Ph.D. Course Selection, Year-to-Year Progression, and Examinations

The material here is an informal step-by-step walk through the program that should be fully understood by all students and advisors. This should be read in conjunction with the Formal Requirements of both the department and the Graduate School. Most importantly ...

FIRST YEAR:

Helpful info and resources for new students is available at the Graduate School first year student resource site. Some key specifics for Statistical Science students are:

• Select first year courses. Students with more advanced backgrounds may select more advanced courses, after detailed consultation with their advisors. Some students (less commonly) take a non-STA course (see Duke course schedule). A normal course load is 3 or 4 courses (3 for students with TA/RA duties). All course selections must be approved by the advisor. No courses are required of any student, and advanced courses, as well as independent study, may be taken by incoming students, again subject to discussion and approval through the advisory process. Core STA courses most incoming students select from are:

  Fall	Spring
  205, 213, 253, 290, MTH203, ...	215, 244, 291/2, 293/4, 395, ...

  All students are encouraged to register for STA 395 Readings in Statistical Science each semester. This is an informal seminar with talks on research given by students and faculty.

• Some PhD students enter the program with background in mathematics that includes preparation in analysis at or above the level of the Duke course MTH 203 Basic Analysis and so do not take the analysis course. Other students take MTH 203 concurrently with STA 205 and 290 in the first semester in order to ensure mathematical preparation.
• Most PhD students enter the program with statistical backgrounds above the introductory level so will typically not take STA 213. Some students do, however, take or audit STA 213 as a "refresher".
• Non-native English speakers take English tests (written and oral) just prior to the start of fall semester. Students not passing at sufficiently high levels will register for one or more English courses for credit, in fall or spring, or both. See the details under the English language proficiency tab at the Graduate School Ph.D. details page.
• Register for and attend the required Responsible Conduct of Research (RCR) program - including a required retreat just prior to semester and attendance at several lectures during the year. See the details under the RCR tab at the Graduate School Ph.D. details page.
• Primary TA/RA assignments are expected to demand an average over semester of 10-12 hours per week (departmental expectation), sometimes more and sometimes less. (In addition to regular TA/RA appointments, a student may - with approvals of the DGS and advisors - engage in additional, occasional paid work as part-time TA or grader, or in statistical consulting, for up to 5 hours per week.) Time committed to TA/RA/consulting appointments is subject to the absolute limit (per Graduate School policy) of no more than 19.5 hours/week.
• In early Fall, first year U.S. students that have not earlier applied for available external, national fellowships should seriously consider applying U.S.fellowships, including those of the National Science Foundation, the National Physical Science Consortium, the Hertz Foundation, and others. Eligibility is for U.S. students applying to graduate school and for first year PhD program students only.
• First Year Exam (FYE), in early May during exam week. Three hour written exam (covering material within the scope of STA 205, 213, 215, 244, 290) and 24 hour take home applied statistics exam.
• Satisfactory Progress: in courses, TA/RA duties, applied statistics, FYE.
• Summer: Vacation, internship, RA at Duke, etc. Consult advisors and other professors as well as students about intership and RA opportunities, and monitor email announcements.

SECOND YEAR:

• Select second year courses and more advanced courses, including STA courses and, in some cases, courses from other departments. All course selections must be approved by the advisor. Core STA courses to select from are:

  Fall	Spring
  214, 216, 291/2, 293/4, 390, 395, ...	376, 291/2, 293/4, 390, 95, ...

  All students are encouraged to register for STA 395 Readings in Statistical Science each semester. This is an informal seminar with talks on research given by students and faculty.

• Discuss potential Prelim Exam topics with advisor, and with other professors, often taking STA 291/2 Independent Study in one or both semesters to develop expertise in an area in preparation for the Prelim.
• Propose Prelim Committee. The Committee must be approved by the Graduate School at least 60 days prior to the scheduled Prelim Exam (though can be changed arbitrarily, with approval, once established). Students should aim to have the initial Committee and named chair/Prelim advisor, approved by the end of February, and schedule the exam in early May.

  The Committee proposed must follow these Graduate School and departmental regulations on Committee membership for what will become the Thesis committee following a successful Prelim.

• File intent to earn MS in Statistical Science "en route" to Ph.D., if that is the intent (most students gain the MS en route).
• Prelim Exam. Some students take the Prelim earlier; some take it later in summer session after the second year. In some cases the Prelim can be deferred until fall semester of the third year.

