“You are our eyes and ears”: A new tool for observing parent-child interactions in large samples

Abstract

Differences in parent-child interactions have implications for a range of developmental outcomes, yet it has traditionally been difficult for large cohort studies that traverse multiple domains to include such detailed behavioural observations. We describe a new method for observing parent-child interactions specifically designed to be a component of a more comprehensive collection of data about child health and development. Participants were mothers and their two-year-old children who were part of the Growing Up in New Zealand study. During a series of brief, structured parent-child conversation tasks, observers were trained to rate mothers’ warmth, use of open-ended questions, talk about emotions and ‘linking’ talk, children’s emotional expression and mothers’ overall use of discipline. Reliability was established before and reviewed mid-way through the one-year data collection wave. We observed differences in parent-child interaction construct ratings as a function of socio-demographic variables, ethnicity and child gender that were in agreement with published research. Inter-scale correlations and correlations between observer ratings and maternal self-report measures provide preliminary evidence of convergent and discriminant validity. Specifically, higher maternal self-reported affiliation and more frequent book reading were significantly correlated with observer ratings of maternal warmth, maternal language style and children’s emotional expression, and negatively correlated with observer ratings of maternal discipline. Higher maternal self-reported parenting hostility was negatively correlated with observed maternal warmth and language and positively correlated with observed maternal discipline. This observational method is a potentially useful technique for obtaining independent measures of parent-child interactions during the preschool years within large cohort studies.