Dark Patterns in Shopping, Education & Health Apps

Dark patterns are deceptive or manipulative design techniques in technology. They pose challenges for policymakers because they impact user autonomy and financial well-being. However, the lack of quantifiability of dark patterns complicates regulation efforts, and much of the existing research remains theoretical, lacking practical application. This study aims to address these gaps by identifying and quantifying various types of dark patterns in day-to-day software applications across different domains, namely shopping, health and fitness, and education. Drawing from the conceptual framework outlined in previous research, the top three apps in Canada within each application domain were selected for analysis. By identifying dark patterns and assessing the attributes of the identified dark patterns, each app was assigned a darkness score. The findings revealed that Temu, Yuka-Food & Cosmetic Scanner, and Duolingo reported the highest darkness scores within their respective categories, with scores of $7.5,5$, and 6.5, respectively. Furthermore, the shopping category exhibited the highest mean darkness scores, indicating a greater prevalence of dark patterns in this domain. Across all categories, the modification of decision space emerged as the most commonly influenced choice architecture. This study contributes to a deeper understanding of where regulatory efforts should be focused in addressing dark patterns in applications. It also sheds light on the nature of different types of dark patterns, their utilization across various domains, and the similarities and differences in their implementation. Moreover, by quantifying darkness in applications, this research underscores the importance of quantifiability in the evaluation of dark patterns, providing valuable insights for policymakers, researchers, and practitioners.