Miriam Metzger and colleagues received grant funding through the National Science Foundation ($1,199,958) on social media advertising.
“Methods and Tools for Effective, Auditable, and Interpretable Online Ad Transparency”
Lead PI: Damon McCoy, New York University
PI: Miriam Metzger, UCSB
PI: Blase Ur, University of Chicago
PI: Michelle Mazurek, University of Maryland
Targeted online advertising is now a ubiquitous presence on many websites. Search engines such as Google and online social networks such as Facebook provide powerful targeting abilities and algorithms for advertisers to deliver their messages to specific end users. These targeted ads impact people’s lives in a variety of ways, such as what job, housing, and credit opportunities they may encounter. At a larger level, they may also have implications for democracy as a whole if used in election interference efforts.
Prior work on understanding targeted online advertisements has largely depended on ’black box’ measurement methods that are difficult to scale and sustain. However, there is hope that this situation can be improved, as platforms have recently started making transparent to researchers and end users more data about advertisements. For example, Facebook has made ad-sponsor information available and transparent for all active advertisements, and Facebook, Google, and Twitter have started to provide some targeting information to end users. Despite these efforts, initial studies find significant shortcomings in the transparency mechanisms provided thus far. This makes clear a need for improved methodologies to ethically investigate these mechanisms so that we can better understand their current limitations—and thus how to overcome existing limitations to make transparency useful to researchers, journalists, civil-society groups, and end users. The good news is that the platforms have demonstrated a willingness to expand and improve their transparency efforts. One goal of our research in this project is to assist them in doing so.