Graduate Training Programs in Biostatistics

7. Graduate Training Programs in Biostatistics

 

This version of the biostatistics portion of the handbook applies to students who entered the program in the fall of 2019. For students who matriculated prior to this, please refer to the previous version of the handbook that is specific to the student’s matriculation year found at http://www.med.upenn.edu/ggeb/GGEB_Handbook.shtml. Questions about the program for any cohort should be referred to the student’s academic advisor, the Program Chair, or the GGEB Chair.

The PhD program in biostatistics is designed to prepare students to be independent researchers in the development of statistical methodologies and in the appropriate and innovative application of these methodologies to biomedical research problems. In the first five semesters of the program, students complete a series of courses in both theory and applied methodology, engage in individually mentored research experiences, explore statistical collaboration, and complete the qualifications examination. Within this period, students also identify a dissertation research problem and an advisor and present a research proposal as part of the candidacy examination. Students typically defend their dissertations and graduate within five years of matriculation.

The MS program is designed to prepare students to be practitioners of biostatistical methodologies. MS students must formally apply to enter the PhD program. The MS program closely parallels the first two years of the PhD program and requires four full-time semesters of course work, exploration of practical aspects of collaboration, and an independent research project which serves as the MS thesis.

    1. Elements Common to MS and PhD Programs
      1. Academic Advisor

Each incoming student is assigned an academic advisor who serves as the student's  primary mentor, advising in course selection and related academic matters. The program attempts to match students to advisors who have similar backgrounds and interests. A student may change advisors at any time by request to the Program Chair.  A PhD  student’s dissertation advisor, once selected, assumes the role of academic advisor during the later years of study.  Furthermore, at any time a student may refer questions about his   or her program to the chair of the Academic AdvisingCommittee.

At the beginning of the academic year, each student, in collaboration with his/her advisor, prepares a proposed academic program, including courses to be taken, courses to be transferred, and timelines for examinations and dissertation preparation.

      1. Non‐Credit Requirements

The department hosts a weekly biostatistics research seminar that invites speakers from other universities, industry, and government. All students are expected to attend at least six seminars per semester, provided there is no conflict with courses. Other non-credit requirements include Responsible Conduct of Research, CITI, and HIPAA training. These requirements are described in detail in Section 6.3. PhD students must also serve as teaching assistants which is also a non-credit requirement (see Sections 7.3.9 and 7.5, respectively).

      1. Transfer of Credit

Only courses considered at the graduate level may be transferred from previous training.     At least eight course units of the total program required for the MS degree must be  completed while enrolled in a graduate program at UPenn. Because the MS program  requires only 12 total course units, no more than four may be satisfied by transfer credit. A maximum of eight units may be transferred from previous training towards the PhD degree. Courses proposed for transfer credit must be relevant to training in biostatistics and may include courses in theory or methods. Transfer of credit must be approved by the Program Chair and the GGEB Chair.

Combined degree students receive an automatic credit transfer (MD-PhD 10 units, VMD- PhD 14 units). These credits can only count towards advanced elective courses.

      1. Auditing

Auditing a course is not allowed for any students in the PhD program. For MS students, auditing of a course is strongly discouraged. If a MS student wishes to audit a course he/she must consult their academic advisor, the course director, and prepare a written request to the Program Chair explaining reasons for the proposed course audit. Final approval must be obtained from the GGEB Chair.

    1. Master of Science (MS) in Biostatistics
      1. Course Requirements

Candidates for the MS degree must complete 12 units of course credit and prepare a Master’s Thesis. Required courses cover probability, mathematical statistics, and statistical methods including categorical data analysis, linear models, survival analysis, and applied data analysis. MS students will have the option to take the written qualifying exam but it is not required to obtain a master’s in biostatistics.

 

The MS in Biostatistics typically requires four semesters of formal course work. Students must complete nine units of required courses, three units of electives, and the Biostatistics in Practice and project requirements (see Section 7.2.3). The required courses are described below. The courses in bold type are the “core” courses for the MS degree.

