207P
Comparing Child and Caregiver Assessments of Child Well-Being in a Sample of Maltreated Children
The key aims of public child welfare in the United States are safety, permanency, and child well-being. Child well-being has proven to be challenging to define and measure, which has impeded service accountability and limited progress in prioritizing well-being outcomes. Federal efforts have promoted child well-being as a significant area of focus; largely due to research that has identified the impact of maltreatment on child socio-emotional development. As measurement strategies are considered, an important decision is whether to collect this type of information from children, from their caregivers, or both. However, research on the congruency between child and caregiver reports on child well-being is limited within child welfare populations. Further, the broader social indicators movement highlights the critical importance of capturing a child’s subjective perception of his or her own well-being. The purpose of this research was to assess agreement between caregiver proxy report and child self-report on a widely-used measure of pediatric health-related quality of life, as an indicator of child well-being.
Methods:
Caregivers and their children aged 5-11 years (N dyads=128) responded to the Pediatric Quality of Life Inventory (PedsQL 4.0) as part of a larger study of families with open child welfare cases. The 23 individual items on the PedsQL 4.0 are used to create a total score and domain scores for physical and psychosocial health including subscales for emotional, social, and school functioning. Differences between child and caregiver were assessed by comparing mean scores using paired t-tests. Reliability between raters was assessed using several correlational analyses (weighted Kappa, intraclass correlation [ICC], and Cronbach’s alpha).
Results:
Results indicate poor agreement overall between child and caregiver. Based on child self-report, 57% of children were in the clinical range for total health-related quality of life compared to 19% based on parent proxy report. Child self-report total score (M = 68.0, SD = 16.0) was significantly lower (t = -5.6, p < .001) than the caregiver total score (M = 76.7, SD = 13.1). The largest difference in mean scores was for the emotional functioning subscale. Analyses assessing reliability indicated discordance between child and caregiver ratings. Weighted Kappas for individual items ranged from .03 to .29 and ICCs ranged from .09 to .26 for total and domain scores.
Conclusions and Implications:
This study found that ratings of child well-being may differ substantially between child and caregiver. Assessment of child well-being should include the voice of the child to capture subjective well-being. In this sample, children self-reported much lower well-being on average for both physical and psychosocial health compared to caregivers’ report. Discordance is likely impacted by differences in the way adults and children cognitively process items and the unobservable nature of certain domains of child well-being, but may also reflect a disconnected caregiver-child relationship and caregiver social desirability bias. Additional work is needed on measurement, but collecting information from the child’s perspective, particularly among those who have been victimized, is critical to monitoring well-being in child welfare settings.