Institutions considering allowing students to opt out of data sharing should consider very carefully whether this can create or further amplify the inequalities that learners face. Known as consent bias, the problem is that students who choose not to participate (or decide not to participate) may differ systematically, so the conclusions or actions taken based on the data will unfairly bias l one of the student groups.
Currently, students generally don’t feel like they can control access to the data their college collects about them. According to the Student Voice survey conducted this summer by Inside higher education and College Pulse, with support from Kaplan, only 22% of students thought they could restrict access to this, while 9% didn’t and the vast majority – 69% – weren’t sure. .
But rushing to provide withdrawal mechanisms can actually harm students, not protect them. I am convinced that this is a real problem that we are facing.
We emailed 4,000 University of Michigan students in 2019 to see if they would opt out (or participate) in data sharing to support learning analytics tools. The answers, detailed in this recent International Journal of Artificial Intelligence in Education item, demonstrated a clear consent bias: women were more likely to respond than men, and students who identified as black were less likely to respond, while students who identified as white were more likely to respond . And these response differences added to the already skewed demographics of the student body.
The issue of confidence that the institution will use the data responsibly is the main factor guiding students’ decision to consent.
Women trusted the institution or the instructor significantly more, although they were also more concerned with the practice of collecting personal data. Black students, however, expressed less confidence in the institution. This indicates a clear opportunity for institutions to build bridges of shared understanding and transparency with students around data collection, and supports initiatives like the UM ViziBlue dashboard, which gives students information about the types of data that is collected about them and how it is used and shared.
This concern for exclusion goes beyond a student’s individual success. Machine learning methods are increasingly used to power educational support tools, and the data on which these algorithms are based affects how they work in students. If a student refuses to share data or refuses to enroll, we should expect the support tools to become less accurate for similar students at the institution.
Given the consent bias we are seeing, this should give institutions real concern that individual choice alone will have a disproportionately negative impact on groups of learners who may be historically disadvantaged.
Thus, the choice to opt out of data sharing must be weighed against the obligation of the student – and the institution – to support the broader learning environment. In light of this, the ethical choice is to proactively engage students through data to show them how it is being used to support their learning and that of others.
Indeed, it is a teaching opportunity for us as educators and sits at the crossroads of confidentiality, data mastery and civic engagement. Instead of encouraging withdrawal, we should educate about the positive impacts of data through transparency, while listening to the concerns and ideas students bring to the discussion. Not only will this allow us to support the equitable impact of data-driven educational materials, but it will also allow us to strengthen our relationship with students by building trust in the institution and our educational mission.