Abstract
A realist framework, integrating qualitative and quantitative methods, is advocated for the sociology of education. The case for a 'numbers and narratives' approach is supported by an analysis of several fundamental problems. Realist forms of sociological explanation require specification of properties at the three distinct levels of social structure, individual disposition, and action within social practice. It is argued that the quantification of these properties is most satisfactorily achieved within a materialist theory of measurement. Many of the problems that arise with quantitative work are shown to have their origin in an inadequate theory of measurement and to be sustained by an unsound concept of 'statistical explanation'. It is concluded that the most satisfactory response to these problems, however, lies not in the celebration of qualitative work, situated in a relationship of opposition, but in the development of an integrated structure-disposition-practice explanatory scheme for the sociology of education.

This publication has 0 references indexed in Scilit: