Symposia
Child / Adolescent - Trauma / Maltreatment
Lindsay R. Druskin, M.S.
West Virginia University
Morgantown, West Virginia
Jane Kohlhoff, Ph.D.
Associate Professor
University of new South Wales Sydney
Villawood, New South Wales, Australia
Christopher Owen, M.S.
Doctoral Candidate
West Virginia University
Morgantown, West Virginia
Robin Han, M.S.Ed.
Doctoral Candidate
West Virginia University
Morgantown, West Virginia
Samantha N. Holbert, None
Research Assistant
West Virginia University
Bridgeport, West Virginia
Cheryl B. McNeil, Ph.D.
Professor
West Virginia University
Morgantown, West Virginia
Child abuse (CA) is a pervasive, global problem impacting millions of children (Stoltenborgh et al., 2015). As children 0 to 2 years are at highest CA risk (National Child Abuse and Neglect Data System; NCANDS; 2019), research on parent risk factors is particularly needed in families with children in this age range. In conducting CA risk assessments, observing parent-child interactions can capture key behaviors that are less susceptible to parent bias and may differ from parent perceptions of their own parenting. Thus, a multi-method approach that combines self-report with observational measures is crucial in ensuring valid assessment of CA risk. We hypothesized that single parent status, parenting stress, and child disruptive behaviors would predict higher CA risk and that observed harsh parenting behaviors would add significant incremental prediction of CA risk when accounting for all other factors.
Participants were 84 mothers and their toddlers (48% male, Mage = 19 months) clinically referred for behavioral problems in Australia. Parenting stress was reported on the Parenting Stress Index (Abidin, 2012), disruptive child behavior frequency and problematic nature on the Eyberg Child Behavior Inventory (ECBI) Intensity and Problem scales, respectively (Eyberg & Pincus, 1999), and CA potential on the Brief Child Abuse Potential Inventory (BCAP; Ondersma et al., 2005). Observed parent negative talk (NTA) was measured with the Dyadic Parent-Child Interaction Coding System (Eyberg et al., 2009) during 20-minute parent-child interactions.
A hierarchical multiple regression technique was used to assess the relative impact of Marital Status, Parent Stress, ECBI Intensity, ECBI Problem, and NTA (entered last) on the BCAP. When NTA was included in the regression model, the value for R Square increased to .640. The addition of NTA significantly (p = .039) contributed 3.5% additional variance in BCAP abuse risk accounted for by only Marital Status, Parent Stress, ECBI Intensity, and ECBI Problem.
Findings highlight the importance of observing parent-child interactions, as observed negative parenting predicted CA risk on top of the variance explained by several known risk factors. It is critical to include observations as part of the standard assessment battery for parents at risk for CA, as they provide unique information above and beyond questionnaires that identify a parent in need of critical intervention to ensure the safety and wellbeing of their child.