Search

This paper critically examines the intersection of privacy concerns in children’s apps and the support required by developers to effectively address these concerns. Third-party libraries and software development kits (SDKs) are widely used in mobile app development, however, these libraries are commonly known for posing significant data privacy risks to users. Recent research has shown that app developers for children are particularly struggling with the lack of support in navigating the complex market of third-party SDKs. The support needed for developers to build privacy-friendly apps is largely understudied. Motivated by the needs of developers and an empirical analysis of 137 ‘expert-approved’ children’s apps, we designed DataAvalanche.io, a web-based tool to support app developers in navigating the privacy and legal implications associated with common third-party SDKs on the market. Through semi-structured interviews with 12 app developers for children, we demonstrate that app developers largely perceive the transparency supported by our tool positively. However, they raised several barriers, including the challenges of adopting privacy-friendly alternatives and the struggle to safeguard their own legal interests when facing the imbalance of power in the app market. We contribute to our understanding of the open challenges and barriers faced by app developers in creating privacy-friendly apps for children and provide critical future design and policy directions.

Despite efforts to augment or replace the 2-dimensional spreadsheet grid with formal data structures such as arrays and tables to ease formula authoring and reduce errors, the flexible grid remains overwhelmingly successful. Why? We interviewed a diverse sample of 21 spreadsheet users about their use of structure in spreadsheets. It emerges that data structuring is subject to a complex network of incentives and constraints, including factors extrinsic to spreadsheets such as the user’s expertise, auxiliary tools, and collaborator needs. Moreover, we find that table columns are an important abstraction, and that operations such as conditional formatting, data validation, and formula authoring can be implemented on table columns, rather than cell ranges. To probe this, we designed 4 click-through prototypes for a follow-up study with 20 participants. We found that although column operations improved the value proposition of structured tables, they are unlikely to supplant the advantages of the flexible grid.

Smart homes are under attack. Threats can harm both the security of these homes and the privacy of their inhabitants. As a result, in addition to delivering pleasant and aesthetic experiences, smart devices need to protect households from vulnerabilities and attacks. Further, the need for user-centered security and privacy design is particularly important for such an environment, given that inhabitants are demographically-diverse (e.g., age, gender, educational level) and have different skills and (dis)abilities. Prior work has explored different usable security and privacy solutions for smart homes; however, the applicability of user eXperience (UX) principles to security and privacy design is under-explored. This research project aims to address the on-going challenge of security and privacy in the smart home through the lens of UX design. The objective of this thesis is two-fold. First, to investigate how UX factors and principles affect the security and privacy of smart home users. Secondly, to inform product design through the development of an empirically-tested framework for UX design of security and privacy in smart home products. In the first step, we explored the relationship between UX, security, and privacy in smart homes from user and designer perspectives: through (i) conducting a qualitative interview study with smart home users (n=13) and (ii) analyzing an ethnomethodologically informed study of six UK households living in smart homes (n=6); and, we then explored the role of UX in the design of security, privacy and data protection in smart homes through qualitative semi-structured interviews with smart home users, designers and business leaders through two rounds of interviews (n=20, n=20). In the second step, using conceptual framework analysis, we systematically analyzed our previously collected data and the literature to construct a framework of design heuristics for consent and permission in smart homes. We applied these heuristics in four participatory co-design workshops and reported on their use. We further analyzed the use of the heuristics through thematic analysis highlighting how the heuristics were used, their purpose, and their effectiveness. By bringing UX design to the smart home security and privacy table, we believe that this research project will have a significant impact on academia, industry, and government organizations. Our thesis will improve design practices for security and privacy in domestic smart devices while addressing wider challenges, opportunities, and future work.