CAREGIVER MENTALIZING OPERATIONALIZED AS MEDIATIONAL LEARNING AMONG CAREGIVERS OF ORPHANS AND VULNERABLE CHILDREN IN SOUTH AFRICA

Parental mentalizing is an important component of caregiving quality that is thought to facilitate sensitive and responsive caregiving and the development of adaptive social cognition and emotion regulation in children, and ultimately to promote children’s mental health and resilience. However, there is less research about parental mentalizing beyond infancy and early childhood, and a lack of observational measures of parental mentalizing for middle childhood and adolescence. Recently, an observational measure of caregiver behavior called Observing Mediational Interactions (OMI), which assesses caregivers’ mediation of children’s learning during dyadic interactions, has been suggested as a behavioral operationalization of caregiver mentalizing. The present study had an overall aim to evaluate the reliability and validity of the OMI as a behavioral measure of parental mentalizing among caregivers during middle childhood. Reliability and construct validity (convergent and discriminant validity) of the OMI were evaluated in a population of caregivers of orphans and vulnerable children (OVC) in South Africa, a particularly at-risk group for mental health difficulties. Children in this study (N = 88) were 7-11 years old and enrolled in one of four community-based organizations (CBOs); caregivers were their CBO careworkers (N = 18). Participants were assessed at three time points (baseline, 6-month, 12-month) as part of a larger intervention trial. Careworker-child dyads filmed an interaction at each timepoint, which was coded using the OMI. To evaluate convergent validity, careworkers completed self-report measures of parental mentalizing and their responses from an interview were also coded for mentalizing. Children also completed a task to assess social cognition. To evaluate discriminant validity, children, their parents, and careworkers completed a measure of child emotional and behavioral difficulties. Reliability for OMI components (affective components and five mediational components: focusing, affecting, expanding, rewarding, regulating) was assessed using intra-class correlations, and adequate reliability was demonstrated in 12 of 18 OMI components across timepoints. Cross-sectional associations between OMI components and convergent and discriminant validity measures were mostly non-significant. However, significant negative associations between Expanding and Regulating behaviors and measures of parental mentalizing, and significant positive associations between affective components and measures of parental mentalizing, point to directions for future research.Psychology, Department o