Last weekend, from Sept. 30 to Oct. 2, top statisticians and mathematicians came to Washington University in St. Louis for the first-ever Workshop on Higher-Order Asymptotics and Post-Selection Inference (WHOA-PSI). Todd Kuffner, an assistant professor of mathematics and the organizer of the event, first conceived of the workshop a year and a half ago.
“My vision for the workshop is something that many people initially regarded as a little crazy,” Kuffner says, “in the sense that I am merging two groups of people who usually do not talk to one another and, until I put them in the same room together, are largely unaware of the potential benefits of interaction.”
He explains, “Higher-order asymptotics is often incorrectly regarded as purely mathematical, and it is not well-understood by methodological or applied statisticians. My belief is that the collective wisdom of the higher-order asymptotics community is exactly what is needed to further refine the exciting new ideas coming out of post-selection inference, so that what eventually reaches the practicing data analyst is a set of inferentially valid and highly-accurate procedures. In this way, the workshop is forward-looking, and trying to anticipate the future needs of users of statistics.”
When John McCarthy, chair and the Spencer T. Olin Professor of Mathematics, first heard Kuffner’s proposal, he thought it sounded like a big task. Too, the invitation list included top statisticians from across the globe, and McCarthy doubted such high-profile names would come. He says, “Instead, they all accepted!”
WHOA-PSI was the first workshop of its kind, dedicated completely to post-selection inference, and Kuffner credits this fact with the workshop’s popularity. “This is one of the most exciting emerging fields in statistics, and there is tremendous enthusiasm in the community,” he says. “I am very fortunate to have a supportive organizing committee who let me pursue my vision for the workshop.”
“We have a small group of statisticians in the Department of Mathematics, but we are doing some great work here,” says Kuffner. “I wanted to bring many of the top senior and junior statisticians around the world to our campus, both to showcase what we have to offer, and to enable more of our faculty and students to participate.” In the end, nine WashU faculty and sixteen WashU graduate students attended the workshop and met with colleagues including more than fifty faculty, post-docs, and graduate students, from 34 other universities in 6 countries. Dean of Arts & Sciences Barbara Schaal, the NSF, the Department of Mathematics, and the Division of Biostatistics all generously supported the event.
The list of visiting scholars also contained several big names. Though there in no Nobel Prize in Statistics, Kuffner explains, the equivalent award is the COPSS Presidents’ Award, and the list of speakers and organizers for WHOA-PSI included 6 COPSS Presidents’ Award winners. Kuffner was thrilled to have such prestigious scholars involved, including Xiao-Li Meng (Dean of the Harvard Graduate School), Nancy Reid (University of Toronto, foreign associate of the National Academy of Sciences), Jianqing Fan (Princeton University), Xihong Lin (Chair of Biostatistics at Harvard), Peter McCullagh (University of Chicago, Fellow of the Royal Society), Robert Tibshirani (Stanford University, elected to National Academy of Sciences), Don Fraser (University of Toronto, Officer of the Order of Canada), and Lawrence Brown (University of Pennsylvania, elected to American Academy of Arts and Sciences and National Academy of Sciences).
“This is great for both students and faculty at WashU,” says McCarthy, “exposing them to the leaders of the field, and, conversely, exposing those leaders to the Washington University statistics community. Statistics has been growing very rapidly here - we now have 19 M.A. students in statistics, compared to one or two five years ago. We had 4 statistics students in our incoming PhD class. I think the workshop is another big step in establishing our international presence in statistics.”