Elizabeth A. "Betsy" Tolley, Ph.D., M.S., B.S.
Biostatistics & Epidemiology
633 DOCTORS OFFICE BUILDING
66 NORTH PAULINE STREET
MEMPHIS TN 381630000
- Ph.D., Virginia Tech, Blacksburg, VA, Genetics
- M.S., Virginia Tech, Blacksburg, VA, Animal Science
- B.S., Virginia Tech, Blacksburg, VA, Animal Science
Dr. Tolley is a tenured Professor of Biostatistics and Epidemiology in the Department of Preventive Medicine at the University of Tennessee Health Science Center. She holds a doctorate in population genetics with co-majors in statistics and physiology from the Department of Animal Science at Virginia Tech, Blacksburg, VA, and received post-doctoral training in building statistical models in the Departments of Animal Science, Economics, and Statistics at N.C. State University, Raleigh, NC. She is course director of the two-semester graduate-level course in Biostatistics for the Health Sciences I and II and for the graduate-level course in Linear Regression Models offered in the Masters of Epidemiology degree program. She has mentored numerous graduate students at the masters and doctoral levels. Dr. Tolley has served as a biostatistician, co-investigator, and consultant on many NIH grants. She also provides biostatistical consulting and collaboration services for many basic science and clinical investigators as part of her assigned faculty duties. Dr. Tolley has served as the biostatistician on clinical trials and prospective cohorts. She has expertise in designing these types of studies and in analyzing data and interpreting results from such studies to address the research questions proposed by the clinical investigators. Early in Dr. Tolley’s career as an academic consulting biostatistician, she collaborated with Dr. Antonius Miller, then a faculty colleague at the University of Tennessee, Memphis. Together they built pharmacodynamics models for the joint purposes of maintaining a therapeutic level of etoposide within a narrow range, while avoiding toxicity characterized by severe neutropenia. Using new patients they validated their models or identified threats to validity, or causes of model instability and over-optimism. Throughout her career, Dr. Tolley has analyzed data representing a wide range of biomarkers (including Vitamin D) and cell-signaling peptides from clinical investigations. Advances in analytical techniques have given investigators the ability to analyze many different substances simultaneously from a single specimen obtained cross-sectionally or longitudinally, thereby creating numerous challenges for biostatisticians who are asked to “make sense” of the data. Generally, investigators wish to identify which biomarkers are mechanistically associated (i.e., potentially diagnostically associated) with disease outcomes or which ones are predictive of disease outcomes. Throughout her career, Dr. Tolley has also worked with clinical investigators to develop mechanistic or diagnostic models of disease outcomes and predictive models of such outcomes. While the regression methods (linear, nonlinear, logistic, or survival) are often similar, the processes for building mechanistic models differ from those for building predictive models. For example, collinearity among independent variables can pose serious problems for mechanistic models, but does not pose as much of a threat to the validity of predictive ones. Dr. Tolley''s peer-reviewed articles provide diverse examples of clinical prediction and mechanistic (or diagnostic) statistical models that Dr. Tolley built in collaboration with clinical investigators.
A list of Dr. Tolley''s peer-reviewed publications can be found at the following URL: http://www.ncbi.nlm.nih.gov/sites/myncbi/1ZmeDV_SEm5AH/bibliography/47569292/public/?sort=date&direction=ascending