Graduate Degree Programs

• Ph.D. in Statistics
• M.A. in Statistics
• M.S. in Medical Statistics

Program for the Degree of Ph.D. in Statistics

The doctoral program in Statistics is administered by the School of Medicine and Dentistry, and hence School of Medicine and Dentistry regulations apply.

Entering Ph.D. students need undergraduate preparation in mathematics, including mathematical analysis (advanced calculus) and linear algebra, and a year of probability and statistics. Normally, doctoral students are initially considered M.A. candidates; this non-thesis degree can be completed in three or four semesters or, in some cases, in one calendar year. Ph.D. studies consist of additional specialized courses and seminars and supervised research leading to a dissertation. There is no foreign language requirement. Computer expertise is developed in the program.

All M.A./Ph.D. students take a comprehensive examination at the beginning of the second year. Ph.D. students take another written examination at the beginning of the third year. Both examinations cover material in the areas of probability, inference and data analysis.

After beginning research on a dissertation topic, Ph.D. students take an oral qualifying examination, consisting largely of a presentation of a thesis proposal to a faculty committee, the student's Thesis Committee. Upon completion of the dissertation, doctoral candidates present their work at a public lecture followed by an oral defense of the dissertation before the Thesis Committee.

Prior to completing degrees, most students have some publications underway, including some work related to their dissertation research, possibly other methodological work done in collaboration with other members of the faculty, and often some applied papers with scientific researchers in other fields. In general, the Ph.D program requires a minimum of four years of study, with five years of study being more common.

Course work in statistics is concentrated in three areas -- probability, inference and data analysis. Beginning students should expect to spend all of their first year, most of their second year and some of their third year taking formal courses. The balance of time is spent on reading and research. Students entering with advanced training in statistics may transfer credits at the discretion of their advisor. A typical program for an entering student without previous training is as follows:

Year 1: Fall

• BST 401 Probability Theory (4 credits)
• BST 411 Statistical Inference (4 credits)
• BST 464 Applied Linear Regression (4 credits)
• BST 497 Seminar in Statistical Literature (1 credit)
• BST 590 Supervised Teaching (2 credits)
• IND 501 Ethics in Research (1 credit)

• BST 426 Linear Models (4 credits)
• BST 466 Categorical Data Analysis (4 credits)
• BST 497 Seminar in Statistical Literature (1 credit)
• BST 520 Current Topics in Bioinformatics (4 credits)
• BST 590 Supervised Teaching (3 credits)

• BST 477 Introduction to Statistical Software I (0 credits)
• BST 478 Introduction to Statistical Software II (0 credits)

• BST 402 Stochastic Processes (4 credits)
• BST 479 Generalized Linear Models (4 credits)
• BST 450 Data Analysis (4 credits)
• BST 497 Seminar in Statistical Literature (1 credit)
• BST 590 Supervised Teaching (3 credits)

• BST 412 Large-Sample Theory and Methods (4 credits)
• BST 513 Analysis of Longitudinal and Dependent Data (4 credits)
• BST 531 Nonparametric Inference (4 credits)
• BST 497 Seminar in Statistical Literature (1 credit)
• BST 591 Reading Course at the PhD Level (3 credits)

Notes:

• Training in the use of statistical software (BST 477, BST 478) is offered during the first six weeks of the summer as a computing rotation (no formal credit).
• BST 497 Seminar (1 credit) is offered every semester. Ph.D. students are required to register for at least six semesters. This course (1) provides students with experience in organizing, preparing, and delivering oral presentations, (2) introduces students to the process of searching the statistical literature, (3) enables students to acquire knowledge of a focused area of statistical research, and (4) introduces students to the research interests of members of the faculty.
• All Ph.D. students are required to have at least four credits of supervised teaching and/or supervised consulting (BST 590, BST 592).
• Advanced courses listed as BST 511, 512, 550, or 570, for varying numbers of credits, are offered depending on interests of students and instructors. Recent examples include:
  • Monte Carlo Methods and Modeling of Biomedical Dynamic Systems
  • Permutation Tests
  • Frailty Models
  • High-Dimensional Data Analysis
  • Statistical Methods in Epidemiology
  • Smoothing Methods
  • Introduction to ROC Methodology
  • Statistical Inference Under Order Restrictions
  • The Bootstrap, the Jackknife, and Resampling Methods
  • Semiparametric Inference

Program for the Degree of Master of Arts in Statistics

The requirements for entry into the M.A. program are the same as those for entry into the Ph.D. program. The M.A. degree requires satisfactory completion of at least 32 credits and a final comprehensive written examination; no thesis is required. Of the 32 credits, at least 24 must be in departmental courses primarily at the 400 level or above, including at least one semester of BST 497. A typical program of study would include most of the courses shown above in Year 1 of the PhD program. A balanced program is worked out with the student’s advisor. The final comprehensive examination is administered during the summer following completion of coursework.

Program for the Degree of Master of Science in Medical Statistics

The M.S. Program in Medical Statistics is primarily intended for students who wish to follow careers in health-related professions such as those in the pharmaceutical industry and biomedical or clinical research organizations. The program consists of one core year (two semesters) of coursework as well as an internship/applied project (BST 470), which is normally taken in the summer after the core program. There are no thesis or language requirements. The degree requires 32 credit hours consisting of all the 400-level courses listed below; substitutions may be made with approval of the faculty program advisor.

The internship/applied project requirement is met by either completing a statistics internship in industry or by working with medical center investigators on an applied project. Students are required to write a formal report summarizing the findings from their research project. These findings are presented in a public lecture, which is followed by an oral qualifying examination. The student's M.S. committee will then determine the student's qualifications for the M.S. degree. For details on the applied project and oral examination, please refer to the M.S. Guidelines document.

The program is open to students with a substantial background in statistics. A bachelor's degree from an approved program is required.

Prerequisites: Three semesters of calculus, a course in linear algebra (similar to MTH 165), a course in probablility (see STT 201), a course in mathematical statistics (see STT 203), and a course in applied statistics (similar to STT 212) are required.

A typical program for an entering student without previous advanced training is as follows:

Fall

• BST 411 Statistical Inference (4 credits)
• BST 421 Sampling Theory (4 credits)
• BST 464 Applied Linear Regression (4 credits)

Spring

• BST 465 Design of Clinical Trials (4 credits)
• BST 466 Categorical Data Analysis (4 credits)
• ELECTIVE (4 credits)

• BST 470 Internship/Applied Project (8 credits)