Big Data: Some Ethical Concerns for the Social Sciences

Michael Weinhardt (2021)

Social Sciences, 10, 36. Open Access

Abstract. While big data (BD) has been around for a while now, the social sciences have been comparatively cautious in its adoption for research purposes. This article briefly discusses the scope and variety of BD, and its research potential and ethical implications for the social sciences and sociology, which derive from these characteristics. For example, BD allows for the analysis of actual (online) behavior and the analysis of networks on a grand scale. The sheer volume and variety of data allow for the detection of rare patterns and behaviors that would otherwise go unnoticed. However, there are also a range of ethical issues of BD that need consideration. These entail, amongst others, the imperative for documentation and dissemination of methods, data, and results, the problems of anonymization and re-identification, and the questions surrounding the ability of stakeholders in big data research and institutionalized bodies to handle ethical issues. There are also grave risks involved in the (mis)use of BD, as it holds great value for companies, criminals, and state actors alike. The article concludes that BD holds great potential for the social sciences, but that there are still a range of practical and ethical issues that need addressing.

Extract: “BD also opens up completely new possibilities for political interference, state control, and surveillance. In the political realm, the Facebook/Cambridge Analytica scandal showed how the harvesting of personal information on social networks may be used to develop data-driven political campaigns, even if they are designed to misinform or spread falsehoods to influence election results (Confessore 2018). The knowledge gained on the connection between social and political attitudes and preferences proved to be useful to influence people’s political preferences. This helps political campaigns in micro-targeting their political ads, but also was exploited by Russian intelligence agencies in Western elections by using bots and false accounts to spread false and misleading information to undermine unwanted and promote more desired candidates and parties, respectively. This poses grave risks not only at the individual level, but also to whole political systems.
State actors have for some time now recognized the vast potential that comes with the digitalization of life and communication. Historically, official statistics have been developed by state actors to govern their populations (Diaz-Bone 2019) and administrations use BD for this purpose too (Thévenot 2020). For example, database matching has always been a desire of police forces and security agencies as they claim it helps them identifying and catching criminals and terrorists. With the advent of widespread security cameras in private as well as public spaces, the developments in face recognition software, computing power, and bandwidth, the police increasingly receive the power to track almost anyone anywhere for (hopefully) legal purposes.” (p. 6)

Link to the article