Department of Statistics, Operations and Management Science, University of Tennessee

Program Philosophy

The Ph.D. degree is the highest academic degree in the field of statistics. As such, it represents substantial achievement in multiple areas:
i) broad knowledge of the field of statistics,
ii) ability to apply statistics in practical situations to problems of industry, and
iii) ability to develop new statistical methods, as demonstrated by successful completion of a Ph.D. dissertation containing original research on statistical methods motivated by industrial problems.

The curriculum prepares students for achievement in each of the above three areas by a comprehensive program of coursework in statistics, and in the basic functional areas of business.

Ability to apply statistics in practical situations to problems of industry is demonstrated either by prior experience, by working in the Statistical Consulting Center or other appropriate context, or through the student's own consulting experience.

Program Structure:
Overview

Important: Students are generally advised to start in the Fall semester due to our course sequences. Students are seldom admitted in other semesters.

There are no formal admission deadlines (International applicants please consult the special information for international applicants and the admission deadlines of the Graduate School), but financial aid deadlines and availability may be of importance. We strongly suggest applicants to apply before May 15 for Fall semester enrollment.

Program Structure:
The Curriculum

The student must fulfill the general requirements for the Ph.D. degree with a Major in Business Administration established by the Graduate School and the College of Business Administration.

The Statistics Ph.D. Program Director will evaluate the previous training and experience of each candidate to place him or her properly within the Ph.D. program. The candidate for the doctoral degree will complete a minimum of 48 semester hours of coursework beyond the Bachelor's degree, in addition to 24 hours of doctoral research and dissertation.

The curriculum has four essential components: The Business Core (giving the student a broad background in business), Research Methods (giving the student fundamental tools for research), the Statistics Curriculum (giving students their basic knowledge in statistics), and Collateral Area, (giving students substantial knowledge in a field outside of statistics).

The Curriculum:
Business Core

A business core, including a basic knowledge of economics, marketing, management, finance, accounting and logistics, is a requirement of the program. Students who have earned an MBA from an accredited institution, or who have completed the first year of the U.T. MBA program, will be able to waive the business core requirements. Students who do not have MBA experience must complete all or part of these requirements, depending on their backgrounds and individual circumstances. A number of alternative strategies may be considered concerning the number and type of courses required. This will be determined on an individual basis with the approval of the Ph.D. Program Director.

The Curriculum:
Research Methods

The Ph.D. program requires each student to master the tools and methods of statistics research and to demonstrate competence in conducting such research. Specifically, graduates must be able to identify in current literature newly developed methods and tools appropriate to formulate and solve industrial problems. This will include the capability of modifying methods to fit the problem at hand and to explore and understand the limits of applicability of methods and tools.

Required Courses: For admission to candidacy, a student must have completed 9 semester hours in research methods consisting of:

662 Computational Methods in Statistics (3) Introduces and emphasizes up-to-date computational methods in statistics using open architecture interactive computational languages supplemented by other statistical packages with graphical capabilities. Topics covered include introduction to statistical computing, numerical methods for linear models and generalized linear models, nonlinear statistical methods, matrix computations and special matrices, essentials of Monte Carlo simulation, and resampling techniques. Prereq: knowledge of a programming language and 572 or consent of instructor.

663 - 64 Advanced Statistics Theory I, II (3,3) Coverage of advanced topics including univariate models and theory of distributions, the general theory of estimation and the method of maximum likelihood, sufficient statistics, small and large sample efficiency of estimators, confidence and tolerance intervals, testing statistical hypotheses, Bayesian methods and estimation, linear model theory and model selection. These topics are covered within the context of a series of significant problems in applied statistics. Prereq: 564, Math 445.

