Smart toys and children's privacy: usable privacy policy insights from a card sorting experiment

Smart toys are new to the Internet of Things market, and its connectivity to the cloud have raised concerns about children's privacy. Parents and legal guardians have striven to protect the privacy of their owns. However, current approaches for privacy control still lack usability for lay people. In this paper, we have explored the use of Card Sorting to enhance the usability of a privacy control for smart toys. Our goal was to identify and describe benefits of this technique to the design of more usable privacy controls. For this reason, we conducted a case study with voluntarily participants. We chose a parental control model from the literature to be the subject of evaluation for the experiment. Therefore, we extracted 19 units of information from its interface, and put them into cards for the Card Sorting evaluation. After the experiment, we obtained 30 valid responses. From these responses we performed a cluster analysis to understand the best alternative to group privacy related contents. Our contributions include a new model for nutrition label style mobile parental privacy controls for smart toys, suggestion of Google Material Design icons to be applied as indication for groups of privacy policies and, finally, a six steps process to perform Card Sorting with cluster analysis that does not rely on users' discussions to compose the Information Architecture hierarchy.