Validation of the Kidsights Measurement Tool: A parent-reported instrument to track children's development at the population level

Abstract

Disparities in child development between groups of children arise early and reflect social inequities in early environments, geography, and other factors. To track and address these disparities, valid and reliable tools are needed that can be implemented at-scale and across populations. However, no population-level measures of child development appropriate for children from birth to age five years have been developed and validated in the United States to date. The Kidsights Measurement Tool (KMT) is a parent-report, population-level tool for children birth to age five years intended to track group-level differences in the developmental status across normative aspects of children's motor, cognitive, language, and social/emotional development. This study reports on validation of KMT as a feasible tool that can be implemented in large-scale surveys to track disparities in early childhood development. Using a sample of N = 5,001 initial parent reports residing in Nebraska and across the United States, we find strong evidence that the KMT can detect disparities in child development birth to age five, as indicated by expected criterion associations with parent education and mental health, as well as child's race and ethnicity. In addition, we found that the KMT is strongly associated with gold-standard direct observation instruments (i.e., the Bayley Scale of Infant Development and the Woodcock-Johnson) that measure similar developmental constructs both concurrently and one-year 12-24 months later. Finally, the KMT exhibits strong reliability even after controlling for age, and we find no evidence that measurement noninvariance threatens valid inferences about group difference. Taken together, our findings indicate that a parent-report measure can generate valid and useful estimates for tracking disparities in early child development at the population level.

Fig 1. Item composition.

Kidsights Measurement Tool domain representation by children’s age. Notes: Cog. – Cognition development, Lang. – Language development, Phys. – Physical development. Socemot. – Socioemotional development.

Fig 2. Reliability.

Marginal reliability ($r_{X{X}^{\prime }}$of scores at child’s age (N = 5,001).