Graduate Student Handbook

The Ph.D. in Statistics

The following guidelines identify the basic structure of the Ph.D. in Statistics at the University of Arizona.  At the core of the program is a fundamental grounding in both statistical theory and methodology; however, extensive flexibility via course electives allows students to tailor their final programs of study to their own interdisciplinary interests.  The student’s advisor, along with the Program director, are available to discuss individual selection of these electives.  It is GIDP policy that the student holds final responsibility for being aware of and responding to all GIDP and Graduate College policies, requirements, formats, and deadlines as they pertain to progression towards and completion of the Ph.D.

The basic requirements for entrance into the Ph.D. program are:

(a) An M.S. in Statistics, Biostatistics, Mathematics, or Applied Mathematics, or an advanced degree in a field that makes significant use of quantitative methods.
(b) Scores on the Graduate Record Examination (GRE -- general test only) that exceed 75th percentile Quantitative and 50th percentile Verbal.
(c) For students applying from non-English-speaking countries and who do not have a degree from a U.S. institution of higher learning, scores on the Test of English as a Foreign Language (TOEFL):  pBT score above 600 and a pBT Section 1 score above 60, or iBT score above 100 and an iBT Section 2 score above 27.  The International English Language Testing System (IELTS) examination can be submitted in lieu of the TOEFL. The minimum IELTS score requirement for admissions consideration is a score of 7, with a score of no less than 6 on any individual module.

A minimum of 74 units of coursework is required, made up as follows:

1.  Core Ph.D. Courses; minimum 33 units as follows:

            18 units from the set of Core M.S. Statistics courses:
                        Theory of Probability (MATH 564)
                        Theory of Statistics (MATH 566)
                        Advanced Statistical Regression Analysis (MATH 571A)*
                        Design of Experiments (MATH 571B), or
                                    Research Design & Analysis of Variance (PSYC 507C)
                        Data Management and the SAS Programming Language (EPID 576D)
                        Statistical Consulting (EPID 685)

            along with an additional set of 17 units of Core Ph.D. Statistics courses:
                        Real Analysis (MATH 523A)
                        Probability Theory (MATH 563)
                        Theoretical Statistics (MATH 567A)
                        Theory of Linear Models (EPID 684A)*
                        Simulation Modeling and Analysis (SIE 531)
                        Scientific Writing Strategies, Skills & Ethics (PHCL 595B)

Students entering with an M.S. in Statistics or an M.S. from a comparable field may petition for waiver of and/or credit for any of the units from the Core M.S. course list, up to a maximum of 12 units.  These credits are referred to by the Graduate College as Transfer Coursework.  Transfer coursework must be approved in advance by the Program director and by the Graduate College, via the College’s Transfer Credit Form.  Approved transfer coursework must be listed as such on the student’s Doctoral Plan of Study (DPOS; see below).

2.  Additional Elective Courses; minimum 12 units from any of the following (a maximum of 6 units may come from STAT 599 and/or STAT 900):
            Advanced Geographic Information Systems (GEOG 520/RNR 520)
            Advanced Quality Engineering (SIE 606)
            Advanced Statistical Natural Language Processing (LING 582)
            Advanced Topics in Biostatistics (EPID 676)
            Applied Biostatistics Analysis  (EPID 576C), or
                        Advanced Topics in Biological Statistics (ECOL 581), or
                                    Applied Biostatistics (RNR 613/ENTO 613/INSC 613)
            Applied Econometric Analysis (AREC 549/ECON 549)
            Applied Time Series Analysis (GEOS 585A)
            Bioinformatics and Genomic Analysis (ECOL 516/MCB 516)
            Biostatistics for Research (EPID 576B)
            Categorical data analysis (STAT 574C/MATH 574C)*
            Clinical Trials and Intervention Studies (EPID 675)
            Detection and Estimation in Engineering Systems (ECE 639)
            Econometrics (ECON 522A), or
                          Advanced Applied Econometrics (AREC 559)
            Econometrics (ECON 522B)
            Engineering Decision Making Under Uncertainty (SIE 522)
            Environmental Statistics (STAT 574E/MATH 574E)*
            Fundamentals of Optimization (SIE 545)
            Game Theory and Mathematical Programming (MATH 579), or
                        Game Theory (SIE 543)
            General Linear and Mixed Effects Models (EPID 684B)*
            Generalized Linear Modeling (EPID 684C)*
            Graphical Exploratory Data Analysis (PSYC 597G)
            Independent Study (STAT 599)
            Information and Noise in Quantum Optics and Photonics (OPTI 528)
            Introduction to Geostatistics (GEOG 574/MATH 574)
            Multivariate Analysis in Management (MGMT 582D)
            Multivariate Methods in Educational Research (EDP 646A)
            Neural Networks (ECE 631)
            Numerical Analysis (CSC 575A/MATH 575A)
            Principles of Image Science (OPTI 637)
            Queuing Theory (SIE 525)
            Reliability and Quality Analysis (A ME 574)
            Research (STAT 900)
            Social Statistics (SOC 570B)
            Spatio-Temporal Ecology (ECOL 518)
            Statistical Genetics for Quantitative Measures (ANS 513/EPID 513/GENE 513)
            Statistical Mechanics (PHYS 528)
            Statistical Methods in Psychological Research (PSYC 507B), or
                        Educational Tests and Measurements (EDP 558)
            Statistical Natural Language Processing (LING 539)
            Statistical Packages in Research (EDP 548)
            Stochastic Differential Equations (MATH 565C)
            Stochastic Methods in Surface Hydrology (CE 655/HWR 655), or
                        Applied Stochastic Processes (MATH 568)
            Stochastic Modeling I (SIE 520)
            Stochastic Processes (MATH 565A)
            Stochastic Processes (MATH 565B)
            Survival Analysis (EPID 686)
            Theoretical Statistics (MATH 567B)
            Theory of Measurement (EDP 658A)
            Theory of Measurement (EDP 658B)
            Time Series Analysis (MATH 562)
            Topics of Optimization (SIE 649)

