Q&A with Sam Wang

Y. Samuel Wang will join the Department of Statistics and Data Sciences during the 2021-22 academic year. He comes to Cornell from the University of Chicago’s Booth School of Business, where he served as a principal researcher under Mladen Kolar. Wang completed his Ph.D. in statistics at the University of Washington in 2018 after receiving his undergraduate degree in both applied math and economics at Rice University.

What is your research about?

I have broad interests across statistics, machine learning and data science, but much of my work is in the subfield of "graphical models." In this area, researchers consider how each variable in a complex system might be dependent or independent of the other variables. I primarily work in theory and methods, so I'm not tied to a specific application area. However, as examples, the methods I work on could be applied to functional magnetic resonance imaging (fMRI) data to discover how different regions of the brain interact, or to financial data to see how the performance of some stocks affect the performance of other stocks, or to systems biology data to see how certain proteins might regulate other proteins. I also enjoy thinking about statistics and data science applied to social science questions. For instance, one of my current projects seeks to measure gender bias in co-authorship team formation.

How did you become interested in your field?

One of the things I enjoy about being a statistician is the blending of "art" and "science." When proposing a new statistical method, you typically also provide mathematical proofs that guarantee performance of that method. For instance, these guarantees typically say things like: if the data you are analyzing has properties X, Y and Z, then the procedure will "work well" as the size of your data grows. And so, there is one side of the work that is very rigorous and "scientific."

However, even when comparing different methods with similar theoretical guarantees, the practical performance can vary quite a bit depending on whether or not the data satisfies the assumptions you have made. Thus, there is also an "art" to statistics which involves thinking about what assumptions are reasonable, how to best represent real processes with mathematical models, and how to communicate complex results in an interpretable manner. The extra bonus is that you often get to work collaboratively and apply these tools across a wide variety of fields and disciplines. As John Tukey put it, "The best thing about being a statistician is that you get to play in everyone's backyard."

What impact do you hope your research will have?

I aim to develop methods that are both practically useful to researchers and rigorously justified. Ultimately, I hope my work empowers other researchers to better analyze and glean insights from their data.

What attracted you to the ILR School?

There are so many great things about the Cornell Statistics and Data Science Department, but if I had to narrow it down, I would say the department's broad view of statistics and supportive colleagues. Researchers in the department are doing cutting edge work in statistical theory and methodology, but they also celebrate interdisciplinary collaborations and apply statistical tools to substantive scientific questions. In addition, the department has a reputation for being quite collegial and welcoming. Even though the entire job search was done virtually, it was clear that folks enjoyed being at the department and working together.

What are you most excited for about your time at ILR?

The people! I always enjoy hearing about problems that researchers in other fields are trying to solve and thinking about how statistical tools might contribute towards that end. So, I'm excited to be at ILR where researchers are working on a broad range of topics.

If you could share one piece of advice with your students, what would it be?

In his lecture at the 2013 Joint Math Meetings titled "The Lesson of Grace in Teaching," Francis Su, a professor of math at Harvey Mudd said that the beginning of the biggest life lesson he ever learned is, "Your accomplishments are NOT what make you a worthy human being."

We ought to be good stewards of the talents and opportunities we are given, and striving to accomplish lofty goals can be a very good thing. However, our intrinsic value as humans is not a result of our accomplishments. Knowing that our value is not on the line gives us the freedom to try hard things when a successful outcome is uncertain, to sincerely welcome honest feedback on our work, and to celebrate the accomplishments of our peers and competitors. This is something I hope to convey to each student I teach and/or work with (and indeed often need to preach to myself).

Besides your work, what's something that you're passionate about?

When I'm not thinking about statistics, I enjoy playing soccer, riding my bike, and participating in my church community. Also, having grown up in Dallas, my fandom for the Cowboys has been a life-long lesson in perseverance.

What was the best part of your college experience?

I did my undergrad at Rice University, where social life revolved around the residential college system. Each student was randomly assigned to a college, so the resulting mix was a cross section of the student body and facilitated friendships amongst students of various backgrounds, majors and interests. Looking back, I'm thankful that this "randomization" gave me the opportunity to form deep friendships with people who I might otherwise not have met.