The Ethics and Governance of Computational Social Science

It is by now common knowledge that the availability and use of new types and greater amounts of digital data permeates and fundamentally transforms many aspects of human life (Brady, 2019). Examples of the use of “big data” are plentiful and range from the use of digital data for the purpose of assessing creditworthiness to the provision of child welfare (Hurley & Adebayo, 2017; Redden, Dencik & Warne, 2020). In an increasingly data-rich world, the boundaries of which types of information can be extracted from data are shifting significantly (Behnke et al., 2018).

 It is not only private companies and governments who explore ways of using digital data for different purposes. Social science researchers have joined their ranks (Lazer et al., 2009). The number of papers studying social phenomena through the use of digital data and computational methods has exploded – manifesting in the rise of Computational Social Science (CSS) (Lazer et al., 2020). CSS, can be broadly defined as the “study of social phenomena using digitised information and computational and statistical methods” (Wallach, 2014). It offers an immense, if not paradigmatic, potential for changing the way we view knowledge production in the social sciences. New questions can seemingly be answered in a granular, cost-effective and quick manner (Salganik, 2019). This promising potential of CSS to result in quick, accurate and inexpensive knowledge production has, however, been accompanied by several shortcomings - among them the lack of “ethical guidance” for CSS researchers.

In brief, existing research ethics institutions neglect risk emanating from the analysis (and merging) of data collected and instead focus on risks emanating at the stage of data collection (Metcalf, 2019). Another issue stems from the fact that informed consent is usually obtained for a singular event and only in relation to the study participant’s own body (in medical sciences) or data (in medical and social sciences). This changes with regard to CSS. Consent given by one individual can also allow researchers to scrape data from that person’s network of friends (Metcalf, 2019). Another development is the role of "private" data collectors such as social media platforms. The role of researchers is changing accordingly. In many cases they no longer collect the data on which their research is based, but purchase or scrape it. 

Against this backdrop, Seliem’s dissertation explores the genesis and development of scientific disciplines and the extent to which CSS can be considered as such in Germany, Austria and Switzerland. He investigates the relationship between law and ethics in the governance of CSS, and the ethical questions occupying CSS researchers themselves. Seliem’s research further explores in detail when digital data use in general, and CSS research in particular, creates public value. Alongside being part of the Research Platform, his work is also embedded in the Digitize! Project.