Kathy Ensor, Lance Waller, and Brittany Segundo are leading a major National Academies consensus study called Frontiers of Statistics in Science and Engineering: 2035 and Beyond. It aims to assess the current state of the statistical sciences and find emerging opportunities for the discipline and its stakeholders as they look ahead to 2035 and beyond. In a recent episode of Practical Significance, Ensor, Waller, and Segundo joined cohosts Donna LaLonde and Ron Wasserstein to share their perspectives on this important study and its implications for the future of statistics.
Donna LaLonde: Tell us about your day jobs.
Lance Waller: I’m a professor in the department of biostatistics and bioinformatics in the Rollins School of Public Health at Emory University.
Kathy Ensor: I am the Noah G. Harding Professor of Statistics at Rice University, where I’ve spent my entire career. I was the 2022 president of the ASA and continue to champion the ASA in our community.
Donna LaLonde: Brittany, our colleague from the National Academies, is supporting the Frontiers of Statistics in Science and Engineering: 2035 and Beyond study co-chaired by Lance and Kathy. Brittany, we’d like to hear about you and your work with the Committee on Applied and Theoretical Statistics.
Brittany Segundo: I’ll start with a bit of context about the National Academies of Sciences, Engineering, and Medicine and what we do. We were founded in 1863 during the Civil War as a nongovernmental, nonpartisan entity to advise the government on scientific, engineering, and medical matters of national inquiry.
We’re broadly known for our honorific societies, but we also have a robust research arm called the National Research Council. That’s where I work, under the NRC on the Board of Mathematical Sciences and Analytics. In this role, I direct CATS, where we translate state-of-the-art statistics and analytics research into policy questions across a variety of domains.
One thing the NRC is really known for is its consensus studies. These reports respond to a statement of task or charge and provide key messaging and, in some cases, recommendations to sponsors and federal agencies. The project we’re working on right now—the Frontiers of Statistics—is one of these consensus studies.
Our hope is that the Frontiers of Statistics study becomes a landmark assessment of the field. Other disciplines routinely take advantage of these opportunities to assess their domains—we see this often in astronomy and physics at the National Academies—and we’re very excited to do something similar for statistics.
We broached the idea of a forward-looking assessment of statistical sciences with our federal contacts, and they showed real appetite for a study of this magnitude and scope. We’re very fortunate to be sponsored by a diverse array of offices across the National Security Agency, National Institutes of Health, and National Science Foundation. At NSF, we’re funded by six directorates: Mathematical and Physical Sciences; Computer and Information Science and Engineering; Biological Sciences; Engineering; Geosciences; and STEM Education.
This is truly a cross-disciplinary effort, looking at the field of statistics and its broad impact.
Ron Wasserstein: Lance and Kathy, why were you willing to take on this effort, and why is it so important that you’re willing to set aside other things to lead this project?
Lance Waller: Kathy and I are big advocates for the field of statistics in all its many forms and interactions. We’ve both served on the CATS committee, and I’m fortunate to be co-chair now with Liz Stuart at Johns Hopkins.
CATS is one of two statistics committees in the National Academies—the other is the Committee on National Statistics, which focuses on federal statistical infrastructure such as the Census Bureau and Bureau of Labor Statistics. CATS is the rest—data collected for many reasons, from meteorologic data to fisheries data, and these bring challenging questions that are really fun to think about. I’ve enjoyed being part of CATS, having federal groups reach out with questions, and then working with 12–15 colleagues to make recommendations on what they might try next.
When this opportunity came up in conversations with David Manderscheid at NSF, the division director for the Division of Mathematical Sciences, we were excited to identify individuals with a broad range of experience and interests across statistics. It just seemed like it would be a lot of fun. And second, the academy staff, like Brittany, make it a delight to work in this kind of environment—sitting together to think about where our field is moving and where it’s going.
Kathy Ensor: Lance and Brittany brought the study to fruition and then invited me to join, which was a wonderful opportunity. So why do this? Our day-to-day lives are changing rapidly, but at the center of it all is the strength of our discipline and what we bring to the world every day.
This study gives us the chance to focus the conversation on science, technology, and the frontiers; what’s happening; and what’s new on the horizon. How much fun to think about that with such an amazing group of people! That is the “why.” It is a lot of work, no doubt, and we still have much ahead of us.
Our profession is always forward-looking. We are a forward-looking society and group of scientists, and that is why we continue to evolve as society’s needs for us evolve.
Donna LaLonde: Brittany, would you give us a behind-the-scenes look at the Frontiers of Statistics in Science and Engineering: 2035 and Beyond?
Brittany Segundo: The first step of any activity at the National Academies is to assemble a committee, and it is the biggest privilege—and my favorite part of the job—to work with people like Kathy and Lance and bring together diverse leaders in the field.
Once the committee is assembled, we begin an aggressive information-gathering campaign to understand the scope and breadth of the field. For this study, we’ve hosted public panels on topics including causal inference, trends in statistical journals, privacy, and precision agriculture. As we write the report, a second wave of information gathering often emerges—filling gaps and reaching out to experts who can help guide us. These conversations are also open to the public.
Uniquely for this study, we’ve issued an open call for perspectives. Whether you are a student, dean, professor, or practitioner, we want to hear what you are working on and how you are pushing the frontiers. This isn’t about my thoughts or Kathy’s or Lance’s; it’s about representing what is happening across the field and capturing the trends that show where statistics is headed.