The Curriculum:
Statistics Curriculum

The following foundational courses (or equivalent as determined by the Ph.D. Program Director) are required:

Statistics Foundation:
Stat 563 - Introduction to Mathematical Statistics (3)
Stat 564 - Theory of Statistical Inference (3)
Stat 566 - Statistical Techniques in Industrial Processes (3)
Stat 567 - Applied Reliability (3)
Stat 571 - Statistical Methods (3)
Stat 572 - Applied Linear Models (3)
Stat 573 - Design of Experiments (3)
Stat 575 - Time Series (3)
Stat 579 - Applied Multivariate Methods (3)
Stat 592 - Internship

Concentration Area (Required Courses):

Stat 691 - Graduate Seminar (3), following the completion of two courses

Also, required: Two courses picked by the Student, subject to approval by Major Professor: Stat 666, 673, 674, 679.

The Curriculum:
Collateral Area (Cognate Area)

In addition to the common program requirements, the Ph.D. program requires each student to master an area of study outside the field of statistics, either inside or outside of the College of Business Administration (CBA). Each student is expected to investigate course offerings across the College and University and develop a program of study related to his/her long-term research, teaching, and work interests or dissertation.

A minimum of 9 semester hours of graduate coursework of 400 (or higher) level coursework in a supporting area is required. Each cognate area will require either an exam or a project.

Collateral areas are, for example, business, engineering, or science.

Although the graduate student will write his/her Ph.D. dissertation in an area of statistics, the research topic may be related to one of the collateral areas. Any student pursuing such options should discuss them with his/her faculty advisor and obtain the approval of the Statistics Ph.D. Program Director.

The following are illustrative of the types of courses recommended. Other courses may be appropriate.

Collateral Area Courses:
Inside the CBA:

Economics
Econ 511-12: Microeconomic Theory (3,3)
Econ 581: Mathematical Methods in Economics (3)
Econ 582: Elements of Econometrics (3)
Econ 583: Elements of Econometrics II (3)
Econ 683: Time Series Econometrics (3)

Management Science
MS 531: Mathematical Programming (3)
MS 532: Stochastic Models in Management Science (3)
MS 533: Computational Mathematical Programming (3)
MS 681: Special Topics (3)

Finance
Fin 511: Strategic Management for Creation of Financial Value (3)
Fin 525: Investment Analysis & Portfolio Management (3)
Fin 641: Seminar in Finance I: Capital Markets (3)
Fin 652: Seminar in Asset Pricing and Markets (3)

Operations and Management Science
OMS 540: Statistics and Operations Management (3)
OMS 541: Operations Management (3)

Collateral Area Courses:
Inside the College of Arts and Sciences:

Computer Science
CS 541: Database Management System (3)
CS 551: Pattern Analysis (3)
CS 552: Image Analysis (3)

Mathematics
Math 445-46: Advanced Calculus I, II (3, 3)
Math 453: Matrix Algebra II (3)
Math 471: Numerical Analysis (3)
Math 525: Statistics I (3)
Math 526: Statistics II (3)

Collateral Area Courses:
Inside the College of Engineering:

Electrical Engineering
EE 471: Introduction to Pattern Recognition (3)
EE 472: Introduction to Digital Image Processing (3)
EE 504: Random Process Theory for Engineers (3)

Industrial Engineering
IE 515: Advanced Production and Inventory Systems (3)
IE 527: Lean Production Systems (3)
IE 556: Data Mining in Engineering and Manufacturing (3)

Qualifying and Comprehensive Examinations

Qualifying exams are taken to demonstrate appropriate competence in the Statistics Foundation outlined in the curriculum above. They consist of a Theory Qualifying Exam and a Methods Qualifying Exam, both of which must be passed at a Ph.D. level. These examinations cover selected topics out of the Statistics Foundation. Students may be allowed to retake an exam a second time if they do not pass it on the first attempt.

The purpose of the qualifying examinations is to provide the opportunity for students to review, deepen and strengthen their understanding of key statistical principles and techniques.

After the student has passed the Theory Qualifying Exam and the Methods Qualifying Exam at the Ph.D. level, and has completed substantially all coursework except dissertation hours, the student must take and pass the comprehensive exam before registering for dissertation hours.

The statistics comprehensive examination covers important dimensions of statistics methods and theory.

