The objectives of this ongoing research are 1) to develop new or improved quantitative methods for assessing the effects of one or more environmental exposures, especially carcinogens; and 2) to apply these methods to epidemiological and experimental data. Improved statistical techniques wil be sought in three problem areas: 1) methods for relating disease-free survival to single and multiple exposures, and application of these methods to epidemiological cohort data; 2) methods for dealing with missing values and other types of incomplete observations in multiple regression analyses of retrospective and prospective epidemiological data; 3) methods for using animal cancer bioassay data to estimate human risk from environmental carcinogens. In areas 1 and 2, particular attention will focus on methods for accommodating exposure errors when estimating disease risk due to environmental toxicants and other lifestyle factors. Data describing the mortality experience of the US uranium miner cohort, data on cancer incidence and mortality among a cohort of former college students, and data generated in animal experiments will be focal points for the development and application of improved techniques in these three areas.
Showing the most recent 10 out of 11 publications