Validation of the Global Scales of Early Development (GSED) tool in rural Western Kenya

Early childhood development (ECD) is a key determinant of long-term health, education, and wellbeing. However, one major global challenge is the lack of ECD assessment tools validated for use in low- and middle-income countries. To address this gap, the World Health Organization (WHO) launched in 2023 the Global Scales for Early Development, an open-access tool designed to generate reliable, valid, and internationally-comparable data on ECD for children aged 0-3 years globally. In this study, we examined the concurrent and convergent validity of the Global Scales for Early Development-long form (GSED-LF) for use with children aged 0-24 months in Kenya.

We analyzed baseline data collected in October-November 2023 as part of a cluster-randomized controlled trial evaluating a parenting program for improving ECD in rural Western Kenya. Primary caregivers (91% mothers) with a child under 24 months were enrolled across 64 villages in Busia and Homabay counties. The GSED-LF was administered to all children (N = 647). In a randomly selected sub-sample of children (N = 116), the Bayley Scales of Infant and Toddler Development (Bayley-III) and the Caregiver Reported Early Development Instruments (CREDI) were also administered to compare their scores with those from the GSED-LF. Concurrent validity of GSED-LF was assessed in terms of its correlations with Bayley-III and CREDI. Convergent validity of GSED-LF was examined with respect to parenting outcomes, including parental stimulation, home caregiving environment, and maternal mental health.

GSED-LF scores had moderate associations with those on the Bayley and CREDI across the domains of cognitive, language, and motor development. GSED-LF had small associations with socioemotional development and relatively weaker concurrent validity for younger children under 12 months. GSED-LF also demonstrated good convergent validity in terms of showing moderate associations with maternal and paternal stimulation and the home caregiving environment.

Overall, this study demonstrated the feasibility and initial validity of the GSED-LF as a direct assessment tool for use in rural Western Kenya. Additional psychometric analyses across diverse settings are needed to strengthen the reliability and validity evidence of the GSED-LF and establish it as a robust, globallyapplicable tool for assessing ECD in resource-limited settings.