The comprehensive exam involves a written component and an oral component based on the student's intended direction for research, a set of broad, comprehensive questions are formulated. This written component will either be in class (6 hours or longer) or take home, as determined by the Statistics Ph.D. Program Director. After this written component of the examination is passed, the oral component is given. Students may be allowed to retake an examination component if they do not pass it on the first try.

The Statistics Ph.D. Program Director is responsible for overseeing the design, administration and evaluation of the comprehensive exam. It is the Director's duty to set exam dates at least eight weeks in advance and to form an examination committee. The faculty members of this committee will be responsible for developing exam questions and constructing, administering and grading the examination. Those faculty serving on the examination committee will be announced by the Ph.D. Program Director at least six weeks before the exam.

Students typically are informed of the outcome of the exam within two weeks of the date the exam is given. Grades are determined by majority vote of the examination committee and consist of pass with distinction, pass, fail or pass with qualifications. The latter rating may be used by the faculty to require further work from students exhibiting marginal performance. Such work is determined by the faculty and may include additional coursework, independent study, or preparation for the delivery of designated written or oral assignments.

Admission to Candidacy

Students who have maintained at least a "B" average in course work, successfully completed the comprehensive examination, and secured acceptance by their doctoral committee of a dissertation research proposal may apply for admission to candidacy. Admission to candidacy must be approved at least one full semester prior to the date the degree is conferred (admission in the Fall semester permits graduation in the following Spring semester).

The application for admission to candidacy must include a list of all courses taken for the degree. Graduate courses accepted from other institutions must be clearly identified, as well as the date of acceptance of the research proposal by the doctoral committee. The application must be signed by the student's doctoral dissertation committee, the Statistics Ph.D. Program Director, the College of Business Administration Director of Graduate Studies, and the College of Business Administration Dean's office, before submission to the Graduate School for approval.

Doctoral Dissertation in Statistics
and Related Requirement

General Definition

The dissertation in statistics represents the culmination of an original research project completed by the student. The organization, method of presentation, and the content of dissertation are important in conveying to others the results of such research.

General Requirements

1.	After completing the required course work toward the Ph.D., the student must pass the doctoral comprehensive exam, and have his/her dissertation proposal accepted by his/her Doctoral Committee to gain admission to candidacy for the Ph.D. degree. The Doctoral Committee must include at least four faculty members, with at least two from the Statistics Department and at least one from another department. At least three of the doctoral committee members must be approved to direct doctoral research.
2.	The Graduate School requires a minimum of two consecutive semesters including summers in residence, defined as full-time registration on the campus where the program is located. During this time the student should make substantial progress toward the degree. The Department Head must certify on the Admission to Candidacy form that the residence requirement has been met. The student must register continuously for dissertation hours (a minimum of 3 hours per term) from the time the doctoral research proposal is approved, admission to candidacy is accepted, or registration for dissertation hours is begun, whichever comes first, with the sole exception of time spent on internship. This includes summer sessions and the semester in which the dissertation is approved and accepted by the Graduate School. A minimum total of 24 dissertation hours is required before the dissertation will be accepted. A student who will not be using faculty services and/or university facilities for a period of time may request leaves of absence from dissertation research up to a maximum of six semesters. The request will be considered by the Graduate School upon request of the Department Head.
3.	A student should be registered for the number of dissertation hours representing the fraction of effort devoted to this phase of the candidate's program. Thus, a student working full time on the dissertation should register for 9 hours of 600 level courses per semester.
4.	A dissertation must be written according to the regulations in the UT Knoxville Guide to the Preparation of Theses and Dissertations (current edition). The dissertation must be approved by the faculty advisor before the defense of dissertation is scheduled. The completed dissertation must be submitted to all committee members at least two weeks prior to the oral examination.
5.	The doctoral committee conducts the defense which must be scheduled through the Office of Graduate Admissions and Records. The examination is announced publicly and is open to the University community at large, and covers the dissertation and coursework submitted to satisfy degree requirements. The possible outcomes are: (i) Pass the defense, dissertation acceptable. (ii) Pass the defense subject to making minor changes in the dissertation as specified by the doctoral committee. A re-examination is not required. (iii) Fail the defense. The student will be given instructions by the doctoral committee on the actions necessary to correct the deficiencies in the dissertation. A re-examination may be scheduled no sooner than the following semester. Failure to pass the examination on the second attempt will result in the student being dismissed from the Ph.D. degree program. The doctoral committee records its decision on the Graduate School final examination form. Two copies of the dissertation must be submitted to and accepted by the Graduate School. Each copy must include an approval sheet, signed by all doctoral committee members, which certifies to the Graduate School that they have examined the final copy and found that its form and content demonstrate scholarly excellence.

