Exploring the impact of a ‘confining’ imaginary of user-recommendation systems on platform usage and relationship development among dating app users

Algorithmic recommendation systems (ARM) on dating apps serve users with a personalised feed of profiles from other users based on the inferred preferences of the user being served. Despite concerns linking ARM to problematic dating app use and negative social outcomes, it has been suggested that critical awareness of ARM's limitations, such as that ARM restrict user choice (i.e. a ‘confining’ perception of ARM, or CP-ARM), can mitigate problematic usage and reduce negative social outcomes. This study tested such a prediction with semi-structured interviews (N = 20) and a subsequent survey (N = 349), which yielded surprising results – while CP-ARM can indirectly decrease compulsive use of dating apps by lowering the perceived usefulness of dating apps, it can directly increase compulsive use, which can be attributed to a sense of helplessness in controlling digital media use. Consequently, compulsive use can decrease the intention to commit in Internet-initiated romantic relationships. The finding suggests that researchers should not assume that critical awareness of algorithms leads to less problematic usage and better social outcomes but situate the inquiries in a broader socio-cultural context where everyday life is increasingly mediatised by various social platforms and individuals find it difficult to opt out.