Department of Preventive Medicine

Division of Biostatistics

Ph.D. in Biostatistics

This program is designed to produce a biometrician who has a deep knowledge of statistical theory and methodology, and who is able to apply this knowledge creatively to the solution of statistical problems arising in the biological and health sciences. The program requires a solid core of courses in the theoretical aspects of statistics, and a wide range of courses in the methodological aspects of statistics. Graduates of this program are prepared for careers in teaching, research, and consulting in biostatistics. This program encourages applicants who have undergraduate degrees in mathematics, statistics, biostatistics, or other related fields to apply.

The goal of the Ph.D. program in Biostatistics (BIOS) is to train biostatisticians who will have in-depth knowledge of statistical theory and methodology and the ability to apply this knowledge creatively to statistical problems in the biological and health sciences. At the same time, the increasingly specialized lines of biomedical and population research each include unique study designs, modes and complexities of data collection, as well as analytic, computational and methodological challenges.

By its nature, biostatistics is an applied discipline involving the use of statistical and mathematical ideas in service of medical research in such areas as epidemiology, genetics, basic biology, and clinical effectiveness. As the areas of application and the mathematical tools are rapidly evolving, however, methodological research is also a central focus of the work of a biostatistician. The goal of our Ph.D. program is therefore to provide

  • a solid grounding in the standard techniques of statistical analysis and study design
  • expertise in the ways of conducting statistical methods research, and
  • extensive experience in applications to one or more major areas of medical research.

In the process, it is expected that the candidate will have made a substantial contribution to novel statistical methodology, generally in the form of one or more publications in appropriate statistical journals. He or she should also have had a broad exposure to multiple areas of application so as to be able to communicate effectively with biomedical researchers, being able to understand the basic concepts of problems being studied, to frame the scientific questions being asked in statistical terms and identify appropriate solutions, to explain difficult mathematical concepts to non-mathematical investigators, and thereby to be able to collaborate effectively on all aspects of a project, from concept, to grant applications, data collection, data management, data analysis, and ultimately paper writing.

Specifically, our program offers training in the following areas:

  1. the principles of mathematical statistics, including distribution theory and statistical inference
  2. the standard techniques for analysis of continuous, categorical, time-to-event, univariate and multivariate, cross-sectional and longitudinal data, etc.
  3. particular emphasis on the types of data most commonly encountered in epidemiology, genetics, and clinical trials
  4. hands-on experience in statistical consulting settings, including problem identification, finding appropriate solutions, and communicating these ideas to investigators
  5. modern methods for statistical computation, including standard packages, programming, and techniques for dealing with computationally intensive problems
  6. methodological research techniques, including computer simulation and distribution theory

We aim to position our graduates to be competitive for academic positions in departments of biostatistics or similar fields, in public service (e.g., research positions at NIH, regulatory agencies, or public health offices), or in the private sector (e.g., biotechnology).

Preparation for the Screening Examination

Prior to the Screening Examination a mentor who will serve on the Guidance Committee must be identified. The Screening Examination will consist of an Applied portion and a Theory portion. Examination questions for the Applied portion will be drawn from the Core courses for the MS degree in Biostatistics. A portion of the examination questions for the Theory portion of the Screening Examination will be drawn from Core Courses. Additional examination questions may be drawn from other Biometry Core/Elective courses (such as PM518b or PM552) or from electives representing the research focus of the student. The content area of the Theory portion requires approval of the Biostatistics/Epidemiology program Directors. A student failing the Screening Examination may be given a second opportunity to retake either or both portions of the Screening Examination. Students failing the examination for the second time will terminate with the MS degree upon satisfactory completion of 33 units and an acceptable thesis.

Preparation for the Qualifying Examination

A formal Guidance Committee will be formed, consisting of five faculty members; up to four from within the Department of Preventive Medicine (one designated as Chair), and at least one from an outside department offering a Ph.D. degree. The Committee will recommend courses in preparation for the Qualifying Examination based on the student’s research and dissertation topic. At least two semesters of PM 610 are recommended. Prior to the Qualifying Examination, a written draft of the background and methods sections of the proposed dissertation must be submitted to the Guidance Committee. The dissertation should be oriented toward a theoretical-methodological application to a problem area in the biological or health sciences.

Qualifying Examination

An oral Qualifying Examination will ascertain the student's competence in communicating the rationale of the research proposal and plans for conducting the dissertation research. In addition, the Qualifying Examination will test the student's integration of knowledge in biometry and the health sciences.

Dissertation and Oral Defense

Upon passing the qualifying examination, the Ph.D. candidate and his/her chair will recommend a three-member Dissertation Committee. The dissertation should be completed within two years of the Qualifying Examination. The Dissertation Defense is an oral examination based on a rough draft or final version of the dissertation. The Dissertation defense is administered by the Dissertation Committee and other faculty and students are invited to attend.

University of Southern California University of Southern California