Welcome and Plenary Session (8:00-9:00AM)
Effective Statistical Power
Julia O’Neill, Direxa Consulting LLC
ABSTRACT: The most powerful statistical methods are those actually adopted. In the 38 years since I fell in love with Design of Experiments in Bill Hunter’s Stat 424 class at the University of Wisconsin, I have had the opportunity to study and work with many of the most influential leaders in applied statistics. Gerry Hahn established a high standard among these role models, and it is my great honor to be selected for the award recognizing his contributions.
My applied statistics work has contributed to some technological advances with impacts on the environment and on human health. Along the way I’ve observed and tested a range of approaches to realizing the adoption of statistics. In this lecture I will share a synthesis and distillation of some of the lessons and principles that are the basis for my statistical consulting practice. I will also tell some stories illustrating what has proven most effective from my work in vaccine development and biotech.
Statistical thinking and strategies played a key role in the unprecedented acceleration of the Moderna COVID vaccine Spikevax, with more than 1 billion doses now administered worldwide. Approaches gleaned from my previous work in Design of Experiments, Total Quality, Six Sigma, and a broad range of statistical techniques, were instrumental in moving the vaccine from initial sequencing to Emergency Use Authorization in 11 months. Solid scientific and statistical work were essential in demonstrating the safety and efficacy to reviewers from health authorities globally.
Tess-celestial: Immersive Science Approach to Mathematics
Radmila Sazdanovic, North Carolina State University
ABSTRACT: Bridging the gap between theoretical, computational, and experimental mathematics and other sciences starts with effectively sharing and communicating research ideas. Tess-calestial is an exhibit that can be viewed in large-scale immersive visualization environments or online via a responsive web design. It explores the interplay of art, nature, culture, shape, perspective, and mathematics through tessellations of the hyperbolic plane showcasing that mathematics is contemporary, dependent on societal needs and technological developments and mediated by cultural and historical influences.
Youden Address (4:00-5:00PM)
DOE Software: Time for a Revolution?
ABSTRACT: Since 2000, much research has been done on the design of industrial experiments. Some of that research has inspired DOE software developers to offer new kinds of experimental designs (e.g., split-plot designs and definitive screening designs). As a result, the experimental designs available to practitioners through DOE software nowadays are more efficient than ever. There has, however, not been any major innovation in the software design itself, or in the way the software is delivered to the user. For instance, the user interfaces are roughly the same as 20 years ago. Also, all DOE software packages are still standalone packages, that do not exploit the opportunities offered by the internet and by artificial intelligence. At the same time, most will agree that DOE is heavily underused. Inevitably therefore, DOE software is also heavily underused. That DOE and DOE software has not gained as much popularity as we would like has various possible causes. One of these causes might be the current DOE software. What if DOE software was really user friendly, and provided guidance to non-expert practitioners when selecting designs for efficient experimentation and when analyzing experimental data? What if there was a web-based DOE software facilitating collaboration? What if DOE software would help with the interpretation of the results and the identification of the true effects, possibly by exploiting artificial intelligence? Is it time for a revolution in DOE software, for a new kid on the DOE block, a challenger of existing software packages?
From the Foundations of Quality Management to the Future: Quality Science, Data Science, and Implementation Science
A. Blanton Godfrey, North Carolina State University
ABSTRACT: What most of us consider modern quality management began almost exactly one hundred years ago when Walter Shewhart wrote a short one-page memo and attached a drawing of a control chart. Over the years, we have added many new tools, methods, and procedures that comprise our current quality management systems. Some of these are well understood and widely used while others have been forgotten, misunderstood, or under used. In this presentation we will quickly review the most important of these tools, methods, and procedures and make a strong case for adding some to our current quality management systems. In the second half of the talk, we will talk about three major directions for future quality management systems.
Reception with SPES Special Panel Session (3:15-5:15PM)
The Impact of Artificial Intelligence Technologies On Quality and Statistics
Panel discussion: Simon Mak and David Banks, Duke University and Annie Booth and Alyson Wilson, North Carolina State University
ABSTRACT: Cutting-edge artificial intelligence (AI) technologies have astounded the world with their remarkable prowess in tackling intricate real-world challenges. These technologies have permeated novel domains within the physical and engineering sciences. As the relentless integration of AI persists, it begets a pressing inquiry into its profound implications on the very fabric of quality and statistics, wielding the potential for transformative outcomes, both boon and bane. Brace yourselves for a riveting panel that assembles diverse viewpoints and firsthand encounters with AI’s seismic influence on the intricate tapestry of quality and statistics. (This abstract was written with the assistance of ChatGPT).