Date of Award

9-6-2024

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Psychology

First Advisor

Scott Leon

Abstract

The benefits of psychological assessment are particularly salient for children and youth in the child welfare system, as youth in contact with the system are more likely to experience mental health symptoms, disorders, and comorbidities as compared to youth in the general population. These assessments often utilize multiple reporters, as emerging evidence indicates that utilizing multiple more effectively captures the mental health experiences of children and youth. However, the extant literature has consistently observed only low-to-moderate (i.e., .2-to-.3) agreement across informants, including among dyads in contact with the child welfare system. Several factors are associated with level of agreement across reporters, including race, gender, age, behaviors and symptoms of interest, and relationship of the respondent to the child. Specific factors associated with child welfare involvement also contribute to discrepancies across reports, including maltreatment history and foster caregiver/biological parent status. The current study seeks to expand on this research about multi-informant reporter discrepancies by examining reporting discrepancies among caregivers, teachers, and 12-year-old adolescent dyads with prior and/or ongoing child welfare involvement, as well as capturing how teacher-caregiver discrepancies change over time via growth mixture modeling and how trajectory predicts child-rated mental health at age 18. In keeping with the extant literature, low-to-moderate discrepancies were observed across caregiver-child, child-teacher, and teacher-caregiver dyads regardless of symptom clusters. Discrepancies did not differ according to race or gender. Next, growth mixture modeling indicated that the best fit for internalizing symptoms was a three-class solution; trajectories included caregiver-teacher dyads with 1) consistently low levels of disagreement, 2) consistently high levels of disagreement, and 3) dyads with high to moderate decreasing disagreement. Regarding externalizing behaviors, the three-class solution was also most appropriate and included caregiver-teacher dyads with 1) consistently low levels of disagreement, 2) low to moderate increasing disagreement, and 3) moderate to low decreasing disagreement. Finally, in terms of total symptoms, the four-class solution was most appropriate and included caregiver-teacher dyads with 1) moderate to high increasing disagreement, 2) low to moderate increasing disagreement, 3) consistently low disagreement, and 3) high to moderate decreasing disagreement, though these trajectories differed depending on whether caregiver or teacher reported higher symptoms. Notably, only trajectories with increasing levels of disagreement predicted child-rated mental health outcomes at 18 when controlling for baseline mental health symptoms. Specifically, when trajectories including increasing discrepancies ending in the moderate range, children reported higher mental health symptomology at age 18; when trajectories including increasing disagreement ending in the high range, children reported significantly lower mental health symptoms at 18. The implications of these findings for research and clinical use are discussed.

Download

Share

COinS