The Becoming-Statistic: Information Ontologies and Computerized Adaptive Testing in Education. (October 2016)

• Record Type: Journal Article
• Authors: Sellar, Sam; Thompson, Greg
• Other Names: de Freitas Elizabeth guest-editor.; Dixon-Román Ezekiel guest-editor.; Lather Patti guest-editor.
• Is Part Of: Cultural studies, critical methodologies. Volume 16:Number 5(2016:Oct.)
• Page Start: 491
• Page End: 501
• Publication Date: 2016-10
• Subjects: data -- Deleuze -- CAT -- control society -- information -- ontology; Culture -- Research -- Periodicals; Popular culture -- Research -- Periodicals; 306.05
• DOI: 10.1177/1532708616655770
• Languages: English
• ISSNs: 1532-7086

Abstract

This article examines how computerized adaptive testing functions in relation to learning in control societies. We first document the transition from static and discrete forms of statistical work that characterized Foucault's disciplinary societies into the continuous, predictive analytics that have emerged as the powerful form of statistical work in Deleuze's control societies. We then explore the function of information science ontologies in adaptive testing and learning applications from the perspective of Deleuze's philosophical ontology. Working between these two conceptions of ontology enables us to open a critical space in which to posit the need for an alternative ontology of number in education. Focusing on the case of Pearson, the world's largest edu-business, we consider how the "datafication" of education is presenting opportunities to exploit information assemblages for profit. The primary focus of analysis is Pearson's Next Generation Assessment agenda, which focuses on the development and implementation of computerized adaptive testing within a broader digital learning environment. Next Generation Assessment is theorized as an information assemblage that functions according to an axiomatic modeling of numerical data enabling the production and communication of information throughout proliferating data infrastructures in education. We argue that the shift from "becoming a statistic" in disciplinary society to "the becoming-statistic" in control society is facilitating the development of digital learning platforms that risk limiting the conditions for learning in the creative sense of this term.