Statistical Methodology Program

The primary goals of the Statistical Methodology Program are to investigate design and statistical analysis issues related to epidemiologic studies in the general area of environmental health and genetic epidemiology. Specifically its objectives are:

• To conduct statistical research on epidemiologic study designs for investigating the human health effects of environmental exposures and on host factors (genetic, nutritional, socioeconomic, etc) which modify the response to environmental agents;
• To develop methods of analysis suitable for studying such relationships;
• To collaborate with other investigators to identify emerging methodologic problems related to new and on-going research projects and to assist in the application of new methods to the design and analysis of such projects

The program's environmental epidemiology oriented research has spanned such topics as the design of nested case-control studies and related methods of sampling within cohort studies, the analysis of exposure-time-response relationships, and study design and analytic techniques for dealing with exposure measurement error, longitudinal studies, and multi-level ecologic studies.

In the genetics field, the program have also been considering questions relating to both study design and analysis. The design work has focused on family studies for gene characterization, such as the NCI Collaborative Family Registry for Breast and Colorectal Cancer Research. The analysis methods are focused on gene-environment interactions and the analysis of censored age-at-onset phenotypes, particularly those using Markov chain Monte Carlo and Generalized Estimating Equations methods.