Students must meet all prerequisites for any elective courses they wish to undertake, or must secure instructor permission prior to registering for the course.  Courses may be added to or removed from this list by action of the GIDP Curriculum Committee, after approval by the GIDP Executive Committee.

Where needed to suite a particular or specialized need in an individual student’s program of study, petition may be made to the GIDP Executive Committee for approval of a course not listed above for use as an elective.  The student must be in good standing and be enrolled in the Statistics GIDP.  The burden of proof for admitting such a course rests with student, and the decision of the committee will be final. Note that introductory, elementary-methods courses that do not expand the statistical frontier are not generally approved for credit towards the Ph.D. in Statistics.

3.  A minimum of 9 additional units for the Ph.D. minor.
Graduate College requirements stipulate that a minimum of 9 units be applied for the Ph.D. minor.  Minor requirements are fixed by the minor department or program; some Minor programs require upwards or 12 or even 15 units for completion.  (A Ph.D. Minor in Statistics cannot be counted towards a Ph.D. in Statistics.)  The selection of the Ph.D. Minor field is to be made by the student in consultation with her/his advisor and the Program director.  The Minor should reflect the student’s transdisciplinary interests, and wherever possible should be coordinated with the student’s additional Statistics electives.  A non-exhaustive list of potential Ph.D. Minors, and their requirements, is given in Appendix 4.

4.  Dissertation credit:  minimum 18 units of STAT 920
As per Graduate College requirements, a minimum of 18 units in the Ph.D. program of study must include dissertation credits.  These are used to undertake the Ph.D. research.  Registration for any units of STAT 920 is restricted to students who have passed the Ph.D. Comprehensive Examination (see below).  Students who wish to undertake research coursework prior to passing the Comprehensive Examination may consider STAT 599 and/or STAT 900 as possible alternatives; however, a maximum of only 6 units may from STAT 599 and/or STAT 900 may be applied to the Ph.D. curriculum.

5.  Language requirement
The GIDP in Statistics has no formal, non-English, language requirement.

* course being developed

The Doctoral Plan of Study, or DPOS, represents a tailored guideline for the courses that will satisfy the student’s Ph.D. degree requirements; it may be amended if circumstances so require.  The DPOS should be formulated and submitted during the student’s third semester in residence in the GIDP:  by the end of second semester in residence, the student should choose an advisor from among the Regular faculty of the GIDP in Statistics (see Appendix 1).  [Until such time as the student has chosen an advisor, the chair of the GIDP Recruiting & Admissions (R&A) Committee, or another GIDP faculty member designated by the GIDP Chair, serves as the student’s temporary advisor.]  After consultation with the chosen advisor and the GIDP Chair, the student determines an expected list of courses to be taken toward the Ph.D. degree, and submits these on the DPOS to the Graduate College.  The DPOS must be on file before the student can sit for the Ph.D. Oral Comprehensive Examination (see below.) 

Components on the DPOS will include the 33 units of core Ph.D. courses and the minimum 12 + 9 = 21 units of additional electives and Ph.D. minor coursework that comprise the student’s own interdisciplinary specialization.  The Ph.D. Minor must be officially recognized and correctly listed on the DPOS form.  No more than half of the total units listed on the DPOS (including transfer units; see below) can be in courses graded with an S or P grade.  The 18 units of dissertation research (STAT 920) are tracked separately and should not appear on the DPOS.

