Equity and inclusion are at the core of the ADA’s work. Our projects seek to uplift student voices, connect students with support, and identify biases in faculty evaluation materials and institutional practices that aren’t serving the most vulnerable. We apply an equity lens throughout the research process, including attributes like race, gender, and financial aid status when appropriate.  

ADA’s text analytics work uses both state-of-the-art algorithms, which are efficient and objective, and input from human subject matter experts to ensure the results are presented accurately and fairly. When disseminating our findings to students, instructors and staff impacted by those findings, we work to identify the limitations of our analyses and seek input to make sure we represent the perspectives of our diverse community. 

Whenever feasible, predictive models are optimized to actively combat historical inequities, not just disregard them. Predictions are only shared once they are deemed free from algorithmic bias and unfairness. To support diverse predictive modeling initiatives, we seek input from not just subject matter experts but also those with lived experience to understand the issues of equity as they manifest. Insight into the process of quantifying and mitigating bias is included in the methodology of each analysis.