ABSTRACT
Data on the differential distribution of students between different institutions is important for three quite distinct reasons. First, it is essential for controlling input in studies of differential institutional effects. Second, it is important because the students at an institution are a major determinant of the 'environment' which influences the behaviour and experience of each student within it. Third, data on the differential distribution of students is important because, in so far as it reflects the sum of individual students' choices, it may provide important clues to explain how these choices were made. Moreover, in so far as the differential distribution by ability is due not to selection but to self-selection, it must be remembered that at the time most students apply. The colleges of education had a distribution fairly similar to that of the civic universities, except that they recruited a particularly small proportion of men from independent schools.
