MDI Scholars Program Partners With Ethics Lab

Written by Carrie McDonald, MDI Journalism Intern

This academic year, the Massive Data Institute (MDI) partnered with Georgetown’s Ethics Lab to incorporate training for the MDI Scholars on integrating ethical questions and considerations within their research process. 

MDI Scholars engage in interdisciplinary research to influence public policy outcomes, often focusing on improving equity for marginalized communities disproportionately burdened by various risks, harms, and disadvantages. This purview necessitates centering ethical considerations throughout every stage of the research process. Research that involves people or will be used by people must be designed to ensure that it benefits society and minimizes biases. 

“Ethical considerations shouldn’t be bolted on to the side but rather be included into the process of creation when doing research and creating systems,” said Dr. Kate Wojtkiewicz, a postdoctoral fellow at Ethics Lab. 

For example, some of these ethical obligations include gaining informed consent, addressing algorithmic bias, and implementing best practices for data protection to mitigate potential harm.

“Data science is enormously technically complicated, yet saturated with ethical considerations,” Ethics Lab Director Dr. Maggie Little said. “Our hope is that the [MDI] Scholars can now see data science projects through an ethical lens, and see how technical design choices can help or hinder those ethics.”

It is critical to adhere to high ethical standards to protect the safety, dignity, and rights of study participants; doing so will also improve the quality and usability of a study’s results. For example, machine learning systems are trained with existing data, but when these data fail to represent marginalized communities, algorithms produce outcomes that consistently put certain demographics at a disadvantage. Considerations like this are critical as artificial intelligence (AI) is becoming more prevalent in decision-making. 

“Many data science projects involve using data that are about people – their preferences, their opinions, their behaviors,” MDI Director Dr. Lisa Singh said. “And there are times when we need to pause to think about how and whether we should be conducting the analysis we are doing. Are we helping the world? Or are we violating privacy expectations or causing other harm to people and communities?”

During some of the monthly MDI Scholars meetings, the Ethics Lab team framed conversations for MDI Scholars to explore these questions. Students reviewed best practices, generated ethical questions specific to their MDI research projects, and discussed specific case studies involving salient ethical debates. The MDI Scholars included an “ethical considerations” section on their research showcase posters to share their reflections on how these ethical considerations apply to their own projects. 

“The sessions helped me think of more ethical implications of my research beyond the scope of the specific work I’m doing right now,” MDI Scholar Alicia Gopal (CAS ’25) said. “Ethical considerations are crucial to our research as we aim to use tax data to analyze the impact of certain government and nonprofit programs on wages. We need to consider things like privacy, who files taxes, who participates in these programs, and how this data can be misused.”

Another MDI Scholar, Bernardo Medeiros (CAS ’24), explained how ethical considerations were vital in his project focused on blending data to improve predictions on forced migration patterns. 

“One question we’ve been thinking a lot about is how the models we’re building might be misused by bad actors, and what we can do to make sure our predictive models stay in good hands,” Medeiros said. “These considerations are important because the people whose movement we’re predicting are often in vulnerable situations.”

MDI and Ethics Lab plan to continue their collaboration in the summer and find ways to further connect Ethics Lab’s expertise to MDI’s research workflows.

Ethics Lab
MDI Scholars