Abstract

This is a response to an article by Stephen Gorard in a previous issue of the journal. It addresses the issue of how achievement gaps in educational performance between ethnic and other groups, and changes in these, can best be measured. The approach recommended by Gorard is compared with that of the authors he criticises. The conclusion reached is that both approaches are of value: that they provide different kinds of information. Which is the most useful on any occasion depends on the purpose guiding the inquiry, and on the assumptions that can be made about the social processes being modelled.