Social Constructions of Data and the Implications of Data Representation

Abstract

The ever growing institutionalization of computing power in our society has recently spawned a critical body of work that examines the likes of algorithmic bias, the commercial role of trace data, and the turn to computational ways of knowing. What has received less attention to date, however, is the investigation of the generation, storage, retrieval, and analysis of data itself--the mechanics of making, using, and interpreting data for various organizational ends. Organizations increasingly have high expectations for and, consequently, invest heavily in creating large datasets that, through strategic algorithmic processing, are envisioned as a key, if not the key, means for them to inform, mediate, or autonomously carry out decisions.

This panel draws together researchers conducting empirical work on the social constructions of data and data representations. The proposed papers all take up the topic of the social construction of data, yet differentiate themselves by showing a broad range of organizational contexts in which these social constructions play out. We present work that investigates how data are constituted and then comparatively evaluated as objective representations of quality. Similarly, we showcase how stakeholders operationalize a machine learning tool to produce (and contest) actionable “evidence.” In parallel, we present two papers that question data’s relationship to broader questions such as urban governance and time management. In sum, we aim to create a critical, constructive conversation around data, data production, data representations, and “data-driven” logics--all of which are too often viewed as objective and scientific, and consequently left under interrogated.

Discussant

• Carsten Østerlund

Individual Presentations

• Organizing for Data: Crafting Data Elements in the ‘Data-Driven’ Hospital - Kathleen H Pine, Arizona State University; Melissa Mazmanian

• What’s in a Number?: Integrating Machine Learning into a Clinical Context - Madeleine Elish, Columbia University; Elizabeth Anne Watkins, Columbia University

• Data-Driven Representation in the Era of Big Data - Leah Horgan, University of California, Irvine

• Time as Data: Silicon Valley’s Quest for Temporal Optimization - Ingrid Erickson, School of Information Studies, Syracuse University; Judy Wajcman, LSE

• Towards Understanding Academia and Digitalisation in Their Relationship of Strong Relationality - Seppo Poutanen, University of Turku, Turku School of Economics

Session Organizers

• Ingrid Erickson, School of Information Studies, Syracuse University

• Kathleen H Pine, Arizona State University