Ph.D. Dissertation Proposal

Each Ph.D. student in Statistics, after completing coursework, passing the doctoral comprehensive examination, and selecting the dissertation advisor and committee members according to the above requirements, must prepare a comprehensive proposal in his or her area of concentration.

A typical dissertation proposal format might include all or some of the following:

Concise statement of the problem
Originality of the proposed topic
Review of prior related work
Methods to be used to accomplish the work Major expected accomplishments
Estimated schedule
Satisfaction of residency requirement
Bibliography

The faculty advisor arranges a meeting with doctoral committee members for oral examination and proposal examination, so that the student can respond to committee members' questions about the proposal. If the doctoral committee does not approve the dissertation proposal, the student may be given up to three opportunities to rectify the situation. Typically, the approval process of the dissertation proposal will be within one year of completion of the coursework and the doctoral comprehensive examination.

Ph.D. Dissertation

A student must complete a doctoral dissertation representing original research driven by applications. The oral Dissertation Examination is the defense of the Ph.D. dissertation.

In addition to his/her dissertation defense, the student is encouraged to submit articles from his/her research to reputable journals for publication.

Graduate School Time Limit

The comprehensive examination must be taken within five years, and all requirements (including the dissertation) must be completed within eight years, from the time of a student's first enrollment in a doctoral degree program.

Incomplete Grades, Deadlines and Readmission

The Graduate Catalog defines an I (Incomplete) grade as follows: "I (no quality points), a temporary grade indicating that the student has performed satisfactorily in the course but, due to unforeseen circumstances, has been unable to finish all requirements. An I is NOT given to enable a student to do additional work to raise a deficient grade. All incompletes must be removed within one semester, excluding the summer term. If a supplementary grade report has not been received in the Office of Graduate Admissions and Records at the end of the semester, the I will be changed to an F. The course will not be counted in the cumulative grade point average until a final grade is assigned. No student may graduate with an I on the record."

Although most students should be able to complete all requirements for a Ph.D. within 4 years (3 years with the appropriate course work from a Master's in Statistics, for example), all students are given up to 8 years from the time of entering the program to complete the degree. Failure to complete the degree within 8 years will result in termination from the program. After 8 years, content of courses taken may have grown out of date. Thus, doctoral course credit earned in years prior to the 8-year limit is forfeited. An extension may be requested under justifiable circumstances. The Graduate School must give final approval of an extension.

Continuous Enrollment Requirement

The Graduate Council of the University adopted a policy in 1977 concerning graduate students whose education is interrupted for a period of more than three semesters. "A student who has not attended the Graduate School at the University of Tennessee, Knoxville for more than three semesters must apply for readmission. Since readmission is not automatic, a readmission application should be submitted at least two weeks prior to the desired reentry date. A student who has attended another accredited institution since enrollment at UT must submit an official transcript showing all course work and any degrees earned at that institution. The student will be notified when the application is received and when action has been taken by the department and the Graduate School. If readmission is denied, the student may receive graduate credit for work taken during the initial semester. However, future registration will not be permitted until the student is fully readmitted to the Graduate School."