Ron Wasserstein: Kathy and Lance, fast forward to 2035. If there was one statistical development or application you hope will have been realized, what would that be and why?
Kathy Ensor: I’ll answer more broadly, focusing on our profession and how we evolve. It’s a good time to remind the audience of the ASA’s vision: a world that relies on data and statistical thinking to drive discovery and inform decisions. In other words, statistics drives evidence-based decision-making. We live this vision daily, and I believe it will continue in 2035.
Since entering this remarkable field at the intersection of science, engineering, and data, I’ve witnessed tremendous advances: DNA sequencing revolutionizing genomics and personalized medicine; breakthroughs in neuroscience reshaping our understanding of the brain; and urgent work on weather, resilience, and sustainability. At the same time, smart connected cities are showing a commitment to technologies that improve the human condition and foster equitable, empowered communities.
Looking ahead, I hope to see statistics and statisticians remain at the forefront of these transformations, advancing a continually evolving discipline essential for discovery, decision-making, and design. We must engage collaboratively to help science, technology, and society understand the world more deeply and navigate it more wisely.
Lance Waller: If I summarize what interests me most about the field of statistics, it is that we combine and evaluate the best data available to get the best answers to the questions at hand. Some questions are motivated by new ways to measure things; we must keep up with technology and hopefully influence it. Other questions are different from what we’re used to, as we’ve seen with AI. There’s a lot of movement.
What I like about our field is we try to get the best answers we can while maintaining a healthy respect for uncertainty. We have, I hope, professional humility, recognizing that we can make mistakes but working to minimize and quantify them, understand them, and do better than before.
There is a lot of room for optimization—I was trained in optimization, myself—but if we wait for the best answer, sometimes we short circuit getting a good answer for a good outcome. In the end, our field brings reality to how we implement data-based, evidence-based decisions.
Ron Wasserstein: Lance and Kathy, what are the opportunities and challenges for statistics regarding artificial intelligence?
Lance Waller: This is an ongoing discussion in our field, and all of us are involved in it. There are excellent working groups and conversations in our community. Our committee knows we can’t pull all of that together, but we want to capture the general sense. As you said, it’s moving very fast—driven by statistical ideas and computing ideas—and we’re trying to reflect some of the visionary thinking. It’s not the only frontier, but it is certainly changing the way we work.
I have a metaphor that not everyone likes, but I’ll share. If you think about the family tree of statistical developments, its roots are in probability and the probabilities of making mistakes. We grew branches from assumptions such as collecting the best data you can—data was expensive, so you wanted to get the most out of the least. That gave us a beautiful set of branches, like the design of experiments, fundamental to our field. But some assumptions are different now.
You can pull data together from many sources. It may not be exactly what you want, and more data is not always better, but sometimes it’s better than ignoring it. Regrowing the tree with new assumptions may create new branches alongside the old.
The roots remain strong, and understanding uncertainty is essential. There will be new versions of that, which will drive where we go. It will happen through collaboration, not only within statistics, and many AI developments have already grown out of statistics and its important components.
Kathy Ensor: Sure, let me jump in. As part of our information gathering, we held a panel of leading statisticians from industry. It was amazing, and the message was clear: Not only are research statisticians innovating and helping create this new AI world, but we also need to become expert users of it, integrating these tools into our day-to-day workflow. From this leading group of industry statisticians, there was no middle ground. That was very clear.
I take that back to my own work. I still try to be an active researcher, asking both practical and visionary questions. I’m genuinely excited that AI tools let me code faster, answer questions more quickly, and, in turn, ask richer and better questions. It’s a fascinating time to be a statistician. Sometimes I wish I were 20 again to fully embrace it. Younger statisticians will ask new questions or even the same questions in new ways.
That’s the excitement: We can answer bigger questions or the same questions better. Not only are we advancing artificial intelligence with our research, but using it allows us to do more and to do it better.
Ron Wasserstein: What impact do you hope the study will have?
Brittany Segundo: I hope the study demonstrates the significance of statistics to science and engineering, which is the title of the study. But statistics also plays such a crucial role in enabling technology and in collaborations with other domains. And it is key for American competitiveness. I hope this study continues to demonstrate the brilliant work statisticians are doing and how vital that work is to our success.
Kathy Ensor: Our profession is at the forefront of science and technology advances. And this report should provide insight and lay a foundation for funding for the profession and why investment in statistics is so critical at this juncture. And I’m optimistic. These are very dynamic times for the whole science enterprise. But I do feel like our discipline and the thoughts we’re bringing forward in this report will really help advance society through the contributions our discipline brings. And so, we’re trying to lay out a funding foundation for the next few years, for sure.
Lance Waller: We have a resilient discipline, and my hope is the study shows we can rise to challenges by pushing things already on the edge and filling in gaps where data may not grow so fast. It encourages using what we have in better ways and not only pushing frontiers outward but making the work better.
The toolbox of statistics is something constantly being built with new versions to answer longstanding questions better and new questions we’ve never thought of. By defining how to use new types of data and combining things, the study seeks to consistently define what’s next for us. Even though defining frontiers is hard, we believe we have the tools to make our job exciting, rise to challenges, and further expand our field.
Leave a Reply