Big Data and the brave new world of social media research

TLDR

The Facebook emotional‑contagion study ignited a debate over the ethical and social implications of Big Data research, revealing a neglected tension between free will and the deterministic expansion of scientific knowledge. The essay defines data, Big Data, and the transformation of the human environment by scientific knowledge, then examines the rapid rise of data‑driven human‑behaviour analyses and the differing implications of basic versus applied research. It explains how Big Data analyses are shaped by the definitions of data and the nature of available data sources, which both enable and constrain such research. The authors conclude that Big Data research will become increasingly pervasive, demanding heightened awareness from data scientists, policymakers, and the public about its contexts and unintended consequences.

Abstract

The recent Facebook study about emotional contagion has generated a high-profile debate about the ethical and social issues in Big Data research. These issues are not unprecedented, but the debate highlighted that, in focusing on research ethics and the legal issues about this type of research, an important larger picture is overlooked about the extent to which free will is compatible with the growth of deterministic scientific knowledge, and how Big Data research has become central to this growth of knowledge. After discussing the ‘emotional contagion study’ as an illustration, these larger issues about Big Data and scientific knowledge are addressed by providing definitions of data, Big Data and of how scientific knowledge changes the human-made environment. Against this background, it will be possible to examine why the uses of data-driven analyses of human behaviour in particular have recently experienced rapid growth. The essay then goes on to discuss the distinction between basic scientific research as against applied research, a distinction which, it is argued, is necessary to understand the quite different implications in the context of scientific as opposed to applied research. Further, it is important to recognize that Big Data analyses are both enabled and constrained by the nature of data sources available. Big Data research is bound to become more widespread, and this will require more awareness on the part of data scientists, policymakers and a wider public about its contexts and often unintended consequences.