Development of a screening tool to identify babies at risk of language delay in India: A preliminary study

ABSTRACT Early identification of language delay is important as it has a serious impact on a child’s life in terms of educational, social, and emotional development. Among the early language screening tools, there are some parent-administered tools; however, they are not culturally appropriate or freely available. This article documents the development and preliminary validation of a quick and easy-to-administer language screening tool for babies from 6 to 18 months of age. Parents of 100 babies ranging in age from 6 to 21 months were included in the study. The babies were classified into five screening levels according to their age. The items of a Screening Test of Early Language Development-Test version (STELD-T) were created and validated through expert opinion. The STELD-T was administered along with the Receptive Expressive Emergent Language Scale (REELS). Internal consistency using the Kuder-Richardson Formula-20 ranged from 0.457 to 0.853 across the five levels, acceptable owing to short tool length and item heterogeneity. Kappa coefficients indicated 0.459 to 0.875 agreement between the STELD-T and the REELS indicated satisfactory criterion validity. After calculating the percentage of babies with a “refer” result as well as Kappa statistics with three different pass-refer criteria, a pass-refer criterion of 75% seemed to be appropriate for screening. The STELD seems to be a reliable and valid tool to screen language development in babies from 6 months to 18 months of age in urban areas of Maharashtra. Items representing a range of language skills including pragmatics make it a unique tool.