Test selection, adaptation, and evaluation : a systematic approach to assess nutritional influences on child development in developing countries

Background. Evaluating the impact of nutrition interventions on developmental outcomes in developing countries can be challenging since most assessment tests have been produced in and for developed countr ys ettings. Such tests ma yn ot be valid measures of children’ sa bilities when used in an ew context. Aims. We present several principles for the selection, adaptation, and evaluation of tests assessing the developmental outcomes of nutrition interventions in developing countries where standar da ssessment tests do not exist. We then report the application of these principles for an utrition trial on the Indonesian island of Lombok. Sample. Three hundred children age 22–55 months in Lombok participated in a series of pilot tests for the purpose of test adaptation and evaluation. Four hundred and eighty-seve n4 2-month-ol dc hildren in Lombok wer et ested on the finalized test battery. Methods. The developmental assessment tests wer ea dapted to the local context and evaluated for an umber of psychometric properties, including convergent and discriminant validity ,w hich wer em easured based on multiple regression models with maternal education, depression, and age predicting each test score. Results. The adapted tests demonstrated satisfactor yp sychometric properties and the expected pattern of relationships with the thre em aternal variables. Maternal education significantly predicted all scores but one, maternal depression predicted socio-emotional competence ,s ocio-emotional problems, and vocabulary ,w hile maternal age predicted socio-emotional competence only.