The development and reliability of an observational tool for assessing mother-child interactions in field studies- experience from Pakistan.

Abstract

This study describes the development and reliability testing of an observational tool to measure mother-child interactions with toddlers aged 2 years in a rural low-income country setting. The development protocol comprised five phases with iterative revisions: (1) identification of the theoretical framework for responsive behaviours and selection of items; (2) field testing; (3) expert review; (4) training of the data collection team; and (5) piloting. The final tool was a structured live observational measure assessing a 5-min interaction of a shared picture-book-reading activity. Maternal behaviours assessed included affect, touch, verbal statements and language stimulation; child behaviours assessed included affect, communication and attention. Following development, the mother-child interaction tool was administered on a cohort of 1390 children at 2 years of age. Using a video strategy, inter-observer reliability assessed by the Bland-Altman test for mother-child dyads suggested moderate agreement between expert and field assessors on total scores (r = 0.681**, P < 0.001, n = 154). Significant associations of the total interaction score correlations using Pearson's' correlations were found with the Responsiveness (r = 0.271**, P < 0.001, n = 1345) and Involvement (r = 0.325**, P < 0.001, n = 1345) subscales of the Home Observation for Measurement of Environment-Infant Toddler Inventory, maternal knowledge (r = 0.203**, P < 0.001, n = 1345), maternal depression (r = .-063**, P < 0.001, n = 1345), child cognitive development (r = 0.392**, P < 0.001, n = 1345) and language development (r = 0.620**, P < 0.001, n = 1345) assessed using the Bayley Scales for Infant and Toddler Development, Third Edition. The authors conclude that this tool can be reliably used by trained assessors to measure mother-child interactions in field studies.