An increased global interest in early childhood development in the past decade has led to a growing demand from program implementers and researchers for simpler measures of child development which can be used at scale. These simpler measures have the potential to be used in planning and evaluating programs or interventions aimed at improving the health and development of infants and young children throughout the world. A working meeting with the focus on identifying indicators that could be used within and across populations (countries) to estimate the proportion of young children under the age of three, who reach their developmental potential, was hosted by the GCDG from 14th- 16th April, 2014 at the Jamaica Pegasus Hotel, Kingston, Jamaica. The meeting was attended by a group of investigators from across the globe, with cohorts with the necessary data for identifying indicators for ECD.
Recognizing that the origins of adult health and well-being are formed very early in life, there is a desire to measure early child development and to incorporate indicators of early child development into national surveys (e.g., MICS, DHS) to be used to assess need for investment, to measure progress, and to measure impact of disasters. Population- based indicators are not used to measure individual children. In 2007, based on stunting and poverty, as indicators of child development, the Group estimated that over 200 million children under age 5 in low and middle income countries do not reach their developmental potential, thus increasing their risk for school readiness problems and low performance that could last a lifetime, contributing to ongoing disparities. However, relying exclusively on stunting and poverty as indicators of child development yields a very low estimate that does not account for other contributing factors, including infections, maternal mental health, maternal education, exposure to violence, access to enriching opportunities for learning, etc. The goal of the meeting was to use longitudinal data from existing cohorts from low and middle income countries, to examine the validity of early child development indicators and to identify whether items or combinations of items from developmental assessments obtained from 0-3 years are adequately robust predictors of later cognitive and behavioural measures.
The outcomes from the meeting resulted in adopting a novel two-stage estimation procedure, based on the Rasch model and the calculation of change scores that can be used to establish “growth charts” of early child development. Stef van Buuren, Senior statistician at the Netherlands Organization for Applied Scientific Research (TNO) in Leiden, showed that data from standardized measures of child development (0-2 years) form a continuous latent variable (D-score) that has interval scale properties. If this observation holds across samples and settings, the resulting trajectory of D-scores can be used to calculate differences within and across ages and countries, much as height for age growth charts are used to indicate rates of stunting. It was decided to apply the methods pioneered by van Buuren to the 13 cohorts identified, which included more than 12,000 children from low and middle income countries from around the world and to evaluate the use of trajectory D-scores to identify global indicators for early child development.
This data-driven approach is being adopted by the Group to determine which items could be used to create simpler measures which could be used at scale to monitor impact of programs and progress in ECD at national and sub-national levels. The goal is to identify the fewest number of indicators that meet the criteria. Considering that there are a number of efforts currently underway to develop child development indicators by organizations including the World Health Organization, Save the Children, UNICEF and USAID, the GCDG plans to collaborate with these groups, moving forward.