 

Theory:

BSTA 620: Probability (1 unit)

BSTA 621: Statistical Inference I (1 unit)

Methods:

BSTA 630: Methods I (1 unit)

BSTA 632: Statistical Methods for Categorical and Survival Data (Methods II) (1 unit)

BSTA 651: Introduction to Linear Models & Generalized Linear Models (1 unit) BSTA 656: Longitudinal Data Analysis (0.5 unit)

BSTS 754: Advanced Survival Analysis (0.5 unit) BSTA 660: Design of Observational Studies (0.5 unit) BSTA 661: Design of Interventional Studies I (0.5 unit)

BSTA 670: Programming and Computation for Biomedical Data Science (1 unit) BSTA 511: Biostatistics in Practice (1 unit)

7.2.1.1 Electives

Students in the MS program choose three additional units from a list of advanced courses  in biostatistics and related topics. At least two of these courses must be quantitative; the third may be in a related scientific field subject to approval by the Program Chair and   GGEB Chair. A partial listing appears under the section on electives for the PhD program (Section 7.3.3). In addition to these electives, BSTA 622 Inference II, which is a required course for the PhD program, may beused as an advanced elective for the MS program.

Courses not described here may be used as advanced electives for the MS program upon receiving approval from the Program Chair and GGEB Chair.

      1. Biostatistics in Practice and the MS Thesis

All MS students must participate in Biostatistics in Practice and complete a Biostatistics in Practice project, which serves as the MS thesis. The project may be completed in any semester. See the description of the Biostatistics in Practice requirements in Section 7.3.8.

      1. Typical Course Sequence for Full‐Time Students in the MS Program

 

 

Semester

 

Required Credit Courses (units)

 

 

 

 

 

 

Year 1

Fall

FT Courses:

BSTA 620: Probability (1)

BSTA 630: Statistical Methods and Data Analysis I (1) BSTA 660: Design of Observational Studies (0.5) BSTA 661: Design of Interventional Studies (0.5)

Spring

FT Courses:

BSTA 621 Statistical Inference I (1)

BSTA 632 Statistical Methods for Categorical and Survival Data (Methods II) (1)

BSTA 651: Linear Models and Generalized Linear Models (1)

Summer

Written Qualifications Examination (Optional for MS Students)

 

 

 

 

 

Year 2

Fall

FT Courses:

BSTA 754: Advanced Survival Analysis (0.5)

BSTA 656: Longitudinal Data Analysis (0.5)

Advanced Electives (1) 

BSTA 511 (1) Biostatistics in Practice

Spring

FT Courses:

BSTA 670: Programming and Computation for Biomedical Data Science (1)

Advanced Electives (2) MS Thesis Presentation

 

    1. Doctor of Philosophy (PhD) in Biostatistics
      1. Funding Policies

Full-time PhD students are eligible for funding in the form of traineeships, fellowships, and research and teaching assistantships. The work associated with these sources of support is an essential part of the graduate training program.

At the beginning of the academic year, each funded student receives a letter describing sources of support and associated obligations. All doctoral students admitted to the program receive support from BGS for up to 21 months. During this time students take courses and explore research opportunities with faculty in the department. Once students decide on a dissertation advisor, the advisor, the student and the Program Chair will work to develop a sustained funding plan that will cover the student’s time during the remainder of their graduate program. A number of funding opportunities are available including federally funded research and training grants, partial TAships and funding through partnerships with industry. The typical length of time in the program is five years. Students in good academic standing have occasionally received funding for up to six years in the program.

Combined degree (MD-PhD, VMD-PhD, and DMD-PhD) students receive financial support from the program until June 30th of the first year in the PhD phase of their program, and are supported by their dissertation advisors starting from July 1st of the same year.

      1. Course Requirements

The PhD in Biostatistics typically requires four and a half semesters of coursework and additional semesters devoted to dissertation research. This is usually accomplished in four to five years of full-time study. The standard course sequence for PhD students consists of 3 units in theory, 7 units in statistical methods, and 5 units of advanced electives. At least 3 of these advanced electives must be in theory and methods and 2 may be taken outside of these areas. In addition, a minimum of three semesters of lab rotations (BSTA 699) is required. In general, students are expected to have completed all required courses by the end of their 3rd year (or equivalent for those who enter with a Master’s degree). In rare cases substitutions may be made. Such alternatives must be pre-approved by the chair of the Curriculum Committee, the Program Chair, GGEB Chair and the director of the course being waived, who is in the best position to evaluate whether the necessary skills are met by the substitution.

