The Ethical Challenges of Publishing Twitter Data for Research Dissemination

Abstract

Empirical research involving the analysis of Internet-based data raises a number of ethical challenges. One instance of this is the analysis of Twitter data, in particular when specific tweets are reproduced for the purposes of dissemination. Although Twitter is an open platform it is possible to question whether this provides a sufficient ethical justification to collect, analyse and reproduce tweets for the purposes of research or whether it is necessary to also undertake specific informed consent procedures. This paper reports on an ethics consultation that formed part of a wider research study and that aimed to identify best practice procedures for the publication of Twitter data in research findings. We focus largely on the UK context and draw on the outcomes of the consultation to highlight the range and depth of ethical issues that arise in this area. We can see Twitter as a case study for a wide number of data sources used in Web Science. This is a highly complex landscape in which questions crystallise around fundamental principles such as informed consent, anonymisation and the minimisation of harm. Furthermore, tensions exist between commercial, regulatory and academic practices, and there are also circumstances in which good ethical practice might compromise academic integrity. There is an absence of consensus in Web science and related fields over how to resolve these issues and we argue that constructive debate is necessary in order to take a proactive approach towards good practice.