De-escalating ‘dataveillance’ in schools

If the transformations in K-12 education over the last two decades had to be encapsulated in a single phrase, “data-driven” would surely be a top contender. Policies like the No Child Left Behind Act inaugurated a vast infrastructure for collecting and analyzing data, and the Every Student Succeeds Act (2015) charged all educators to be facilitators of “data-based instructional decision making” (p. 295). At the same time, technologies for data collection became more sophisticated. New devices and apps have made it easy to amass more, and more kinds of, data about students. Over time, this information has been used to not only document young people’s academic progress but also, increasingly, mine for insights into factors that could influence their learning.

School safety has been one prime focus for data-driven interventions. A thriving market has emerged for technologies that monitor students’ behavior and alert educators at the earliest signs of trouble. These technologies range from the mundane (e.g., software that notifies administrators when certain websites or keywords are used on school computers) to the highly specialized (e.g., microphones that sweep classrooms for sounds associated with stress, fear, or anger). Such technologies are based on a shared belief: With enough data, we can identify and mitigate threats to school safety, perhaps even before they materialize.

The comfort this belief offers comes at a cost, however. Such technologies are expensive, and there is growing evidence that they don’t always work as promised — or, in some cases, at all (Casella, 2018; Gillum & Kao, 2019). In addition, many of these technologies subject young people to invasive forms of data-driven surveillance — or dataveillance (van Dijck, 2014) —  and thus violate students’ privacy and erode trust within school communities. Indeed, educators we’ve worked with in our research on dataveillance technologies often report feeling uneasy about both the volume of data harvested and the ways this information can position students as potential risks to be managed rather than learners to care for and support.