 

Below are the required core courses; the courses in bold type are PhD “core” courses that are covered on the written qualifying examination.

Theory:

BSTA 620 Probability (1 unit)

BSTA 621 Statistical Inference I (1 unit) BSTA 622 Statistical Inference II (1 unit)

Methods:

BSTA 630 Statistical Methods and Data Analysis I (1 unit)

BSTA 632 Statistical Methods for Categorical and Survival Data (Methods II) (1 unit) BSTA 651 Introduction to Linear Models & Generalized Linear Models (1 unit)  BSTA 656 Longitudinal Data Analysis (0.5 unit)

BSTA 660 Design of Observational Studies (0.5 unit) BSTA 661 Design of Interventional Studies I (0.5 unit)

BSTA 670 Programming and Computation for Biomedical Data Science (1 unit) BSTA 754 Advanced Survival Analysis (0.5 unit)

BSTA 511: Biostatistics in Practice (1 unit)

      1. Electives and Independent Study

Students are required to take 5 additional advanced electives; a partial listing of such courses is given below. In addition to this list, other courses offered by departments outside of the Department of Biostatistics, Epidemiology & Informatics may be appropriate advanced electives and may be used as an advanced elective for the PhD program upon receiving approval from the student’s academic advisor and the Program Chair. At most one of the five required advanced electives may be a reading course, and only on a topic not offered as a formal course within a year.

STAT 530 Probability (1 unit)

STAT 531 Stochastic Processes (1 unit)

STAT 921 Observational Studies (1 unit)

STAT 925 Multivariate Analysis: Theory (1 unit) OPIM 930 Stochastic Models II (1 unit)

BSTA 751 Statistical Methods for Neuroimaging (1 unit) BSTA 771 Applied Bayesian Analysis (1 unit)

BSTA 774 Statistical Methods for Evaluating Diagnostic Tests (0.5/1 unit) BSTA 775/STAT 920 Sample Survey Methods (1 unit)

BSTA 777 Systematic Reviews and Meta-analyses (1 unit)

BSTA 782 Statistical Methods for Incomplete Data (1 unit) BSTA 783 Multivariate and Functional Data Analysis (1 unit)

BSTA 785 Statistical Methods for Genomic Data Analysis (1 unit) BSTA 786 Advanced Topics in Clinical Trials (1 unit)

BSTA 787 Methods for Statistical Genetics in Complex Human Disease (1 unit) BSTA 788 Functional Data Analysis (1)

BSTA 789 Big Data (1)

BSTA 790 Causal Inference in Biomedical Research (1 unit)

BSTA 820/STAT 552 Statistical Inference III (1 unit)

BSTA 852/STAT 910 Forecasting and Time Series (1 unit)

BSTA 854/STAT 927 Bayesian Statistical Theory and Methods (1 unit)

      1. Applied Research Requirement, Equivalent of MS Thesis

All PhD students must participate in Biostatistics in Practice and complete a Biostatistics in Practice project, a requirement that students typically satisfy during the first or second year. See Section 7.3.8 for further details.

      1. Teaching Practicum

All students in the PhD program must provide teaching support for courses offered by the Department of Biostatistics and Epidemiology or related programs. This is discussed in detail in Section 7.5.

 

      1. Examinations

The PhD program requires the successful passing of three examinations: the written Qualifications Examination, the oral Candidacy Examination, and the Dissertation Examination. Later sections outline the procedures for each of these.

      1. Lab Rotations

Goals and Objectives

The overall goal of the rotations is to expose students to biomedical research, and in particular research related to statistical methodology early in their training. In addition, students will rotate through a number of different labs, in order to get a broad perspective on research and faculty. Only GGEB faculty are eligible to lead a lab rotation. This will also assist students in identifying their research interests and dissertation topic earlier in their educational process. In addition, both the students and faculty can assess whether they are a good match for possible dissertation advisor/advisee relationships. By the end of 21 months of training (summer of year 2) students who were initially funded by BGS will identify their dissertation advisor, have a foundation for the first topic in their dissertation work, and move off of BGS funding and onto funding that is related to their dissertation work. Students will normally identify their PhD mentor through working with them on a lab rotation. Students who are funded by a training grant during their first 21 months in the program will remain on the training grant throughout their program. Students who are currently funded or who have interests in receiving funding during their dissertation research from a training grant should discuss how to structure their lab rotations with the training grant director. Lab rotations that offer research experience in areas relevant to our training grants will be available each year.

