Privacy and Security Concerns When Social Scientists Work with Administrative and Operational Data

Abstract

Social science research is transitioning from working with “designated data,” collected through experiments and surveys, to working with “organic data,” including administrative data not collected for research purposes, and other data such as those collected from online social networks and large-scale sensor networks. The shift to organic data requires significant innovations in research methodologies. This article reviews the complexities and diversity of organic data and the special efforts that must be undertaken to make those data findable and usable by researchers. In some cases, advanced formal privacy techniques such as differential privacy and secure multiparty computation are needed to work with organic data in a manner that is ethically and logistically permissible, and effort is also required to make studies involving organic data transparent and replicable. These considerations make clear that moving forward, social scientists and information and communications technology (ICT) professionals must work closely to develop appropriate technical controls and ethical frameworks that minimize the risks of research to participants and to society at large.