Data4Good Goes Back to School: Capstone Projects for the Greater Good

With a PhD in statistical astrophysics, David Corliss is lead, Industrial Business Analytics, and manager, Data Science Center of Excellence, Stellantis. He serves on the steering committee for the Conference on Statistical Practice and is the founder of Peace-Work, a volunteer cooperative of statisticians and data scientists providing analytic support for charitable groups and applying statistical methods in issue-driven advocacy.

Sarah Handau. Maddie Young. Haelie Egbert. Edina Kovacs. Eli Baker. Five great students from different backgrounds and unique skills, all working with one goal in mind: identify risk factors and hotspots for human trafficking in Appalachian Ohio. What brought this amazing D4G team together? A capstone project.

With the fall semester starting in just a few weeks, it’s an exciting time for students and professors as they plan for the new year. We’re going to see some changes with the COVID pandemic trending downward in many countries and we move into a “new normal.” And some changes we saw during the pandemic will be continue. New technology, practices, and connections developed to support remote study will form the backbone of the emerging post-COVID environment in education, government, and private sector research.

Opportunities for education projects have become more global, with students working as easily with coral reefs in Australia or STEM education in Africa as an organization in their home town. The work from home environment has also led to more data and other resources becoming available to the public. The possibilities for school projects—including term papers, class projects, and other educational experiences—are endless. The same goes for student groups.

This month, we’re going to look at capstone projects in detail. These projects generally involve working with a team on a long-term task to produce something complex, such as a new statistical application with associated data, modeling, and visualization. The projects normally take one or more semesters and serve as the final project for a program. Capstone projects many times allow for a “real world” experience by providing a project created by a team with diverse skills and backgrounds and leveraging raw, uncurated data to produce a functioning application on a managed timeline. Many capstone programs give students a choice of several potential projects, often including one or more in community service. This makes capstone programs a great opportunity to get started in Data for Good when starting one’s career.

Of course, capstone projects are more than just the students. Professors, private sector partners, and more are needed to make it all happen. Here are some qualities I have found that make for good capstone projects:

• Partnership with industry, government, or some other organization outside the university. The project must complete an analysis or statistical product needed by this outside group, not the professors.
• At the same time, the professor needs to completely own the academic requirements and assessment: This is a class project, not an internship.
• Working with a team is better than working individually, so different people, skills, resources, and backgrounds can support the project. Diversity is so much more than fair play. As with all justice issues, practicing it actively makes things work better.
• While working as a team, however—and this is a key point that is sometimes missed—each person working on a capstone project needs to have ownership of a specific part of it. They need to be responsible for this task and take credit for it going forward. Those who set up and manage student projects need to create opportunities to show ownership and leadership—both for the success of each part of the project and as something students can show future employers.
• Clear goals. The parameters of a successful project need to be described early on and in writing, with each person agreeing to work together to achieve them.
• In addition to goals, good projects have stretch objectives for each person. These are not included in the written goals (i.e., project requirements, without which the project is not a success). The university/outside sector partnership needs to create opportunities for each team member to excel, going above and beyond the required to do something extraordinary.

Capstone projects offer many opportunities for statisticians in academia, government, NGOs, and private industry. I was connected to the human trafficking project for Eyes Up Appalachia, a community foundation in Ohio, through volunteer D4G work in human trafficking, but it also helped me make connections at a major research university from where my employer recruits data scientists. My role was mentor—serving as a subject-matter expert; recommending contacts, data sources, and methods; teaching best practices; and reviewing and encouraging work. It was such a rewarding experience for me to work on this project! I strongly encourage looking at capstone projects at a university you support to check out the opportunities to get involved.

I would like to give a Stats4Good shout-out to Evan Boyle and Holly Bossart, recipients of the inaugural ASA Pride Scholarship. The scholarship was established to raise awareness for and support the success of LGBTQ+ statisticians and data scientists and allies.