Quantitative Analysis and Ethics
Biggs, Max;Freeman...
Quantitative Analysis and Ethics
Biggs, Max; Freeman, Rupert; Parmar, Bidhan L.
E-0527 | Published August 7, 2026 | 7 Pages Technical Note
Collection: Darden School of Business
Product Details
This technical note challenges the assumption that quantitative analysis is objective and value neutral. It provides students with a framework for examining the ethical choices embedded throughout the analytical process, including defining goals, selecting and measuring data, choosing and deploying models, and evaluating outcomes. Through examples involving AI and other quantitative tools, students consider issues such as bias, privacy and consent, explainability, human oversight, and algorithmic fairness. The note helps students recognize how analytical choices distribute benefits and harms and prepares them to make more intentional, responsible decisions when developing or using quantitative models. It is an excellent companion note for the case "Bias in AI Advertising Models" (UVA-E-0518).
After reading and discussing this technical note, students should be able to do the following: - Explain why quantitative analysis is not necessarily objective or value neutral and identify the ethical choices embedded in analytical decision-making. - Evaluate how the selection of goals, measures, and proxy variables can shape model results and create unintended consequences. - Identify ethical concerns related to data selection, including privacy, informed consent, representativeness, and the use of variables correlated with protected characteristics. - Assess trade-offs involved in model selection and deployment, including accuracy, explainability, uncertainty, human oversight, and revisability. - Analyze how quantitative models distribute benefits and harms across individuals and groups and compare alternative approaches to algorithmic fairness. - Apply ethical reasoning to decisions about the design, implementation, monitoring, and revision of quantitative models.