The purpose of this project is to conduct research in statistical methods and computer techniques with particular emphasis on those appropriate for analyzing data from clinical, diagnostic, and prevention trials and epidemiologic studies of cancer. Many of the problems studied under this project arise from the consultative activities of the Section. During the past year, emphasis has been given to a number of projects dealing with sample size determination, including the impact of prevalent cases on prevention studies (with and without prior screening), and determination of sample sizes in the presence of arbitrary hazards. Previous work on an interactive computer program for calculating sample size was expanded to address a wider range of situations and study designs. Another area of research has been methodology for correcting logistic regression for errors in variables (such as occur in determining dietary history). Work has continued on the analysis of cancer maps involving analysis of variations in cancer mortality. Other research has involved competing methods for analysis of qualitative interactions (where the effect of a variable is in opposite directions in different subsets). A model of the cost of clinical trials as a function of time was investigated with regard to implications for either treatment or prevention trials. Finally, work has continued on maintaining and improving software for interactive analysis of complex medical data using sophisticated multiple regression techniques and survival analysis.