Up to 12 units on the DPOS may be from courses taken outside a UA graduate degree program, including courses taken in non-degree seeking status and/or as transfer coursework.  (Note:  The GIDP does not recognize coursework at or below the 400-level  for credit towards the Ph.D. in Statistics.)  Transfer coursework must be approved in advance by the Program director and by the Graduate College, via the College’s Transfer Credit Form.  (Consult the Graduate College’s degree certification staff for more details on transfer coursework requirements.)  Approved transfer coursework must be listed as such on the DPOS.

To proceed towards Ph.D. candidacy in the GIDP, a student must pass a written Ph.D. Qualifying Examination by the beginning of her/his fourth semester of study.  The examination may be retaken only once.  Offered during May and January of each year, the Qualifying Examination is used to assess the student’s potential to successfully complete a Ph.D. dissertation in modern interdisciplinary statistics.  It tests the student’s ability to integrate material from the following core Ph.D. courses, and to use this knowledge in solving pertinent, challenging statistical problems commensurate with Ph.D. status at the level of this material:
                        Theory of Probability (MATH 564)
                        Theory of Statistics (MATH 566)
                        Advanced Statistical Regression Analysis (MATH 571A)*
                        Design of Experiments (MATH 571B), or
                                    Research Design & Analysis of Variance (PSYC 507C)

Each specific examination is constructed and graded by a committee of GIDP faculty appointed annually by the GIDP Chair.  Where possible, this will include the instructors of the pertinent core courses.  A minimum of two examiners grade every question independently.

The exam is administered in two parts over two consecutive days; the first day centers on the more theoretical material from MATH 564-566, and the second day centers on the more methodological material in MATH 571A and MATH 571B/PSYC 507C.  (With prior agreement of the GIDP Examination Committee, this order may be reversed on an exam-by-exam basis, but must be the same for all students who sit for that exam.)  Note, however, that concepts from all four course areas will be addressed freely in either or both parts of any examination, at the discretion of the Examination Committee.

Students have 4 hours on each day to complete the exam questions given on that day. If a student is present for only one of the two days, s/he will be viewed as having attempted the entire exam and graded accordingly.

Replicas of previous exams are available from the GIDP’s Graduate Coordinator for students who wish to make copies for study purposes. 

Students must register in advance for each examination; deadlines will be set by the GIDP Examination Committee prior to each offering.  (Students who register for an examination but do not sit for it at the designated time and place will be viewed as having received a failing grade on that examination, unless prior authorization and approval are acquired from the Program director.)

There are three possible outcomes to the exam:
• PhD Pass: Students may continue towards candidacy in the Program.
• MS Pass: Students interested in pursuing a Ph.D. will be required to retake the exam the next time it is offered and improve to a grade of PhD Pass.  Students who wish to complete only the M.S. degree in Statistics (see below) may use this result to stand for their M.S. exit examination.
• Fail: Students must retake the exam at the next opportunity in order to remain in the Program and be eligible to pursue a Ph.D. degree in Statistics.  Failure, or a score of MS Pass, on a second attempt on the exam results in the student’s dismissal from the Ph.D. program.

After receiving written notice of the Qualifying Exam results, each student will meet individually with the Program director to discuss his/her performance and options for the coming year. Students may request a consultation with a designated member of the Examination Committee to review their exam results.

Note that a Qualifying Examination result of MS Pass on a first attempt indicates that the student has made good progress, but also identifies areas of weakness that must be resolved in order to pursue more advanced studies.  Although students are often initially disappointed at not having achieved a PhD Pass on their first attempt, they should recognize that an MS Pass demonstrates development in their studies.  It is not uncommon for students who retake and pass the exam, after the benefit of further study, to later produce excellent Ph.D. dissertations.

* course being developed

Before advancement to Ph.D. candidacy, a student must pass a written and an oral Ph.D. Comprehensive Examination in both the major area of Statistics and the chosen minor(s).  This examination is intended to test the student’s comprehensive knowledge of Statistics and of the minor field(s) of study, both in breadth across the general field of Statistics and in depth within the area of interdisciplinary specialization.  The Comprehensive Examination is considered a single examination, although it consists of separate written and oral components.

