Mental Models of Online Privacy: Structural Properties with Cognitive Maps

Individuals usually build small-scale representation of reality to help them navigate their environment. Although mental models have been used in HCI before, they mostly occur as analogies and metaphor within the privacy and security research space. The meaning for users, the values associated and reasoning over online privacy have not been investigated before. In our research we explore and depict users’ mental models of online privacy through the content, properties and structure of privacy mental models. We believe mental models provide a framework for understanding user cognitive processing and reasoning and consequently privacy decison-making. In this paper we present an on-going study that use Amazon’s Mechanical Turk and cognitive mapping technique to elicit and illustrate mental models. We compare the cognitive maps generated for two different questions and analyse their structural properties. We find that while a list of concrete privacy evaluations populate the cognitive maps when asked directly about privacy, the examples are generally scarce if not absent when queried about personal importance of the online environment. We also find that the degree of vertices complemented with the source and sink vertices can help to identify key concepts, triggering links and clusters within the maps.