Epidemiology

Raymond Carroll's work has had an important impact on epidemiologic research. This article reviews contributions to theory for the case-control design and to methods for nutritional and radiation epidemiology. Some of these contributions build on Ray's broad-ranging research on regression analysis, measurement error, and missing data problems. Ray has been a welcome visitor at the U. S. National Institutes of Health (NIH), first with the National Heart, Lung, and Blood Institute and later with the National Cancer Institute (NCI), both as a Visiting Scientist and Guest Researcher and as a friendly collaborator who drops by from time to time. At NIH, he has given valuable advice on a wide range of topics and collaborated on many projects not covered by this article, including the analysis of survival data with informative censoring (Wu and Carroll, 1988 [OW-2]), the design of community intervention trials (Gail et al., 1996), the design and analysis of the kin-cohort design for genetic epidemiology (Carroll et al., 2000 Gail et al., 1999), the meta-analysis of surrogate endpoints (Gail, 2000), and agreement of exposure assessments based on quantile groupings (Borkowf et al., 1997), among many others.