Beyond Optimism: The Cruel ‘Iron Cage’ of Data Science in Education

• In Event: Algorithms at Work: The Practice of Prediction - I: Algorithms-in-Practice in Institutional and Organizational Perspective

Abstract

Education policymakers and administrators use data science to measure performance outcomes of students in order to achieve standardized goals. Faith in data science within education more often than not borders on a deterministic understanding of scientific and technical progress. Education data science, with its objective of complete quantification, claims to provide targeted instruction, a closing of the achievement gap, and the creation of more objective measures of achievement, and, thus, be a remedy to the ills of contemporary education. It is the argument of this paper that data science’s goals of ensuring equity, efficiency, and productivity in American school systems merely represent old positivistic views hidden in the veneer of a new data optimism. The introduction of data science into educational practices reinforces current trends in educational managerialism, corporatism, and outcome-based or evidence-based education. This paper, using data from ethnographic interviews with educational practitioners, researchers, and activists, further argues that the case of education data science is not merely a case of techno-optimism, but can be better understood within Lauren Berlant's theoretical framework of ‘Cruel Optimism’. Bringing critical perspectives on the possible impacts of modeling education through reductionist logics of metrication, this paper argues that the application of data science in education has two effects: 1) the exacerbation of latent tensions between the goals of education policymakers and the tools chosen to achieve them, and 2) that data science reinforces the construction of positivistic subjectivities amenable to the analysis of data science.

Author

• Caroline Mason, Rensselaer Polytechnic Institute - STS