Combined degree (MD-PhD, VMD-PhD, and DMD-PhD) receive financial support from the program until June 30th of the first year in the PhD phase of their program, and are supported by their dissertation advisors starting from July 1st of the same year.

Lab Rotation Content

Laboratory rotations should focus on statistical research as opposed to simple data analysis or service/project work. Ideas include implementation of a method from the literature, literature review of methods, running and summarizing simulations, or the analysis of a complicated dataset. It is recognized that, depending upon the background of the students matriculating into the program, the first lab rotation may need to be a lighter introduction to the methodological area such as focusing on literature review, study design, logistics, and data management and/or data analysis.

 

Duration and Number of Individual Rotations and Dissertation Advisor Choice Students are expected to participate in 3-5 rotations in total with a minimum of 2 different mentors. First year students will have 3 rotations: fall, spring, and summer. Students can expect to spend 20+ hours per week during the fall/spring semesters and full time 40+ hours

during the summer. In the second year it is expected that students will settle into a relationship with a potential PhD dissertation advisor with the goal of focusing their research in an area of research related to that of the advisor. Nevertheless, if needed, a fourth or even fifth lab rotation may occur in the second year to help the student decide on a research topic and mentor. In addition to determining an area of research interest, the dissertation advisor must show willingness to, at least partially, support the student.

Laboratory Rotation Assignments

Prior to the arrival of the new students, interested faculty mentors will submit a short summary describing their proposed laboratory rotation (similar to a course description). During new student orientation and the first week of classes, faculty who are interested in serving as a mentor for the upcoming semester will make a brief presentation of their projects. Faculty mentors can also distribute a handout with a more detailed project description to interested students. Each student should schedule meetings with 2 or 3 potential faculty mentors. This should occur within one week after the faculty presentations so that lab rotation assignments can be made and lab rotations can begin in a timely fashion for the Fall semester. Students will rank their choice of projects and the rankings will be reviewed by members of the Academic Advising Committee. The Academic Advising Committee will take this information and facilitate matching students to faculty. It is important to note that each student will be working in three labs during their first year, so students and faculty will have ample opportunity to work with each other over the course of the program and initial matches do not preclude options for other collaborations. The same process will be used to match first year students to faculty mentors for lab rotations in Spring semester.

For subsequent lab rotations (summer of first year, and Fall and Spring of second year) students may identify faculty mentors on their own without participating in this formal matching process. In the case of lab rotations that are arranged outside of the formal match process, students should email the Chair of the Academic Advising Committee, copying their faculty mentor, with their choice of lab rotation mentor. Students and faculty are asked not to arrange lab rotations outside of the formal match process during Fall and Spring of first year.

It is the responsibility of faculty mentors to make sure any required IRB approvals are obtained, and that students have the requisite CITI training to work with any data related to their laboratory rotation project.

Evaluation

At the initiation of each new lab rotation, the advisor and mentee will write a short summary of the goals and expectations for the particular laboratory rotation. This document will be submitted to the Academic Advising Committee for review. Laboratory rotations are taken for credit (1 unit, BSTA 699) and students receive both a written evaluation and a letter grade.

 

Summer rotations also receive a letter grade; however, these will not appear on the formal transcript. In addition, each student will present their progress on their laboratory rotation work at least once per year during a brown bag format “chalk talk”. Attendance will be required for all students participating in various lab rotations and encouraged for more senior students as well as faculty.

Laboratory Rotation Lab Notebook

PhD students are required to keep a formal record/notebook of work being conducted during mentored laboratory rotations as part of the first 21 months in the PhD programs. The lab notebook should be a shared document between the mentor and mentee. This document should be stored in a location that both mentor and mentee can access such as a project directory or a shared PENN Box folder. Both mentor and mentee should continue to have access to the shared folder once the lab rotation has been completed.

The lab notebook should contain dated entries including a list of the goal(s) for each week, and a summary of progress towards those goals. The notebook progress should be reviewed at weekly meetings between the mentee and mentor with new goals set for the following week.

A copy of the lab notebook along with all work products should be retained by the mentor at the end of each rotation.