A student must pass the written portion of the exam before sitting for the oral portion.  The written portion is determined and graded by a comprehensive examination committee, which by Graduate College regulations must consist of a minimum of four members.  The student’s advisor serves as chair of the committee.  The advisor and two additional members must be tenured, or tenure-track, members of the gradate faculty.  (Advisors who do not hold such status must receive special approval from the Dean of the Graduate College to serve in this capacity.)  The fourth member may be tenured or tenure-track, or a special approved member.  (If the advisor is already specially approved due to non-tenure-eligible status, the fourth member of the committee must be tenured or tenure-track.)  All special members must be pre-approved by the Dean of the Graduate College. Any members beyond the fourth can also be tenured or tenure-track, or special approved members.  It is expected, but not required, that the examining committee will overlap with the student’s graduate dissertation committee (see below), in order to foster continuity in the student’s research program.

The typical format for the written portion of the examination is a series of technical and conceptual questions put forth by the committee concerning the student’s expected dissertation research.  If the research theme is not yet sufficiently developed, these question can address broader areas of interdisciplinary statistics.  The student is given the questions in a single examination packet, and allowed a maximum of 24 hours to answer the questions in writing.  Deviations from this format are permitted at the discretion of the examining committee and with prior agreement of the student.

During the 24-hour examination period, all questions and interactions on the examination material, on allied material, and/or on larger issues related to scholarship in statistics or in the minor field(s) of study must be transmitted to the examining committee chair.  Interaction with any other individuals, without prior approval from the GIDP Chair, is considered a breach of the UA Code of Academic Integrity and will be treated accordingly.

The written portion of the exam will be graded by the examining committee, and results transmitted to the student within 14 calendar days of receipt of the student’s answers.  A student who fails her/his written portion may sit for a second attempt; an entirely new set of questions may be drawn up and graded by the committee.  This second sitting must be scheduled within 2 months of the original sitting.  Failure on a second written portion will lead to a student’s dismissal from the Program.

Upon successful completion of the written portion of the Comprehensive Examination, a student must sit for the oral portion of the exam.  The oral portion is again conducted by the student’s examining committee, and must occur no earlier than 1 calendar week and no later than 4 months after successful completion of the written portion.  An approved DPOS must be on file before the student can sit for the oral portion of the Comprehensive Examination.

The faculty committee conducting the oral portion of the examination has both the opportunity and obligation to require the student to exhibit knowledge of

(i)    the specific questions/material posed during the written portion,
(ii)   general comprehension of the minor field(s) of study as it pertains to the student’s research interests, and
(iii)  sufficient depth of understanding in the area(s) of the student’s statistical specialization.

As with the written portion of the examination, a student who fails her/his oral portion may sit for a second attempt.  This second sitting must be scheduled within 2 months of the original oral sitting, but also no later than 4 months after successful completion of the written portion.  Failure on a second oral portion will lead to a student’s dismissal from the Program.

The written and oral portions of the comprehensive examination must be completed successfully no later than 3 months prior to the Final Oral Defense Examination (see below).

Once a student has an approved DPOS on file, has satisfied all required coursework and residence requirements, and has passed the written and oral portions of the Comprehensive Examination, s/he must file an Application to Advance to Candidacy.  This application must be submitted to the Graduate Degree Certification Office as soon as requirements are met but no later than 6 months before the Final Oral Defense Examination is scheduled (see below).  The Office is located on the 3rd floor of the University’s Administration building.  Deadlines for the submission of paperwork pertaining to doctoral programs are available in the Graduate Degree Certification Office, or online.  Forms for “Application to Advance to Candidacy” are available online and from the Graduate Degree Certification Office.

At the time a student submits her/his Application for Advancement to Candidacy, her/his bursar account will be billed a single composite fee (currently $135) for candidacy, dissertation processing, and archiving.  This is a one-time fee and is not reassessed if the anticipated graduation date changes.  Copyrighting of the dissertation is optional and carries an additional fee.

By the time of advancement to candidacy, the candidate must form a Dissertation Committee.  Graduate College regulations require the committee to consist of at least three faculty, all tenured, tenure-track, or approved by the Dean of the Graduate College as equivalent.  The candidate’s advisor, who must be a Regular member of the GIDP in Statistics, serves as the committee chair.  At least one of the other two committee faculty must be from the GIDP in Statistics; s/he may hold Affiliate or Regular status.  One committee member must represent the student’s minor field(s) of study.  A fourth or fifth member from the GIDP in Statistics or from the Minor area(s) may be added to the committee, at the discretion of the candidate and with approval of the committee chair.  Such members can be tenured or tenure-track, or a special approved member.  As above, special members must be pre-approved by the Dean of the Graduate College.  Individual faculty members may decline to serve on committees for academic reasons.  It is encouraged, but not required, that the dissertation committee include as many members as possible from the candidate’s comprehensive examination committee, in order to foster continuity in the student’s research.

