​​​Academic Data Analytics


ADA campus photo grid
Data-Driven Decision-Making

The Academic Data Analytics (ADA) unit is designed to provide data-driven decision-making into the hands of stakeholders at the University of Oregon, while building a body of research that can be generalized to other higher education institutions. The unit emphasizes research on student and faculty success in a variety of research areas and technical methods.

We envision the unit as one that pushes the boundaries of what is possible in higher education research, through the exploration of machine learning and other advanced methods of analysis. The goal is to use these techniques to more efficiently distribute resources (in predicting student success), determining what the main issues are (in using text analytics), and to uncover any issues that are a hindrance to student success (in focusing on a specific subgroup of students or their paths to graduation).

Areas of Focus:

The ADA unit aims to have research products that are actionable, rapidly produced, and in formats that are easy to use for university stakeholders and higher education researchers alike.

The unit will focus on four main technical areas: 


for the purposes of fairly and equitably predicting academic outcomes

2. TEXT analytics

to determine sentiment and themes in written feedback


To measure the effectiveness of programs and policies


to help uncover barriers to student and faculty success

Our Research

Combating inequity is a core priority for ADA, and the unit will have a strong focus on issues surrounding diversity, equity, and inclusion (DEI). All research will begin to attend to these areas through the inclusion of variables such as gender, race, ethnicity, first generation status, and Pell eligibility status. Accordingly, all efforts to predict student success will remain closely focused on bias, fairness, and equity throughout the research process, and researchers will employ state-of-the-art algorithms to quantify and mitigate bias in all models and predictions.

Wherever feasible, predictive models will be optimized to actively combat historical inequities, not just disregard them. Predictions will be disseminated only once they are deemed sufficiently free from all forms of algorithmic bias and unfairness. To support its diverse research initiatives, ADA will seek input from subject matter experts to ensure that staff thoroughly understand issues of bias and equity as they manifest in each area. Researchers will document the process of quantifying and mitigating bias in each analysis and publish this insight as a key part of study methodology. In so doing, ADA hopes that other researchers and practitioners may benefit from its work in this space, advancing equity not only at the University of Oregon, but at other higher education institutions.

What are the characteristics of transfer students, and how efficiently do their credits transfer? How accurately can we predict their graduation outcomes?

What kinds of topics do students write about in their student experience surveys? Has the change to a more structured survey format reduced potential biases in responses?

How accurately can we predict second-term persistence for students who enter the university? How can we use techniques to help mitigate potential biases in our predictions?

How long does it take junior tenure-track faculty members to achieve tenure? How does this length of time vary by faculty characteristics?


For Students:

Are you interested in using data to help improve the academic experience for you and your fellow students? If so, please contact us for internship and other partnership opportunities!