The primarily focus of our research is the application of Bayesian methods to epidemiologic problems, such as diagnostic test evaluation and disease prevalence estimation.
The primary practical advantages of the Bayesian framework are that
- it can always be used for small samples.
- it allows great flexibility in coping both conceptually and computationally with complex statistical models.
- it allows for the combination of current data and available scientific information in a coherent way.
- it allows direct probability interpretation of outcome results and therefore practical and statistical significance can be readily evaluated by scientists and policymakers.
- To develop innovative techniques that assess test accuracy in disease diagnosis at individual- and population-levels with or without a gold standard, facilitate interpretation of test results, and improve disease risk modeling based on diagnostic test results.
- To develop statistical methods to make improved inferences that a herd (or group of herds in a zone, state, or country) is free of important infectious agents based on herd-level test results.
Funding for research described on this site was provided in part by the USDA - CSREES (now USDA - NIFA)