Biostatistics in Practice

The Biostatistics in Practice (BIP) project is required for both the MS and PhD degrees. It offers the student an opportunity to acquire and demonstrate proficiency in statistical collaboration, data analysis and scientific writing. Students typically start their BIP project in the summer prior to the start of their second year and must submit the final written report by April 15th the next year. The project is defined by several elements:  A scientific question or hypothesis arising in medical research; the statistical methodology needed to address the question; the development of a study design and/or analysis of a relevant data set; and    a summary of the results of these analyses. In most cases, a collaborating medical scientist provides the research question and the data. The student, under the supervision of a biostatistics faculty member, identifies the appropriate statistical methods and conducts the analysis. The analysis should be sufficiently extensive and detailed to support a manuscript publishable in the medical literature.

The project consists of two parts. The first is a written report including: a description of the research question; background and significance; a description of the statistical methods applied; the results of the analysis; and summary of the major findings and conclusions. The written report should be at the level of a completed manuscript that includes Introduction, Methods, Results, Discussion, Literature cited etc. Each BIP project will be supervised by one Biostatistics program faculty member, a secondary subject area advisor, and will also be evaluated by another Biostatistics program faculty member (faculty reader) who will be chosen by the primary Biostatistics faculty mentor. The student must provide the faculty readers their report by March 14th to get feedback. Students may use this Biostatistics in Practice project write-up as their MS thesis. The second part of the project is a short (20 minute) oral presentation of the project to the biostatistics faculty and students.

 
 
  Text Box: All data analyses done as part of the Biostatistics in Practice project must have IRB approval. In most cases this requirement is satisfied if the research objectives are part of an existing protocol of the participating medical research collaborator, as long as the student is added to the protocol according to the standard modification procedures required by the IRB.
 

 

      1. Typical Course Sequence for Full‐Time Students in the PhD Program

 

 

 

Semester

 

Funds

 

Plans

 

 

 

 

 

 

 

Year 1

Fall

BGS

FT Courses:

BSTA 620: Probability (1)

BSTA 630: Statistical Methods and Data Analysis I (1) BSTA 660: Design of Observational Studies (0.5)

BSTA 661: Design of Interventional Studies (0.5) BSTA 699: Lab Rotation (1)

Spring

BGS

FT Courses:

BSTA 621: Statistical Inference I (1)

BSTA 632: Statistical Methods for Categorical and Survival Data (Methods II) (1)

BSTA 651: Linear Models and Generalized Linear Models (1) BSTA 699: Lab Rotation (1)

Summer

BGS

Written Qualifications Examination

BSTA 699: Lab Rotation

 

 

 

 

 

 

Year 2

Fall

BGS

FT Courses:

BSTA 622: Statistical Inference II (1)

BSTA 754: Advanced Survival Analysis (0.5)

BSTA 656: Longitudinal Data Analysis (0.5) BSTA 511: Biostatistics in Practice Project (1)

BSTA 699: Lab Rotation (1) OR BSTA 899: Pre-Dissertation Research (0.33-1) (if a dissertation advisor has been chosen)

Spring

BGS

FT Courses:

BSTA 670: Programming and Computation for Biomedical Data Science (1)

Advanced Electives (2)

BSTA 699: Lab Rotation OR BSTA 899: Pre-Dissertation Research (0.33-1) (if a dissertation advisor has been chosen)

Summer

BSTA

Oral Candidacy Examination

PhD Thesis*

 

 

 

 

 

Year 3

Fall

BSTA

Advanced Electives (3)

BSTA 899: Pre-Dissertation Research (0.33-3)

Oral Candidacy Examination (if not completed in summer) PhD Dissertation

Spring

BSTA

BSTA 995: Dissertation Research PhD Dissertation

Summer

BSTA

PhD Dissertation

Years 4-5

Fall

BSTA

BSTA 995: Dissertation Research PhD Dissertation

 

    1. Evaluation and Examinations

Evaluation for the MS degree is based on performance in the required and elective courses, and completion of the MS thesis/Biostatistics in Practice project. Evaluation for the PhD degree isbased on these criteria plus ongoing interactions with the faculty

 

advisor(s), laboratory rotations, successful performance on the written Qualifications Examination, Candidacy and Dissertation Examinations, and the dissertation itself.