Example Ti metable

For Students Entering with an MS in Statistics (including Advanced Calculus)

Prerequisites SUMMER

Acc 201-202 (6 hours)

1st Year

FALL	SPRING	SUMMER
Econ 311 or AgEcon 505	Stat 666-Advanced SPC	Stat Consulting (no credit)
Fin 301	Acc 521	Collateral Elective
Mkt 501	Stat 691 Graduate
Stat 673-Advanced Topics in Design of Experiments	Seminar in Applied Statistics
12 hours	9 hours	3 hours

2nd Year

FALL	SPRING	SUMMER
Stat 663 - Advanced Stat Theory I	Stat 664 - Advanced Stat Theory II	Statistics 592 - Internship
Collateral Elective	L/T 501
Stat 662-Computational Methods in Statistics	Mgmt 571
9 hours	9 hours	3 hours

3rd Year

FALL		SPRING		SUMMER
Dissertation	(6)	Dissertation	(9)	Dissertation	(9)
Collateral Elective
9 hours		9 hours		9 hours

Research Methods:	9 hrs required
Collateral Area:	9 hrs required
Concentration Area:	12 hrs required

Educational Enrichment Activities

Students can anticipate that considerable effort will go toward experiencing and understanding real world statistics through working with real data, solving real problems, and working to improve administrative and industrial processes. As a consequence participation in a number of activities such as those listed below is to be considered part of the educational experience.

Attending talks outside of the Department, (e.g., at the WATTEC Conference) where the focus is on the use of information to address a problem.
Working to flowchart and improve administrative systems (e.g., at Tennessee Valley Authority or U.T.).
Touring plants with an eye toward finding a project within the plant environment, the completion of which would benefit both the student and the plant.
Attending seminars on current industrial problems (seminars given by consultants or past students).
Assisting Center for Executive Education consultants on field trips, to learn more about the consulting environment, and to see plants from the inside.
Attending Center for Executive Education functions (lunches, dinners, seminars) to have personal contact with industry leaders and workers and to appreciate the utility of statistics from their point of view.

Orientation

Each entering Ph.D. student will attend an informal orientation program conducted by statistics faculty and current Ph.D. students concentrating in statistics. The orientation generally includes discussion of:
1) objectives of the Ph.D. program,
2) expectations of the faculty,
3) a sense of the skills and values needed to become a contributor to our targeted industrial market, and
4) other topics related to personal and professional development. The purpose of this orientation is to set the standards and expectations for the Ph.D. program and to give entering students the opportunity to become acquainted with students and faculty in the department.

Travel and Other Professional Activities

Ph.D. students concentrating in statistics are strongly encouraged to attend annual conferences within the discipline to further their professional development. The department will make every effort to fund at least one trip to a national conference that is in conjunction with the student's presentation of a paper at the conference. Attending conferences with an industrial statistics focus and making contact with practitioners in the field is strongly encouraged.

The Dean of the College of Business Administration allocates a portion of CBA private funds each year to this function as part of the annual spending plan. The Dean will consider requests for general CBA funding for those applications approved at the departmental level. Cost sharing by both the student and the department in the award selection is strongly encouraged.

Corporate Partnership Strategy

The Department of Statistics faculty strive to identify industrial advisors to help give direction to the Department. We seek individuals who have their corporation's backing for the task, so that a partnership with the corporation can become a reality. Corporate partners not only help shape the future of the Ph.D. program, but can be expected to provide internships for students and be prime suppliers of problems for dissertation efforts.

Seminar Series

The seminar series in statistics is an important part of the intellectual life of the Department of Statistics. Students in the Ph.D. program are required to attend these seminars on a regular basis.

Prior to graduation, every Ph.D. student will give a minimum of two seminars related to the content of their dissertation.

Faculty Mentors

All students entering the Ph.D. program are matched with a faculty mentor to advise them on course selection and other matters relevant to their educational and career progress until such time as they obtain a major professor.

As students progress and get closer to the point of writing the dissertation, they will typically select a major professor and obtain his/her consent to serve in that role. At this point, the major professor begins to function as the mentor for that student. The major professor must be approved by the Graduate Council to direct doctoral research.

