Design and Evaluation of the UsersFirst Privacy Notice and Choice Threat Analysis Taxonomy

UsersFirst is a privacy threat modeling framework under development at Carnegie Mellon University (CMU) designed to help identify and mitigate user-oriented privacy threats associated with N & C interfaces. In this poster, we report on the first user study designed to evaluate the usefulness of an initial version of the privacy N & C threat taxonomy that is part of UsersFirst and evaluated its efficacy as compared to an existing taxonomy (LINDDUN PRO’s unawareness threat category). This initial version of the UsersFirst Taxonomy is organized around three major categories of threats (delivery, language & content, and presentation & design) and comprises a total of 27 different threat types. We selected privacy N & C interfaces from a well-known e-commerce platform and conducted semi-structured in-person interview sessions with 14 participants who had prior privacy experience. We found that privacy practitioners who use the UsersFirst Taxonomy were able to identify more user-oriented threats associated with privacy N & C than they identify without the use of a taxonomy, and that all UsersFirst Taxonomy users identify more threats than LINDDUN PRO users. Participants assigned to use the UsersFirst Taxonomy also commented it was easy to use and helped them better organize their thoughts during threat identification.