Directory Calendar News Careers Alumni Giving


SIBS Faculty

Roslyn A. Stone, PhD
Marnie Bertolet, PhD
Maria Mori Brooks, PhD
Eleanor Feingold, PhD
Abdus Wahed, PhD
John W. Wilson, PhD

Collaborating Investigators:

Steven Bell, PhD
Michael Fine, MD
Tanya Kenkre, PhD
Mary Marazita, PhD

Roslyn A. Stone, PhD

Professor of Biostatistics

Senior Statistician, Biostatistics and Informatics Core,
Center for Health Equity Research and
Promotion, VA Pittsburgh Healthcare System

SIBS Pittsburgh Program Director


B.S. (1969), Mathematics, Douglass College (Rutgers University)
M.A. (1972), Mathematics Education, University of Kansas
M.S. (1979), Statistics, Montana State University
Ph.D. (1986), Biomathematics, University of Washington

Research Interests

Clustered categorical data; Survival analysis; Guideline implementation;
Cluster-randomized trials; Health services research; Medication safety;
Occupational health; Cohort studies; Statistical education

Serendipity has played a major role in my path to becoming a biostatistician, a career I had never heard of until my second round of graduate school.  Although I’ve always liked math, when I started graduate school in mathematics, I realized that real analysis and abstract algebra were not my “thing.”  I switched into the math education program, and took a couple of courses in educational statistics and experimental design that really caught my attention.  I decided then that if I ever went back to graduate school I would major in statistics.  Meanwhile, I had taken a summer job in Lake Tahoe, CA, that turned into five years of serving cocktails and dealing “21” as well as skiing and other outdoor activities.  I then decided to get serious about a career and moved to Montana, where I got a part-time job teaching middle school math and a second part-time job dealing poker.  Although I enjoy teaching, I did not enjoy teaching middle school math.  I interviewed at Montana State University, where I was offered a teaching assistantship that paid for my M.S. training in statistics, and took several courses in biology.  I discovered that I really liked statistics as well as biology, and enjoyed teaching at the college level.  I then applied to the Ph.D. program in biomathematics at the University of Washington, and was offered a graduate student research position, and subsequently a training grant from the National Heart, Lung, and Blood Institute, to support my education there.  I came to the University of Pittsburgh when I finished my Ph.D.

At Pitt, I quickly became involved with teaching and collaborative projects that utilized my training and interest in generalized linear models, clustered categorical data, and survival analysis.  I have maintained several long-term collaborations with investigators in the areas of occupational health, epidemiology, health services research, medication safety, and disparities research, with specific focus on pneumonia, pulmonary embolism, occupational cancer, diabetes, and chronic kidney disease.  Currently, I am senior statistician in the Biostatistics and Informatics Core of the Center for Health Equity Research and Promotion (CHERP) at the VA Pittsburgh Healthcare System, which focuses on health services issues related to the quality and equity of medical care.  I am the Program Director for SIBS-Pittsburgh.  I review applications for the National Institute for Occupational Safety and Health, and have served on the national Data Safety and Monitoring Board for multi-site VA health services trials.

Biostatistics requires communication skills as well as analytic and technical skills.  I spend about half of my time writing, editing, or reviewing manuscripts and research proposals, and most of the rest in meetings with clinical investigators and students.  My biostatistical training and expertise have allowed me to collaborate on many important and interesting studies.

Marnie Bertolet, PhD

Assistant Professor of Epidemiology
Statistician, Epidemiology Data Center


B.A. (1993), Mathematics, Shippensburg University of Pennsylvania
M.Eng. (1995), Operations Research & Industrial Engineering, Cornell University
M.S. (2002), Statistics, Carnegie Mellon University
Ph.D. (2008), Statistics, Carnegie Mellon University

Research Interests 

Survival analysis; Longitudinal analysis; Clinical trials;
Survey sampling; Bayesian analysis; Cardiovascular epidemiology;
Diabetes epidemiology                      

My research interests focus on the many uses of statistics in public health. My devotion to statistics stems from the diversity of applications to which statistics can be applied.  I came to public health after a career in telecommunications, where I was a reliability and quality engineer doing analysis analogous to “consumer reports” for telecommunication equipment.

My applied research currently focuses on cardiovascular and diabetes research. I apply longitudinal, repeated measure and survival analysis methods to large clinical trials.

My methodological research currently focuses on incorporating Bayesian and survey sampling analyses into existing studies. I am extending this work to the public health setting by investigating similar methods to increase the generalizability of clinical trials.

Maria Mori Brooks, PhD

SIBS Pittsburgh Associate Program Director
Associate Professor of Epidemiology and Biostatistics
Statistician, Epidemiology Data Center


B.A. (1985), Mathematics, Williams College
M.A. (1986), Statistics, Harvard University
Ph.D. (1991), Statistics, University of North Carolina at Chapel Hill

Research Interests 

Survival analysis; Longitudinal analysis; Clinical trials; Epidemiology of cardiovascular disease and diabetes
My work focuses on the design, conduct and coordination of multi-center clinical trials and the statistical analysis of clinical and epidemiological data.  My first job was in the Biostatistics Department at the University of Washington, where I performed statistical analyses for the Cardiac Arrhythmia Suppression Trial (CAST).  This randomized clinical trial showed that two approved drugs used to treat abnormal heart rhythms were associated with higher rates of arrhythmic death compared with placebo. 