  Prelim goals and format: The Prelim assesses the competency of the student in the broad field of statistical science, and preparedness for individual research in statistical science. With agreement of the advisor and Prelim Committee on topic(s), the student will make a seminar-style presentation to the Committee, with questions from the Committee on the material presented, and associated material, as well as any other questions on statistics and related topics the Committee members choose. The Prelim presentation material may comprise review and critique of published research (in one or two papers from the literature), or -- more commonly -- work the student has developed in mentored independent study or as RA. It is common, but not required, that the student presents material s/he has been exploring that is already identified as a likely Ph.D thesis research topic.

• Satisfactory Progress: in courses, TA/RA duties, applied statistics, Prelim.
• Submit the required annual progress form to the DGS Assistant before April 15.
• Summer: Vacation, internship, RA at Duke, possibly attend research conferences, etc.

THIRD YEAR:

• Thesis research.
• The Prelim committee becomes the Thesis Committee. Commitee membership may be changed at any time, depending on the focus of research and advisory relationships between the student and involved faculty. It is usual that the primary Ph.D. advisor also be the Thesis Committee Chair (this is not required, but is almost always the case). The Committee membership, governed by Graduate School and departmental requirements, comprises the Statistical Science Advisor/Chair, at least 2 other professors in Statistical Science and 1 professor from another related department or program. Full details are covered in the summary of Graduate School and departmental regulations on Committee membership.
• Students may take a small number of more advanced courses in STA topics and sometimes a course in an area of immediate relevance to their thesis research from other departments. As in years 1 and 2, course selection must be approved by the thesis advisor. Also, register for STA 291/2 Independent Study and Research each semester with thesis advisor.
• All students are encouraged to register for STA 395 Readings in Statistical Science each semester. This is an informal seminar with talks on research given by students and faculty.
• Meet regularly (weekly) with primary advisor and with committee members, informally, throughout each semester.
• Attend one or more research conferences, presenting posters or talks on thesis research. Post-Prelim students are eligible to apply for a grant for partial financial support from the Graduate School and department for one research conference each year - the application form is available at the Graduate School's financial support web page. Students should always seek additional support from conference organisers and faculty advisors who may have relevant grant funds.
• Present progress reports in informal seminars (often in STA 395) to students and faculty.
• Write thesis. Writing in thesis format should begin as soon as research begins. Keep detailed written (and graphical, and software) records of all aspects of R&D, and evolve records for talks, future papers and thesis presentation. You cannot start writing early enough; beyond the discipline and experience of routine writing, you will find that the act of writing, and editing, your work routinely will help to generate new ideas, rediscover old ones, and to identify open questions and the need for more work on aspects of research you thought you had completed ...
• Satisfactory Progress: in courses, TA/RA duties, research towards thesis.
• Submit the required annual progress form to the DGS Assistant before April 15.
• Summer: Vacation, internship, RA at Duke, possibly attend research conferences, etc.

FOURTH YEAR +:

• Thesis research, and other details as listed above in Year Three. Committee changes may be made if research directions change and to respond to developments in your research interests.
• Statistical Science students are expected to progress to thesis completion and defense within 5 years, and many complete and defend in 4 years (or during the summer following the 4th year). See also the Duke Graduate School policy on timelines and satisfactory progress.
• All students are encouraged to register for STA 395 Readings in Statistical Science each semester. This is an informal seminar with talks on research given by students and faculty.
• Satisfactory Progress: In research towards thesis, presentation of dissertation and oral thesis defense.
• Submit the required annual progress form to the DGS Assistant before April 15.

DEFENSE & FINAL TIMELINE:

It is the responsibility of the student and advisor to accord to the details of thesis format and the timeline for thesis format checks, approval, submission and post-defense deposit, per the Graduate School policy. This includes:

• File the online "declaration of intent" form at least one month before the defense date, and attend to other details of Graduate School time limits related to defense and graduation.
• The advisor and student should agree that the thesis is essentially complete, apart from final touches, several weeks prior to defense.
• The student must submit the electronic version of the thesis online and, once the Graduate School has confirmed the submission, schedule an appointment for the required "Examination Card Interview" at the Graduate School, at least two weeks before scheduled defense; see the Graduate School thesis & dissertation requirements. Once this interview is scheduled, the student must request that his/her advisor send an email or written letter to the Graduate School confirming that the thesis is complete and ready to defend; this must be sent prior to the interview.
• A full copy of the thesis must be given to each member of the committee at least one week in advance of the defense, and preferably sooner.
• After a successful defense, the electronic thesis submission can be updated with the final thesis if corrections have been made; see the Graduate School requirements on final thesis deposit time limits.