Broken Expectations, Broken Concepts: A New Diagnosis of Dark Patterns

Abstract

Interest in analyzing and preventing dark patterns, commonly defined as design that is manipulative or deceptive, has been rising. Most analyses of dark patterns today are descriptive, describing and categorizing examples of these patterns in user interfaces, making it difficult for stakeholders to identify the root causes of more structural and nuanced patterns. In this thesis, I propose a new definition: dark patterns emerge when a given design breaks users’ expectations and harms their interests. To formally define and analyze expectations, I use an existing design framework called concept design, which defines applications in terms of independent building blocks called concepts. I present a new type of design catalog built on concepts to empower organizations to set standards around expectations. I then introduce a formal definition of design extensions within concept design, bringing further precision to dark pattern analysis. This concept-based approach to analyzing dark patterns enables the identification of more subtle dark patterns at both a user interface level and a structural level, more precise analysis of the root cause of darkness within a design, and more nuanced consideration of the impact of cultural expectations on perceptions of darkness.