Algorithmic recommendations in the everyday life of young people: imaginaries of agency and resources

Abstract

Algorithmic recommender systems have become a regularised part of the current information infrastructure, as they have been increasingly integrated into many online platforms. This study aims to examine and thus improve our understanding of young people’s agency and their imaginaries related to algorithmic recommendations within the information infrastructure by applying Giddens’s (Citation1984) conception of resources to the domain of recommender systems. The study is guided by the following research questions: How do imaginaries of algorithmic recommendations shape young people’s agency? How do different resources constitute information and information infrastructure in this context? Thematic interviews were conducted with 20 young Finnish people, and the material was analysed using qualitative content analysis with a focus on agency, resources and imaginaries in relation to algorithmic recommendations on different online platforms. The key findings of the study depict the duality of power relations: recommender systems limit young people’s agency by offering them scarce opportunities for self-expression and determining how their data is being used, yet young people were able to use resources for their own benefit. The notion of allocative resources was useful for noticing how participants’ imagined their personal information being utilised to generate recommendations, while authoritative resources offered insights into how users expressed placing trust on different platforms based on their size and the kind of opportunities participants had for self-expression on these platforms. This study contributes new knowledge regarding the ways algorithmic recommendations shape the agency of young people by making them adapt to the logic of recommender systems.