"This is not a data problem": Algorithms and Power in Public Higher Education in Canada (2403.13969v2)

Abstract: Algorithmic decision-making is increasingly being adopted across public higher education. The expansion of data-driven practices by post-secondary institutions has occurred in parallel with the adoption of New Public Management approaches by neoliberal administrations. In this study, we conduct a qualitative analysis of an in-depth ethnographic case study of data and algorithms in use at a public college in Ontario, Canada. We identify the data, algorithms, and outcomes in use at the college. We assess how the college's processes and relationships support those outcomes and the different stakeholders' perceptions of the college's data-driven systems. In addition, we find that the growing reliance on algorithmic decisions leads to increased student surveillance, exacerbation of existing inequities, and the automation of the faculty-student relationship. Finally, we identify a cycle of increased institutional power perpetuated by algorithmic decision-making, and driven by a push towards financial sustainability.

• The paper finds that algorithmic decision-making driven by financial sustainability intensifies institutional power and heightens student surveillance.
• The study employs qualitative analysis of enrollment prediction and machine learning models, revealing concerns over bias and transparency.
• The paper introduces the ASP-HEI Cycle to illustrate how fiscal pressures and data reliance create a self-reinforcing dynamic affecting equity and academic relationships.

"This is not a data problem": Examining the Impact of Algorithms and Power in Canadian Public Higher Education

Introduction to the Study

A recent qualitative analysis conducted on data and algorithm usage within a public college in Ontario, Canada provides valuable insights into how post-secondary institutions are integrating data-driven practices. This significant shift towards reliance on algorithmic decision-making is largely influenced by internal and external pressures towards financial sustainability. The paper contends that such reliance leads to heightened student surveillance, magnifies existing inequalities, and automates elements of the faculty-student relationship. Notably, it highlights a vicious cycle where algorithmic decision-making perpetuates increased institutional power, driven by the overarching goal of financial sustainability.

Data-Driven Practices and Algorithmic Decision-Making

The paper uncovers that data-driven decision-making at the college aims primarily at resource allocation, revenue forecasting, and potentially commercialization of algorithms. This aligns with government policies and the neoliberal trend towards greater accountability and efficiency in higher education. The adoption of these practices responds to financial challenges but raises ethical concerns related to student privacy, equity, and the transformation of educational values.

Algorithms at Work

The analysis details the college's foray into algorithmic decision-making with three primary models focused on enroLLMent prediction and assessing student success. The use of automated machine learning platforms, such as DataRobot, simplifies model building but limits transparency and control over the machine learning process. Concerns arise regarding model accuracy, bias mitigation, and the absence of clear ownership and responsibility for algorithmic decisions.

Impact on Stakeholder Relationships

The integration of algorithms and data-driven systems redefines relationships within the institution. This shift particularly affects the academic and student experience divisions, with notable discrepancies in perceptions of the college's capabilities in handling data and algorithms. As decision-making increasingly relies on data, divisions with direct data access, particularly those linked to business and administration, gain undue influence over institutional priorities.

Consequences for Students

The emphasis on data and algorithms profoundly affects students through increased surveillance, reinforced inequities, and a reduced human element in education. The paper discusses how the need for extensive data for algorithm training leads to pervasive student monitoring, raising ethical concerns. The opaque nature of algorithmic scoring and categorization can exacerbate existing biases and unfairly impact students from marginalized groups. Moreover, the automation of identifying and supporting at-risk students traditionally a faculty role signals a shift towards viewing education through a business lens, potentially undermining the quality of student support and the faculty-student relationship.

The ASP-HEI Cycle: A New Framework for Understanding Power Dynamics

The paper introduces the "ASP-HEI Cycle" as a conceptual framework to illustrate how algorithms, student data, and institutional power interact within higher education institutions. This cycle reveals how financial sustainability goals dictate the adoption of data-driven practices, which in turn amplify institutional control at the expense of student autonomy and equity. The framework underscores the cyclic nature of these dynamics, where financial pressures drive further reliance on algorithmic practices, perpetuating institutional power and potentially distancing higher education from its core democratic and inclusive values.

Implications and Future Directions

This research underscores the need for a careful examination of how higher education institutions adopt algorithmic decision-making. The findings call for a more human-centered approach in the design and implementation of algorithms, emphasizing transparency, equity, and stakeholder participation. There is a clear necessity to balance efficiency and financial considerations with the educational mission's qualitative aspects. Future research should further explore the consequences of these practices on the educational experience and how institutions can mitigate potential harms while embracing the opportunities presented by technological advancements.

Concluding Thoughts

The paper sounds a cautionary note on the growing reliance on data and algorithms within higher education, highlighting the complex interplay between financial sustainability, institutional power, and the risk to democratic educational values. It presents a critical perspective that invites further dialogue and reflection among policymakers, educators, and researchers on navigating the challenges of integrating technology in education while preserving its core values of inclusivity, fairness, and human connection.