The study removed the element of socioeconomics by considering directly comparable communities. It looked at single-stream, single-age classes; schools where children from all five years of French primary education were in one class; and schools in which children from two or three years were in one class.