Data analytics is spreading fast in higher education. Vendors and proponents have spent a decade advertising the benefits, from early warning systems that identify students in need to adaptive learning that can improve outcomes. Yet, COVID-19 has opened more eyes to the risks and downsides, from deeply invasive remote proctoring services to reckoning with unaccountable third-party tools hastily deployed during the pandemic.

As the urgency of the crisis begins to recede, ethical and strategic questions about data analytics must come to the forefront. Can artificial intelligence be deployed without further marginalizing underserved populations? Can institutions remain open and accountable to their missions while key decisions are made by black box algorithms? Can institutions benefit from the use of data analytics without ceding too much control to vendors? Can open education survive on closed infrastructure?

At the center of many of these questions are textbook publishers, who have used the pandemic to accelerate their transition away from printed books to models more like Netflix or Facebook. The growing adoption of subscription-based digital textbooks creates a pipeline for valuable data to flow into increasingly advanced analytics products, giving publishers insights into campus activities and students’ daily lives. While privacy laws afford some protection to individuals in some jurisdictions, they are woefully inadequate in others and many institutions have only begun to grapple with the broader strategic implications of outsourcing their knowledge infrastructure.

This session will explore the intersection of data analytics and open education practice, drawing mostly from a North American context, but encouraging participants to enrich the conversation with their own perspectives.

Specifically, the conversation will consider open education in the context of a solutions framework for regaining and maintaining community control over education infrastructure published by SPARC (Aspesi 2019). The framework proposes action at three levels:
1. immediate actions that can mitigate the risks and harms of data analytics;
2. thinking strategically about the trade-offs, such as balancing human and AI decision-making; and
3. collective actions, such as how to build, acquire and govern open infrastructure.