Data science programs have grown substantially at universities over the past few years.  How much room for growth is left? 

Data science is a new paradigm for research and discovery; nearly every field is rapidly evolving to incorporate data science approaches as volumes of data are collected from and about the world around us. From that perspective, any institution that does research or teaches current approaches in nearly any field will need to train their students in the fundamentals of working with data. We aren’t even close to that yet. There are still major research universities who are just starting to scope what their data science strategy will be – whether to develop a research and consulting Institute or Center that serves the campus as a whole, or focus on curriculum building for degrees within a department or school of data science. There are notable successes with both approaches. Smaller schools, including community colleges, are also figuring out where they fit in the data science education spectrum. Training for jobs like “data analyst” is in high demand from adult learners looking to switch careers. We are also starting to see data science courses being taught as early as high school and even middle school. Yet in some schools, students are graduating high school without any math beyond Algebra 1. How can colleges and universities level-set their course requirements to make data science accessible and equitable to the broad spectrum of backgrounds of their incoming students? There is much work to be done!

You founded the Academic Data Science Alliance just prior to the pandemic.  What kept you going with the ADSA during those tough early days?

We were fortunate that we had already secured 3 years of funding from two foundations, the Alfred P. Sloan Foundation and the Gordon and Betty Moore Foundation, prior to the pandemic. Both foundations were incredibly generous and understanding of the need to shift spending priorities in light of the pandemic, and even renewed their funding for ADSA this year. We had also established from the outset as a remote organization – we have no physical headquarters. Our staff were already holding all our meetings by Zoom prior to the pandemic. That said, we had built our strategy around relationship building at events and workshops. Pivoting to virtual meant learning the most effective ways to convene and engage with our community in the virtual environment. And that was a moving target. Our first annual events during the pandemic followed the “just move it all to Zoom” approach where we held two consecutive meetings over 5 full days: our Leadership Summit and our Annual Meeting. But by the second year of the pandemic, we recognized that format was no longer sustainable. Zoom burnout coupled with demands on time from personal lives meant we had to meet people where they were and that meant reducing expectations of time. No one could dedicate three straight days to an online conference – not with family and life obligations. And it was nearly impossible to ignore all the other work demands when everyone is always online and no longer out-of-office for travel. In that year, we created a more distributed model of weekly events hosted from Fall 2021 to Spring 2022. Even with this model, participation waned. I think many of us felt the constant drum of work at a much greater intensity. For me personally, what kept me going were all the champions who believed in us from the beginning and supported us with their time and investment throughout those tough years. And I am so grateful to my ADSA team who helped me weather the storm.

How does academic data science differ from corporate data science?

Fundamentally, academia is about training the next generation and providing a home for basic research (that is, research without an immediate application or product). For data science specifically, this means ensuring the students who come through the higher education system have more than the technical capabilities needed for a data science job, but also the ethical awareness and training to take responsibility for the tools and technologies they build. Students should be able to critically assess a data science project for potential negative impacts on individuals, communities, and society and know how to mitigate the potential for harm. Corporations by necessity are driven by a need for profit. This doesn’t mean they can’t promote ethical technologies and data use and provide professional development training to support these, but training is not their end goal. Academic and corporate data science are also deeply intertwined by supply (trained students; basic foundational research) and demand (workforce and technological advancement needs). And this relationship requires a feedback mechanism: academia should be responsive to the evolving needs of the sectors who will hire their students. I also want to note that the Academic Data Science Alliance isn’t only for people in academia. We welcome anyone who shares the values traditionally associated with academia, like collaboration and working openly, and who want to maintain connections to academic colleagues.

Does data science need its own professional society, such as ACM for computer science or the ASA for statistics?

Yes! Professional societies act as a compass for a field. They organize, convene, and support scholars in a particular discipline to stay up-to-date on advancements in the field, help define and refine a field as it grows and evolves, and provide professional networking opportunities. They can also advocate for changes in policies that would benefits their members. ADSA is already filling this role. We provide a welcoming and inclusive community for sharing knowledge, collaborating on resources, and building networks of professional connections. We promote and support responsible data science and advocate for rewarding career paths for data scientists in academia.

Operationally, ADSA is a professional society: we host multiple annual meetings, including our invitation-only Data Science Leadership Summit, our ADSA general meeting “Future Leaders for Responsible Data Science,” and workshops including co-hosting ImageXD with the Berkeley Institute for Data Science and a Career Support workshop with the U.S. Research Software Engineers Association. We are a fiscally-sponsored project of the non-profit Community Initiatives which means we can take in funding from grants or donations, we have an internal Board that helps us stay accountable to our Mission, Vision, and Values, and we offer both individual and institutional paid memberships with nice benefits. We are also currently exploring the option of an external advisory board with members from other sectors outside of academia including industry, government, philanthropy, and other NGO’s.

Outside of the professional sphere, what can you share about yourself that your colleagues might be surprised to learn – any unique hobbies or stories?

I’m delighted you asked this question. I often use “fun facts” questions during rounds of introductions at ADSA meetings, virtual and in-person. There is so much more to a person than their work life. I’ve met an ADSA community member who is a sommelier, another who speaks Kiswahili fluently, another who hosts open music mic nights, and another was the middle school oboe teacher for Jennifer Lawrence, star of the Hunger Games movies. As for me, most people know I love skiing. I once skied 13 different mountains in one season (3 in one day) as part of a friendly competition with my sister. I’ve also cross-country skied in Antarctica, I’ve experienced 2 boat sinkings (not my fault!), and my first name comes from the opera Carmen (my parents are avid classical music listeners). For the stories, you will have to ask me at an ADSA meeting. 🙂