Evaluating a ten questions screen for childhood disability: reliability and internal structure in different cultures

Abstract

This paper uses five strategies to evaluate the reliability and other measurement qualities of the Ten Questions screen for childhood disability. The screen was administered for 22,125 children, aged 2-9 years, in Bangladesh, Jamaica and Pakistan. The test-retest approach involving small sub-samples was useful for assessing reliability of overall screening results, but not of individual items with low prevalence. Alternative strategies focus on the internal consistency and structure of the screen as well as item analyses. They provide evidence of similar and comparable qualities of measurement in the three culturally divergent populations, indicating that the screen is likely to produce comparable data across cultures. One of the questions, however, correlates with the other questions differently in Jamaica, where it appears to "over-identify" children as seriously disabled. The methods and findings reported here have general applications for the design and evaluation of questionnaires for epidemiologic research, particularly when the goal is to gather comparable data in geographically and culturally diverse settings.