'I'm still the master of the machine.' Internet users' awareness of algorithmic decision-making and their perception of its effect on their autonomy. Issue 9 (4th July 2022)

• Record Type: Journal Article
• Authors: Dogruel, Leyla; Facciorusso, Dominique; Stark, Birgit
• Is Part Of: Information, communication & society. Volume 25:Issue 9(2022)
• Page Start: 1311
• Page End: 1332
• Publication Date: 2022-07-04
• Subjects: Perceptions of algorithms -- user autonomy -- transparency -- socio-technical systems; Information technology -- Social aspects -- Periodicals; Information technology -- Economic aspects -- Periodicals; Information technology -- Political aspects -- Periodicals; Internet -- Periodicals; World Wide Web -- Periodicals; 303.4833
• DOI: 10.1080/1369118X.2020.1863999
• Languages: English
• ISSNs: 1369-118X

Abstract

ABSTRACT: Algorithms are an integral part of our everyday lives and shape the selection and presentation of information and communication on the internet. At the same time, media users are faced with a lack of control and transparency when interacting with these systems because algorithms largely remain black boxes to end users. Relying on the notion that algorithms are socio-technical systems that comprise both technical and human components, this paper examines internet users' awareness of algorithms in different areas of internet use and inquires into users' perceptions of the impact of algorithms on their autonomy when interacting online. Empirically, we rely on qualitative interviews with 30 German internet users. Findings indicate that users in general are aware of algorithms operating in a wide range of applications and demonstrate a basic understanding of how these systems work. In line with the third-person effect, users perceive algorithms to have a stronger impact on others' internet use than on their own. Further, users' awareness of algorithms was found to be closely related to their perceived autonomy. When users feel in control of their interactions online, they are less aware of the impact of algorithms governing their interactions. Based on these results, we discuss the implications for transparency measures in algorithm regulation.