      1. The Written Qualifications Examination (MS and PhD)

The following guidelines refer to the written Qualifications Examination (QE) required for all PhD students and optional for MS students in Biostatistics. The Qualifying Examination Committee develops and administers the examination each year and presents the results  to the full faculty.

Passing of the written Qualifications Examination is required for continuation in the PhD program. This examination also satisfies the UPenn examination requirements as outlined below.

      1. MS Program Examination Requirements

Students in the MS program in Biostatistics, are not required to take and pass the written qualifying examination. In order to obtain the degree, students will need to write and deposit an MS thesis and conduct an oral presentation of their work on one of the designated Biostatistics in Practice presentation days. MS candidates will have the option to take the exam if they are considering PhD work but it is not required. If the MS student passes the QE at the PhD level and is subsequently admitted to our PhD program, he/she will not need to take the exam again

      1. PhD Program Examination Requirements

The written Qualifications Examination serves as the university-required “Qualifications Evaluation” as defined in the Graduate Studies catalog. PhD students must pass the written Qualifications Examination as the first step toward full PhD candidacy. The examination is offered each summer, in June, to allow for grading and faculty review such that students will have results before the July 4th holiday each year. All full-time PhD students are expected to sit for the examination after their first year of study, with the option to retake the exam the following year if the exam is not passed the first time.

This one-day written exam will consist of problems stemming from the five first-year core courses and aims to test theoretical and methodological foundations of the course materials. The exam is 6 hours and is closed book/notes. Because the exam is offered only once a year, PhD students who do not pass the exam and elect to take it a second time have the opportunity to take two additional semesters of course work between examinations.

      1. Examination Results and Notification

All matters pertaining to grading and review of the written qualifications examination by the faculty are confidential, but some general procedures are outlined here.

Once grading is complete (typically two weeks after the exam), the faculty of the Biostatistics program meet to review the examination. Course grades, performance in lab rotations, and related materials may also be considered in the pass/fail decision. Because the difficulty of the examination may vary from year to year, there are no absolute cut-offs for passing.

Each student who takes the exam receives written notification of his or her outcome as soon as possible after the faculty grading meeting. No other information is made available before that time. The letter notifies the student of the outcome of the exam, recommendations for continued study, and the process for discussion of the exam with the Examinations Committee if needed.

 
      1. MS Students Applying to the PhD Program

Students enrolled in the MS program who wish to apply to the PhD program must submit a formal application. As the written qualifications examination is typically administered in the summer after the first year of study, it is likely that the results will be known to the Admissions Committee. The Admissions Committee determines how to use the results of  the written examination in the admissions process.

MS students who pass the written qualifying examination and later enroll in the PhD program are not required to take the written qualifying examination again.

      1. Review of the Examination

A student who does not pass the examination may request to review his/her exam paper with the  Qualifying Examination Committee  (QEC) chair and together with his/her academic advisor.  The purpose of such a review is to help the student evaluate the types  of mistakes made and identify areas where further study is needed.. The QEC retains the graded exam bluebooks and does not return them to the students.

A student may appeal the outcome of the qualifying examination to the QEC Chair, who together with the QEC membership evaluates the appeal and judges whether it has merit.    A grade can only be changed in cases of a specific error in scoring. The QEC refers any grade change that could affect the student’s exam outcome to the full Biostatistics program faculty, who decide whether to revise the exam outcome.

Because a minor revision of the score is unlikely to change the exam outcome, a student should only initiate a formal appeal if there were egregious errors in the problem or in its grading, the correction of which would likely lead to a substantial increase in the score.

The QEC Chair can advise the student on the prospects for success of a proposed appeal.  A student who seeks a formal review must request it, in writing, from the QEC Chair within one weekof the issuance of the letter notifying the student of the exam result.

      1. Sickness Policy

A student who has declared an intention to sit for the exam, but at the time of the exam is too ill to attend, may request a deferment by contacting both his/her academic advisor and the Chair of the QEC. A deferment can be granted only if the student provides an official note from a medical doctorexcusing him or her from the exam as a result of the illness.

The request must be made before taking the exam. If the request is approved, the student

may then be allowed to take the exam on or before the first business day when the note indicates that the student may return to work.

