The program goal is to train predoctoral and postdoctoral biostatisticians in biostatistical theory and applications in the area of cancer research. This will be accomplished with a rigorous training program in both biostatistical theory and applications while providing the opportunity for trainees to apply these statistical techniques to data collected from the many different types of studies involving cancer, such as, clinical trials, observational and psychosocial data. The Department of Biostatistics will provide the administrative leadership for this program a n d provide the necessary courses in biostatistical methodology and applications. The trainee must satisfy the requirements of the Ph.D. program in Biostatistics as well as additional courses designed to broaden exposure to problems in cancer. Trainees will serve a two-month internship comparing alternative methods of analyzing the data collected in selected areas of cancer that have been investigated by faculty in Biostatistics and learning the limitations of the various models applied to cancer data. The Biostatistics Training Grant Committee will meet every three months to review trainees and the overall progress of the program. Predoctoral trainees are expected to have completed a baccalaureate degree with a major in mathematics, s t atistics or science. Entry into the program will be based upon undergraduate grades, graduate record examinations, and recommendations. Two predoctoral awards will be given the first year and three predoctoral awards will be given each of the next four years. A postdoctoral award will be given in each of years two through five. The requirements for the postdoctoral award are a Ph.D. in biostatistics or statistics and will be given for a two- year period. Postdoctoral trainees will be expected to take appropriate course work in the areas of biostatistics and/or cancer and will participate in the research of a faculty mentor.
| D'Angelo, Gina M; Weissfeld, Lisa A (2013) Application of copulas to improve covariance estimation for partial least squares. Stat Med 32:685-96 |
