Search for: All records

People go online for information and support about sensitive topics like depression, infertility, death, or divorce. However, what happens when such topics are algorithmically recommended to them even if they are not looking for it? This article examines people's self-diagnostic behaviors based on algorithmically-recommended content, for example, wondering if they might have depression because an algorithm pushed that topic into their view. Specifically, it examines what happens when the sensitive content is not generated by users, but by companies in the form of targeted advertisements. This paper explores these questions in three parts. The first part reviews literature on self-diagnosis and targeted advertising. The second part presents a mixed-methods study of how targeted ads can enable self-diagnostic reactions. The third part reflects on the mechanisms that influence self-diagnosis and examines potential regulatory implications. 

This project illuminates what data youth believe online advertisers and social media companies collect about them. We situate these findings within the context of current advertising regulations and compare youth beliefs with what data social media companies report collecting based on their privacy policies. Through interviews with 21 youth ages 10-17 in the United States, we learn that participants are largely aware of how their interactions on the website or app are used to inform personalized content. However, certain types of information like geolocation or how long data is retained is less clear to them. We also learn about what school and family factors influence youth to adopt apps and websites. This work has implications for design and policy related to companies' personal data collection and targeted advertising, especially for youth.