If a dissertation committee has only three members, all must approve the dissertation prior to or at the final oral defense (see below).  For committees consisting of four or five members, there may be at most one dissenting vote.  All dissertation committee members are expected to attend the oral defense.

For the dissertation, candidates must develop a (formal or informal) research proposal of sufficient academic merit and on a topic of sufficient scholarly impact to satisfy their committee.  Work proceeds on the research via the dissertation course STAT 920, under the direction of the candidate’s advisor and using the skills and knowledge of the larger committee where appropriate.  The final, completed dissertation must represent an original, substantive advance in the theory, methodology, and/or practice of statistics, with focus on the candidate’s interdisciplinary interests.  It is expected that the work will result in one or more published research articles in high-quality, peer-reviewed statistics and subject-matter journals.

To be awarded the Ph.D., a candidate must complete the dissertation and submit to a Final Oral Defense Examination.  The examination focuses on the dissertation itself but can include general questions relating to the interdisciplinary study of statistics contained within the scope of the dissertation research.

The exact time and place of this examination must be scheduled with the Graduate Degree Certification Office at least seven working days in advance of the event.  The candidate must be in good academic standing in order to schedule the defense with the Graduate College.

The chair of the dissertation committee presides over the examination.  The examination is closed to the public, although an initial, open portion may be held during which the candidate presents the dissertation results and entertains questions.  (Questions may not be proffered by the dissertation committee during the open period.)  There is no minimum time limit for the Final Oral Examination, but the entire proceedings may not exceed three hours.  Members of the dissertation committee must be present for the entire examination.  In closed session following the examination, the committee votes to pass or fail the candidate based on her/his performance during the oral defense:  if the committee is made up of three members, at least two must vote to pass for a successful determination; if the committee is made up of four or five members, at least three must vote to pass.  The candidate must make her/himself available to the committee chair within 24 hours of completion of the defense, in order that the results of the examination may be delivered.

Upon successful completion of the Final Oral Defense Examination, the candidate submits the dissertation to the Graduate College electronically or via paper copies for forwarding to the Library of The University of Arizona and to University Microfilms, Inc.  A processing and microfilming fee also must be paid to the University Bursar.  The final, completed dissertation must meet all Graduate College formatting and submission requirements; a bound, printed copy is required for submission to the GIDP Graduate Coordinator.  For more information consult the Manual for Theses and Dissertations, available online or from the Graduate Degree Certification Office.  Upon receipt of the finalized dissertation, the Dean of the Graduate College will recommend conferral of the doctoral degree by the Arizona Board of Regents.

In general, continuous enrollment is expect of all students admitted to a Ph.D. program.  A student admitted to the Ph.D. program in Statistics must register each Fall and Spring semester for a minimum of 3 graduate units from original matriculation until the completion of all course requirements, written and oral comprehensive exams, and the 18 units of STAT 920.  After these requirements are met, doctoral students not on financial assistance and/or needing to maintain appropriate visa status must register for a minimum of 1 unit each semester until final copies of the dissertation are submitted to the Graduate Degree Certification Office.  Students receiving funding such as assistantships, fellowships, loans, grants, scholarships or traineeships may be required by their funding source to register for more than 1 unit to meet full-time status requirements, and should check with the GIDP office regarding such requirements to ensure that they remain qualified for funding.  Doctoral students utilizing University facilities or faculty time during summer sessions must enroll for a minimum of 1 unit of graduate credit.

If any Ph.D. degree requirements (including the Comprehensive Examination and the Final Oral Defense) are completed during Summer term(s), the student must be registered for a minimum of 1 unit of graduate credit during that term.  If any degree requirements are completed during an intersession (Winter session or the Pre-Session), the student must have been registered for a minimum of 1 unit during the preceding semester.

Unless excused by an official Leave of Absence (which may not exceed one year throughout the student’s degree program), all graduate students are subject to this Continuous Enrollment Policy and must pay pertinent in-state and/or out-of-state tuition and fees in order to remain in the program.  If the student fails to obtain a Leave of Absence or fails to maintain continuous enrollment, he or she will be required to apply for re-admission, to pay the Graduate College application fee, and pay all overdue tuition and fees, including cumulative late penalties. Tuition or registration waivers cannot be applied retroactively.

Graduate College policy states that all requirements for the degree of Doctor of Philosophy must be completed within 5 years of passing the Comprehensive Examination.  Should a student fail to complete all the Ph.D. requirements within that time period, s/he may be allowed to sit for another Comprehensive Exam, with permission of the Program, and then proceed to complete any remaining Ph.D. requirements.