Inequity has become an intractable problem in education. Improving the equity of school systems has, rightly, become a central goal of most school systems around the world. This has led to the widespread measurement of inequity in terms of outcomes. Among these, popular in Australia are PISA's international comparisons of the equity of school systems. PISA's quality-equity grid, the comparative 'slope' of inequity and the 'tail of underachievement' are frequently used in policy and academic circles and are the basis for some of the most significant policy decisions.
The increasing use of international comparisons of education equity raises several questions with regard to the nature, usefulness and consequences of such comparisons, and the levels of specificity at which they may be meaningful. The challenge of making the abstract notion of 'equity of educational systems' measurable is compounded when the measures have to be comparable, meaningful, relevant and accessible to policy makers in countries as diverse as Iceland and Israel, imposing further demands on the design of surveys, the data collected, and the way they are represented.
This study examines international comparisons of equity as a knowledge practice using OECD's PISA as a case in point, tracing the processes by which complex patterns of inequity in vastly different contexts are made comparable, mapped, translated and represented as graphs and tables. Drawing on the conceptual and methodological resources of actor-network theory (ANT), this is a sociological study of measurement which traces, step-by-step, the processes used by statisticians as they gradually eliminate uncertainties to produce mobile and decontextualised 'policy objects'.
Two data sets inform this study. The first consists of focused and detailed interviews with a number of PISA experts, and the second an analysis of PISA documents and published articles on the subject of equity measurement.
By turning the analytical eye to the measurement of equity and by problematizing the international comparisons they generate, this study opens up the 'black box' of equity measurement, moving it from the realm of statistical expertise and making it available for wider public debate. The use of ANT in education is relatively novel, and this study adds to the small but growing number of ANT studies in the field.
Abstract: