
A six-profile solution emerged consisting of four risk profiles and two other profiles. Profile #1 was defined solely by alcohol use. Profile #2 consisted of poor parenting attitudes, high levels of depression, and use of both psychological and physical punishment. Profile #3 was marked by minimal social support, poor neighborhood quality, moderate depression, and high personal stress. Profile #4 was distinguished by moderate family cohesion and social supports but parents strongly supported use of physical discipline and lacked both warmth and developmental understanding. Of the other two profiles, one was completely neutral and the other indicated positive parenting attitudes.
Membership in Profiles 1-3 predicted CPS investigation but not Profile #4. Living in poverty increased the likelihood of investigation. Children in Profiles #2 and #4 were likely to exhibit aggression. Girls in Profile #1 were more likely to be withdrawn, while boys were more likely to be aggressive.
Conclusion:
Derived from the first application of LPA to understanding the ecological context of child maltreatment, these results indicate that child maltreatment risk factors tend to organize in select patterns with strong implications for practice and research. Concerning future practice, these findings suggest different starting points for assessment and intervention for both caregivers and children. Assessments need to be more differential in nature not only to address adequately the type of risk, but to inform differential intervention to both parents and children either engaged in or affected by maltreatment. Concerning future research, by identifying how (and which) risk factors co-occur, researchers may assist in the development of more targeted interventions for such parents and allocate already-scarce resources more effectively (Keller, Cusick, & Courtney, 2007)