Although I wish to draw on research into ethnic minority groups for examples to demonstrate the hypotheses I wish to put, I see no reason why the same hypotheses need not apply equally to many other aspects of public sector social research. The background of my paper is that in a period of increasing unemployment in Western Europe, there are many countries which have a significant proportion of their population from ethnic minority groups. Amongst these minority groups there are often problems which arise either from unemployment or from under-employment; from conflicting cultural backgrounds, or from a tendency to live often in near ghetto conditions in the larger towns and cities. Survey research in this context has a dual function. The first is to establish and to demonstrate facts, particularly in a country which has not yet come to terms with the need to identify in its census the ethnic background of its population.