The objective of Week 6 is to examine the special analytic problems associated with observational studies in drawing associations between exposures and disease outcomes. Not only are epidemiologic studies observational, but often the quality of the data is poor. Some of these probelms include measurement error, informative missingness, aggregate data used for defining individual risks, the use of monitoring and surveillance data and the use of geographic information systems (GIS) for spatial mapping. The leading question is whether it is possible to draw inferences about actual effects of exposures on disease outcomes in epidemiologic studies.

The realization of the connection between health and the environment is only a recent phenomenon. It reflects advances in the science of measurement and the development of statistical techniques capable of dealing with complex data. It is a further purpose of the workshop to discuss several case studies of environmental problems relating to health, including the role of small changes in the magnitude of air pollution on mortality.