How do children acquire knowledge about voice assistants? A longitudinal field study on children’s knowledge about how voice assistants store and process data

Abstract

Voice assistants (VAs) are available in the family context and are easy to interact with. Due to the opaqueness of these assistants, however, children in particular might be unaware of potential negative consequences (e.g., with regard to privacy). To gain a deeper understanding of children’s knowledge about VAs, of how to facilitate knowledge regarding VA technology (focusing here on parental interventions), and of how children’s knowledge and understanding affects their communicative behavior towards VAs (focusing here on telling a secret to the VA), we conducted a longitudinal field study with 16 families (N=20 children). The emphasis was on differentiating between how VAs store and process data, as both aspects are important for a sophisticated usage but vary in how they are experienced or understood. The qualitative data demonstrated that children draw observable inferences from VAs’ behavior, which explains why children know more about how data is processed than about how it is stored or accessed. Our results indicate that parental interventions positively affect children’s understanding of data processing, but that this cannot be transferred to data storage, indicating the need for self-explaining systems. Lastly, knowledge about how data is stored negatively predicts the willingness to entrust a secret to the system, which underlines the importance of a basic understanding of how technologies work, also for younger users.