The University of Georgia College of Veterinary Medicine

Office for Research and Graduate Affairs

Statistical Service

The Veterinary Medical Experiment Station has established a statistical consulting service at no cost for faculty and graduate students in the College of Veterinary Medicine. Dr. Deborah Keys has been contracted by the College to provide this service for the past five years and has become an integral part of our research program.

Dr. Keys lives and works in Athens. She obtained her Ph.D. from North Carolina State University in 1998, and has held several positions in both academia and industry and brings a tremendous amount of knowledge and practical experience in statistical analysis for both clinical and basic science research.

Because experiments in veterinary medicine often involve the use of valuable laboratory animals, it is essential that such experiments be well designed. Moreover, the ethical treatment of experimental animals requires that optimal information be obtained from experimental results. Poorly designed experiments can lead to low statistical power, unnecessarily complex statistical analyses, or more seriously, ambiguous results. Collaboration with a statistician during the design phase of an experiment can reduce cost; ensure that the optimal amount of information is obtained from experimental animals, and that a sufficient sample is obtained to detect desired effects with good power.

Research grant proposals often require detailed plans for experiments and analysis of the resulting data. Review of proposals by a statistician before they are submitted would allow for improved quality of experimental design and should give researchers a competitive edge when seeking external funding. This benefit also extends to research papers. Collaboration with a statistician can lead to improved data analysis, improving the quality of such papers, and hence, increasing the likelihood that they will not only be accepted for publication, but also increasing their visibility within the veterinary and scientific literature.

In summary, the College of Veterinary Medicine benefits from this statistical consulting service through:

Improved experimental designs, optimizing information obtained from experimental animals and enhancing the power to detect desired effects;
Improved competitiveness for research grants;

Consulting will be provided free of charge to faculty and graduate student researchers in the College of Veterinary Medicine. Appointments are offered the first and third Tuesdays of each month and can be scheduled by contacting Lisa Reynolds at lisar@uga.edu or by calling 706-296-7818.

Faculty are especially encouraged to seek critical review of all experiment designs proposed in their grant proposals before they are submitted. During the design phase, the details of the experiment will be planned including the definition of the objectives, selection of treatment and what variables will be measured, allocation of treatments to experimental animals, protocols for obtaining the desired data, and a preliminary plan for how the data will be analyzed. If it appears during the course of the experiment that the experimental protocol needs to be changed, advice should be sought from the statistical consultant before any major changes are made. This will ensure that such changes will not compromise the quality of the experiment. Statistical analysis of data is available, but the cost is the responsibility of the investigator. The fee for this service is $75/hour (paid from a designated research or departmental account). A statistical service agreement is signed by both the investigator and Dr. Keys prior to analysis and provides an estimate of analysis time and cost.

This page last updated September 22, 2010.