Protocol for the Adaptation of a Direct Observational Measure of Parent-Child Interaction for Use With 7–8-Year-Old Children

Abstract

ObjectiveParenting sensitivity and mutual parent-child attunement are key features of environments that support children’s learning and development. To-date, observational measures of these constructs have focused on children aged 2–6 years and are less relevant to the more sophisticated developmental skills of children aged 7–8 years, despite parenting being equally important at these ages. We undertook a rigorous process to adapt an existing observational measure for 7–8-year-old children and their parents. This paper aimed to: (i) describe a protocol for adapting an existing framework for rating parent-child interactions, (ii) determine variations in parents’ sensitive responding and parent-child mutual attunement (‘positive mutuality’) by family demographics, and (iii) evaluate the psychometric properties of the newly developed measure (i.e., inter-rater reliability, construct validity).MethodParent-child dyads completed one home visit, including a free-play observation and parent questionnaire. Dyads were provided with three toy sets: LEGO® Classic Box, Classic Jenga®, and animal cards. The Coding of Attachment-Related Parenting (CARP) was adapted for use with 7–8-year-old children, and rating procedures were streamlined for reliable use by non-clinician/student raters, producing the SCARP:7–8 Years. Trained staff rated video-recorded observations on 11 behaviors across two domains (five for parents’ sensitive responding, six for parent-child positive mutuality).ResultsData were available for 596 dyads. Consistently strong inter-rater agreement on the 11 observed behaviors was achieved across the 10-week rating period (average: 87.6%, range: 71.7% to 96.7%). Average ICCs were 0.77 for sensitive responding and 0.84 for positive mutuality. These domains were found to be related but distinct constructs (r = 0.49, p < 0.001). For both domains, average ratings were strongly associated with the main toy used during the observation (p < 0.001, highest: cards, lowest: LEGO®). Adjusted multivariate linear regression models (accounting for toy choice) revealed that less sensitive responding was associated with younger parent (p = 0.04), male parent (p = 0.03), non-English speaking background (p = 0.04), and greater neighborhood disadvantage (p = 0.02). Construct validity was demonstrated using six parent-reported psychosocial and parenting measures.ConclusionThe SCARP: 7–8 Years shows promise as a reliable and valid measure of parent-child interaction in the early school years. Toy selection for direct observation should be considered carefully in research and practice settings.