Validity and reliability of the early development instrument in Indonesia

Abstract

There is increasing interest from international organizations and the research community to use internationally comparable instruments that in turn foster global understanding while providing evidence for local and international policy development. In the field of early childhood, international comparisons have traditionally been limited to indicators such as infant or child mortality and anthropometric data such as stunting and wasting. However, there has been gradual interest in developing international measures that can be used to compare and monitor the holistic development of children. Using both the short and standard versions of the Early Development Instrument (EDI), this paper reports on the process of adaptation of the EDI in Indonesia. Further, it explores the content and construct validity, internal consistency, inter-rater reliability and predictive validity of the EDI using a number of measures including the Strengths and Difficulties Questionnaire, the Dimensional Change Card Sort, and school-based tests of language, mathematics and cognitive performance, collected from a number of informants (caregivers, teachers, and children). We report on data for two cohorts of children: the “younger cohort” were approximately 1 year old (N = 3116) and the “older cohort” were approximately 4 years old (N = 3251) at Time 1. Both cohorts were followed up approximately 4 years later, at Time 2. This study finds that the EDI shows moderate validity and reliability in poor communities in Indonesia and highlights some of the difficulties associated with adapting western instruments for non-western cultures and contexts.