October 8th

Welcome and Plenary Session (8:00-9:00AM)

Enthusiastic Collaboration and Mentoring

Scott Vander Wiel, Los Alamos National Laboratory

Abstract: This is a three-part talk. First, I will reflect on Gerry Hahn who, with others, mentored me during two summers at the General Electric Corporate R&D Center beginning 37 years ago. I will discuss some highlights of this privilege, including Gerry’s enthusiastic and supportive mentoring style and his encouragement to dive into the world of collaborative industrial statistics for improving manufacturing processes and product quality.

Second, I will present new work from my own recent collaborations and mentoring—namely, co-active subspace methods (CoASM) for adjacent computer models. CoASM locates directions in the parameter input space that maximally drive the outputs of two adjacent computer models in the most similar fashion. This new work extends CoASM to functional outputs, inspired by tests of high-explosives that are set off within cylinders made of various metals.

Third, I will offer a variety of insights from a career’s-worth of advancing statistical methods through collaborative problem solving with physical scientists and engineers. in forward-looking research environments. These include the value of teaming with other statisticians, dealing with terminology differences, diving into the science or engineering deep end, coming up for statistical fresh air, and enjoying shared successes.

Luncheon (12:15-1:45PM)

History and Happenings of the Houston Livestock Show and Rodeo

Samantha-Anne Horwitch Nadolny, Houston Livestock Show and Rodeo Speakers Committee

Abstract: TBD

Youden Address (4:00-5:00PM)

Design of experiments: The key to the cause-and-effect kingdom

Christopher Nachtsheim, University of Minnesota

Abstract: The design of experiments is the most powerful and efficient method for discovering and characterizing cause-and-effect relationships in real-world phenomena. We begin the presentation by touching on the origins of DOE, its connection to the scientific method, its power, and the limitations of observational studies. We chronicle the rapid growth of DOE applications and highlight key developments in DOE over the past half century. One of the controversies that we have followed closely over the past half century has been the battle between the evangelists of “optimal” design and the evangelists of “classical” design. That controversy, in our view, has been resolved and we will share our views on the nature of that resolution. We close with some discussion of a project that we have been involved in for the past four years involving a textbook for the design and analysis of experiments, and web-based software system that has been developed in support of the text.

October 9th

Luncheon (11:45AM-1:15PM)

Advancing Science Together: The Quiet Power of Everyday Statisticians

Ji-Hyun Lee, University of Florida

Abstract: In a world that celebrates exceptionalism, the contributions of everyday professionals often go unnoticed. Funding favors “exceptional” work, but real progress depends on the many, not just the few. In this talk, I challenge the idea that only the extraordinary make a difference. As a biostatistician, collaborator, and team scientist, I’ll share how statistical thinking and leadership drive science and patient care. Now, as ASA’s 2025 President, I’ll further this mission by leading initiatives to strengthen our community, advance our field, and champion the power of everyday statisticians.

Reception with SPES Special Panel Session (3:15-5:15PM)

Digital Twins in Industry and Academia

Panel discussion: TBD

Abstract: Digital twins are a new technology that integrates modeling, data collection, and decision making for a complex system by defining an iterative two-way interaction between the physical agent and its digital representation. Digital twins are currently used in many areas of science and engineering, from manufacturing, to healthcare, to operating autonomous vehicles. While they are considered to be a powerful tool, a unified statistical framework for understanding these diverse systems and quantifying uncertainty from various sources is still a work in progress. This panel discussion will focus on the current applications of digital twins in industry and academia, commonalities and differences in their structures, and primary questions of interest to researchers who are working with these systems.