Since 1995, I have been in the Department of Epidemiology at the University of Pittsburgh Graduate School of Public Health.  I teach Epidemiology Methods to students interested in public health.  I also direct the statistical analyses for several studies, including the Bypass Angioplasty Revascularization Investigation (BARI) and the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D).  These trials evaluate the comparative effectiveness of treatment strategies for patients with coronary disease and type 2 diabetes.  I oversee protocol issues, monitor compliance and risk-factor control data, analyze clinical outcome data, as well as write and edit scientific manuscripts.  In addition, I have served on a number of Data and Safety Monitoring Boards for trials sponsored by the National Institutes of Health and by industry.

Eleanor Feingold, PhD

Professor of Human Genetics and Biostatistics

B.S.(1985), Interdisciplinary, Massachusetts Institute of Technology
Ph.D. (1993),  Statistics, Stanford University

Research Interests 

Statistical genetics

After graduating from college with a major that combined math and science writing, I spent three years working for a large electric utility company, doing operations research. I loved using math to answer scientific questions, like how the water in various reservoirs should be managed, but I found that the field of operations research had limited appeal for me. I didn't like converting everything to dollars and maximizing it, and I also felt limited by not considering the randomness of nature in most calculations. Those insights guided me toward graduate school in statistics, and eventually a career in biostatistics, specifically statistical genetics.

I have been working in statistical genetics ever since my dissertation, which developed stochastic process models for analyzing data from a new genetic mapping method. Most of my work is to develop statistical methods for gene mapping - for finding genes that affect risk for various diseases.  I also apply those methods to research on Down syndrome, premature birth, dental health, lymphedema, and a number of other traits and diseases. I have been particularly fascinated over the years by the statistical challenges posed by successive generations of genomics technology. Every new technology requires new statistical methods to clean and analyze the data properly, and until those methods are developed, the technologies are close to worthless.  I am also obsessed with teaching both statisticians and geneticists to choose statistical methods based on what the scientific questions are, rather than mechanically computing p-values that might not even answer the relevant question.

Abdus Wahed, PhD

Associate Professor of Biostatistics


B.Sc. (1992), Statistics, University of Dhaka (Bangladesh)
M. Sc. (1994), Statistics, University of Dhaka (Bangladesh)
M.A. (2000), Mathematical Statistics, Ball State University
Ph.D. (2003), Statistics, North Carolina State University

Research Interests 

Dynamic treatment regimes; Longitudinal data analysis; Missing (Censored) data; Survival analysis; Clinical trials

My research is focused mainly on dynamic treatment sequencing.  Treatment of complex diseases such as cancer, leukemia, AIDS and depression usually follows complex treatment regimes consisting of a sequence of multiple stages of therapies.  A dynamic treatment regime (also known as an adaptive treatment strategy) is a sequence of individually tailored treatments that allows for flexibility to modify treatments based on intermediate response.  During the treatment period, individuals receive time-varying treatment based on their health status and other eligibility criteria (specified prior to the start of the treatment).  One example is leukemia treatment, where patients receive some form of induction therapy and those who achieve remission receive some form of maintenance therapy to maintain remission.  While making treatment decisions at each stage, a physician looks at multiple factors including (i) treatments assigned in prior stages, (ii) response to the treatments in prior stage, (iii) health status (quality of life) of the patient, and (iv) possible treatments for which the patient is eligible at that particular stage. The goal at each stage is to decide on the treatment that will result in the largest patient benefit (i.e. longest survival).  I have been conducting research to develop efficient methodology for analyzing time-to-event and longitudinal data in order to compare different dynamic treatment regimes based on observational or randomized studies.

I have been working as the lead statistician in several multi-center studies including Virahep-C at the Epidemiology Data Center. In this role, I am responsible for (or oversee) the following steps (in the order of how a medical research study is conducted):

Preparation of grant proposals to support the study
Study design (sample size estimation, randomization, etc.)
Protocol development
Regulatory approval such as IRB or FDA
Randomization scheme (if any)
Monitoring patient recruitment
Periodic reports to data safety monitoring board
Periodic reports to the steering committee
Analysis of the data
Manuscript writing

I enjoy playing and working with kids (my own, and those at Pitt), developing biostatistical methods and playing carom and bridge.

John W Wilson, PhD

Assistant Professor of Biostatistics
Statistican, National Surgical Adjuvant Breast and Bowel Project


B.A. (1975), Biology, Yale University
M.S.(1979), Evolutionary Biology, Cornell University
Ph.D.(1982), Evolutionary Biology, Cornell University
M.S. (1987), Biomathematics, University of Washington

Research Interests

Nonparametric statistics;  Statistical methods for clinical trials; Statistics education

Upon completion of my Ph.D. in biology, I discovered that I really enjoyed computer programming, math and statistics somewhat more than I enjoyed biology (a fine time to make that realization...).  I also discovered that statistics jobs were much easier to come by than were biology jobs.  So, I returned to school to learn more about statistics and have never regretted the decision (although I still miss biology fieldwork... That was fun!).

I have been fortunate to be able to combine my interests in biology and math in several interesting settings.  As part of a team studying endangered wombats in South Australia, I estimated the number of wombats in a conservation park where most of the remaining animals were found.  In the northwestern U.S, I participated in a number of projects related to salmon migration in the Columbia and Snake Rivers.  My MS thesis in biomathematics examined the utility of redirecting water flow through hydroelectric dams to minimize the kill of salmon fingerlings by turbines.  Since moving to Pittsburgh, I have been the statistician on research projects in cardiovascular medicine, pharmacology and cancer treatment.  

In teaching statistics, I have tried to emphasize that the most effective use of statistical tools requires some knowledge of the scientific field of application.  In my SIBS teaching, I hope to convey not only that philosophy, but also the satisfaction and enjoyment that come from collaboration with congenial researchers to address interesting research problems.

© 2018 by University of Pittsburgh Graduate School of Public Health

Login  |  Sitemap