New data, old tensions: Big data, personalized learning, and the challenges of progressive education

Abstract

Personalized learning has become the most notable application of big data in primary and secondary schools in the United States. The combination of big data and adaptive technological platforms is heralded as a revolution that could transform education, overcoming the outdated classroom model, and realizing the progressive vision of interest-driven and self-initiated learning. Yet, even supporters concede that, in practice, personalized learning is geared toward behaviorist models of learning. This article explores how this gap between expectation and reality concerning big data can be understood as a reflection of existing tensions within progressive education. To explain the tension, I examine the interplay between Rousseau and Dewey’s theories of education, and the novel opportunities offered by big data. I hold that personalized learning could be understood as an iteration of Rousseau’s vision of well-regulated freedom, in which students’ freedom is perceived as a means toward increasing the effectiveness of their learning. The relegation of decision making to algorithms renders this regulation more feasible and justifiable. Dewey’s critique of Rousseau’s individualized and teleological model of education offers the contours of an alternative role for big data in education, which prioritizes social interaction and the cultivation of democratic citizens. Moreover, due to the increased capacity to operationalize complex learning processes in naturalistic learning environments, big data could allow tackling some of the lingering challenges to implementing Dewey’s approach.