This paper discusses issues in the development of a methodology appropriate for eliciting sound quantitative data from primary school children in the complex contexts of ethnolinguistically diverse developing countries. Although these issues often occur in field-based surveys, the large extent and compound effects of their occurrence in ethnolinguistically diverse developing countries, particularly in research with young children, means that they combine to be imperative for project viability in these contexts. At the same time, the scale of the challenges often also means that they are regarded as unsurmountable and are disregarded at the very time when the need is highest to provide sound quantitative data on which to base education policy. The paper provides a framework of these interlocking issues relating to a survey's overall approach, instruments, the sampling scheme, and implementation. This is illustrated in detail via a large-scale cross-national study carried out with children in each grade level of Years 1-6 in three Pacific Island countries of Vanuatu, Kiribati, and Solomon Islands.