ETHICAL CONSIDERTIONS IN USAGE OF TWITTER DATA

Abstract

Internet-based data, especially the ability to extract and analyze data from social media such as Twitter, is becoming more prevalent in providing data in empirical research. The public nature of Twitter and its more open access to data, compared to other social media, as well as the approval to use that data in accordance with the Twitter’s Terms of Service were considered a sufficiently ethical justification for the use of Twitter data in research. This tendency to collect data from the Internet, that is from social networks, in social research, has prompted a number of scientific societies to develop ethical guidelines for Internet research. This paper provides an overview of the recommendations contained in the ethical guidelines, compared to the requirements set out in the Twitter’s Terms of Service. Additionally, research on social media users’ perceptions regarding the use and publication of their data is analyzed. The tendency to apply the principle of situational ethics is evident in the ethical guidelines, starting from the existing collision between the established ethical principles for scientific research and the privacy policy on Twitter. However, there is a lack of consensus among ethical guidelines in determining the appropriate way of collecting, processing and presenting data in research and scientific publications. Also, in research on the perception of social media users, there is a significantly lower concern about the use of data by Twitter users, compared to users of other platforms. But in spite of that, the prevailing opinion among them is the need for prior consent for the use of their data in research and scientific publications, with a special emphasis on the request for anonymity. Given the complexity of this debate, which ultimately aims to preserve the academic integrity of research, the conclusion points to the need to summarize the various ethical aspects in establishing a methodological approach in studies that use Twitter data.