Administrative Matters

Termination From Program

It is fully expected that a student who enters the program will complete the degree requirements. The faculty is committed to achieving that goal with every student. However, it may become apparent that sufficient progress is not being made despite faculty and student efforts. In this case, termination decision may be considered.

The College of Business Administration requires that a Ph.D. student whose overall GPA falls below 3.0 shall be placed on probation. A student on probation shall be dropped from the program unless his/her GPA is 3.0 or higher at the end of the probationary period (defined as the next semester's coursework).

Other circumstances which may cause termination from the program include 1) failure to complete the comprehensive examination successfully or 2) failure to complete a dissertation within three years after admission to candidacy.

Appeals Process

The student handbook, HillTopics, published and distributed annually, contains statements of UTK standards of conduct and of all disciplinary regulations and procedures. Normally, grievances should be handled at the departmental level through the student's advisor or major professor, then the Ph.D. Program Director, and if necessary the department head. Further appeal may be made to the Dean of the respective college, the Dean of the Graduate School, the Graduate Council, and the Chancellor. A copy of the Appeals Procedure is available in the Office of Graduate Admissions and Records.

Teaching and Research Responsibilities

Teaching

One of the roles of the applied statistician in industry is that of teacher. This teaching may take the form of one-to-one (or one-to-a-few) interactions with clients, or it may involve formal courses offered to employees. Ph.D. students are given a variety of opportunities to develop and practice the skills needed to fulfill the role of teacher. These include consulting experience, and formal "classroom" teaching. Formal classroom teaching opportunities usually arise as part of the responsibilities of graduate assistantships. These may take several forms. A student may serve as the assistant to a faculty member in the administration of a course in statistics. Or, from time to time, as the need arises, a graduate student may be allowed to take full responsibility for planning, conducting, and administering an undergraduate course in statistics. Alternatively, a student may assist in conducting one of the courses offered through the Center for Executive Education (CEE). The scheduling of teaching assignments depends on course demands and variations in departmental resources from year to year. Every effort is made to allocate available teaching assignments so as to satisfy the needs and requests of students. Further information can be found in the Guidelines on Administration of Graduate Assistants, available from the Office of the Dean of the Graduate School.

Research

Not all problems brought to the applied statistician in industry are "textbook" problems that can be solved by applying standard methods available in computer packages. Consequently, applied statisticians must be able to modify and extend existing methods and even develop new methods as needed. The culmination of the Ph.D. program is the dissertation in which students demonstrate their originality and creativity in developing statistical methods. The dissertation is discussed in detail under "Doctoral Dissertation in Statistics." Students may also acquire research experience as part of their graduate assistantship duties by serving as research assistants on faculty research projects. The availability of such assignments varies from year to year, and such assignments are made after considering the needs both of the student and the faculty member involved.

Financial Aid

Since completion of the Ph.D. program normally requires a minimum of three years for students already having an M.S. degree in Statistics, and four years for those without a M.S. degree in Statistics, most students should come to the University of Tennessee prepared for three to four years to be spent in residence.

Students often need financial support for part or all of the duration of their stay. Support is available from a variety of sources, including: i) teaching assistantships with the Department of Statistics, ii) research assistantships with the Department of Statistics, iii) other work opportunities at the University of Tennessee, including the Management Development Center and Statistical Consulting Center, or iv) the student's current employer. Such support may be granted to a student from the above sources based on funds availability and student performance for from one to four years, renewable annually.

Students interested in obtaining support should so indicate at the time they apply for admission to the Ph.D. program.

Academic Honesty and Professionalism

The Statistics Department adheres to established professional standards of honesty, scholarship and professionalism. Absolute honesty is expected, and all students are subject to university-wide policies on this subject.

Faculty

The Department of Statistics has 12 full-time, tenured faculty. Of these, eight are approved to teach courses at the 600 level, and nine are approved to direct Ph.D. dissertations.





	Copyright ©2006 The University of Tennessee · College of Business Administration · Knoxville Tennessee 37996