Colloquium

Department of Statistics

Statistics Colloquium
October 31 (Fri) 11 a.m.
Alavi Commons Room, 6625 Everett Tower

Challenges and Opportunities in Providing a Statistical Support Model to a Large Research Organization (Musings from a Zoetis Biometrics Insider)

Clark D Smothers, PhD
Zoetis (formally Pfizer Animal Health)
Veterinary Medicine Research and Development
Kalamazoo, MI

Prior to 1983, most of the design and statistical analysis used in Pfizer Animal Health was conducted by someone with a PhD in Nutrition. A plant geneticist was employed to begin a more formal statistical group, but that really did not come to fruition until 1987 when former head of KSU Statistics, Art Dayton, was hired to lead the Biometrics group. I came on board in 1989 and together with a biometrician in Sandwich, UK we began working with efficacy and safety studies strictly for product development. Professors George Milliken and Dallas Johnson were key contributors to our philosophies regarding the design and analysis of experiments. George and Dallas, along with other academic partners like Walt Stroup and Ray Littell, helped us forge a common approach to our statistical consulting that we still use today. In those days, we only had a few projects and now we work with hundreds. We consult with Clinical Development, Discovery, Metabolism and Safety, Pharmaceutical Sciences, Biologicals Development, Analytical and Commercial project teams. In addition to animals or pens, experimental units can now be fermentation vats or rows on a 96 well plate.
Today, the Global Biometrics Team consists of 12 graduate-level biometricians. Some of the challenges we face are the same ones we faced a long time ago. Some are not. Opportunities continue to present in underserved areas. We work with researchers and regulators all over the world. The statistical rigor demanded by agencies and our own colleagues continues to increase. Biometricians are seen as an integral part of project teams and the demand for statistical services continues to grow. Providing a consistent, efficient, high-quality model of statistical support requires communication, teamwork, education, influence, technical expertise and the ability to form strong scientific relationships. Please join me as we take a look at some of the challenges and opportunities commensurate with this support model. It is exciting to see the value that appropriate statistics has created over the last 25 years in this organization and beyond.

All statistics students are expected to attend.

Past colloquiums