James Connell Gear, DrPH

James Connell Gear, DrPH

Retired and consulting

Describe your current position
Quantitative support (data analysis, data mining, predictive modeling for consulting clients on an ad-hoc basis.

List your career highlights
– Director of Research Statistics and Associate Professor of Community Medicine, Howard University College of Medicine, Washington, DC., 1987-1989

– Statistical Technical Writer, SAS Institute, Cary NC, 1989-1991

– Research Scientist (Industrial Statistician), Philip Morris, USA, R&D, Richmond, VA, 1991-2002

– Senior Statistician (Regulatory), Monsanto Company, St. Louis, MO, 2002-2004

– Senior Statistician and Director of Predictive Modeling, Midwest Employers Casualty Company, Chesterfield, MO, 2006-2012

What was your favorite thing about the Department of Biostatistics?
I always appreciated the level of scholarship and professionalism of the faculty! I believe we had access to the “best of the best” in biostatistics and statistics in our department, and I was always proud to be a part of such an august body!

Discuss your relationship with faculty and/or staff during or after the program.
I have always counted on my relationship with the faculty as a professional advantage. Individually and collectively, they served as mentors and advisors in professional and technical areas. And my dissertation advisor, Dana Quade, also shared advice in personal areas as well. I count my relationship with the faculty as a blessing!

What would you consider to be your greatest achievement in your professional career and why?
I am proudest of two things:

1) While at Howard University College of Medicine, I was a part of the first study to scientifically demonstrate that the prevalence of glaucoma in a Black population is almost 5 times the prevalence in the general population, leading to co-authorship of a landmark paper in ophthalmologic research.

2) In the early 1990’s, Philip Morris was developing a firesafe cigarette, and the government regulatory agency had required the ignition propensity data (# of ignitions/# of chances to ignite) to be analyzed as normalized binomial proportions. I used logistic regression to model those data. They rejected my analyses, but I insisted that the analyses were appropriate. The regulatory agency had to hire a consulting statistician to confirm the validity of my analyses, which changed their standard for analyzing such data.

What is your first memory of Chapel Hill or the Department of Biostatistics?
When I visited the department before enrolling, I remember talking with David Kleinbaum about the challenges of being a student in the department. He was frank and honest about those challenges, and how difficult it was likely to be. I remember his openness about that, and I remember thinking that “They really must want me here, to tell me the truth like this!” For me, his honesty really was the best policy!