Scandals regarding the handling of data and manipulation of voting behavior through digital channels and social media have increasingly affected the companies operating in this space in recent years, with Facebook being the most prominent example. All major platforms reacted to this issue by increasing their transparency efforts and some today expose political advertising data through an application programming interface (API). Following an action design research (ADR) framework, this thesis project tries to solve data accessibility issues of the Facebook Ad Library API for its identified core user groups by building an IT artifact that provides the data as a simple file download instead of a complex coding challenge. The author argues that the main audience for the exposed data set are researchers and journalists, two groups that do not have the resources to access data in the way required by the API. This hypothesis is confirmed using semi–structured interviews with subject matter experts and elevated to a larger set of problems by analyzing digital literacy in the social sciences and the quality of research APIs more generally. A skills gap with regard to computer science capabilities in the social sciences is confirmed and, based on the creation of the IT artifact and the data collection process, several solutions are suggested. (1) Increase upskilling efforts in the social sciences and teach researchers how to code. (2) Improve the accessibility and quality of research APIs by adopting universal standards and (3) demand from platforms to build more no–code solutions for researchers and journalists. Finally, an updated ADR framework is suggested that includes the stakeholders and collaboration in a researcher–practitioner–user–API provider environment and a more general argument is made regarding who should be the arbitrator of truth with regard to digital political communication.
|Uddannelser||Cand.merc.it Business Administration and Information Systems, (Kandidatuddannelse) Afsluttende afhandling|