      1. The PhD Candidacy Examination

In order to advance to candidacy for the PhD degree in Biostatistics, a student must successfully pass a candidacy examination. This examination satisfies the requirements of the University’s required “Candidacy Examination” defined in the Academic Rules on the Provost website.

The purpose of the candidacy examination is to evaluate whether the student is qualified to proceed to dissertation research. Although the examination is structured around the dissertation proposal, it may also cover any material in the student’s course work. This includes the core courses and other requirements. The examination should emphasize the student’s qualification to continue as well as the content  of the dissertation proposal. Committee members are encouraged to communicate to both  the student and the dissertation advisor suggestions regarding the proposed research, as  well as the feasibility of completing the project in a reasonable time.

Candidacy Examination Committee Membership

The committee will consist of a minimum of three members, not counting the advisor(s), of whom two must be faculty members of the GGEB, and one must be an external (non GGEB) member. A GGEB faculty member will be appointed as the Chair of the committee by the student’s advisor (the advisor cannot serve as the Chair). The role of the Chair is to run committee meetings and to oversee the candidacy examination and final defense. Committee members will be collectively responsible for administering and evaluating the oral Candidacy Examination, reading the dissertation, and evaluating the final defense. Additional content experts from within or outside the GGEB may be added to the committee as needed. The initial constituency and any changes in the membership must be approved by the Program Chair and the Graduate Group Chair. This Committee will be in place at all times during the dissertation phase. If for some reason, a student changes to a different area of research, a new Dissertation Committee must be appointed immediately and must meet within three months to discuss new plans for the dissertation research.

Candidacy Examination Scheduling

To maintain good academic standing, students must advance to candidacy by 18 months after passing their written qualifications examination.   In typical cases this will occur no   later than the beginning of the student’s fifth (spring) semester, although students are encouraged to complete their candidacy examination within one year of passing the written qualifications examination.   It is the responsibility of the dissertation advisor to work with    the student to schedule the candidacy exam once the dissertation research has been   clearly outlined.   The Program Chair will schedule the candidacy examination for any  student who has not met this deadline.

Content of the Dissertation Proposal

The student should provide a written dissertation proposal to the committee at least two weeks prior to the scheduled examination date. The members review the proposal and prepare questions for the exam.

The dissertation proposal should include a review of the literature relevant to the topic to be studied. The proposal may, but need not, include preliminary research results. The paper should be primarily a true proposal and should typically not exceed twenty pages.

The examination lasts up to two hours and typically includes three parts: First, the student gives a formal presentation of the proposal, generally not to exceed 45 minutes (which may be extended if there are questions during the talk). Next, committee members question the student on the proposal or on topics in biostatistics. Once all committee members are satisfied that the questioning is complete, the student is asked to leave the room. The committee then discusses the examination, votes the outcome (see below), and makes recommendations (if any) for future research and study. The student is then readmitted and informed of the outcome.  The committee chair ensures that necessary  forms are signed and returned to the graduate program office.

Candidacy Examination Outcomes

The candidacy exam has four potential outcomes:

  1. Pass. The student has sufficiently mastered the material and advances to PhD candidacy.
  2. Conditional pass. Additional evidence of mastery of the material is required, the nature of this evidence to be determined by the committee. An example would be a revised dissertation proposal that better represents the nature of the problem to be studied. Once all committee members agree that the additional requirements have been met, the outcome of the examination is changed to a “Pass”.   The student is  not required to retake the candidacyexam.
  3. Fail with possibility of retaking the examination. The student has not mastered the required material and must retake the examination within 6 months in order to progress to PhDcandidacy.
  4. Fail without possibility of retaking the examination. The student has not mastered the required material and the committee determines that the student should not continue in theprogram.

Retaking the Candidacy Examination

A student who fails the candidacy exam after taking it a second time, or who does not make a second attempt within six months of the first (unsuccessful) attempt, is automatically withdrawn from the program.

A student who has passed the candidacy exam but wishes to change dissertation topics must prepare a new proposal and submit it to his or her committee. A discussion of the new proposal should be scheduled within three months of its submission. If there is no substantial change in the statistical content area, the candidacy exam need not be retaken, and the committee need not be reconstituted. If there is a substantial change in the statistical content area, the full committee determines whether an additional candidacy exam is required. The student, the advisor for the new proposal, and the other members of the committee also decide on any changes in the composition of the committee.

   Review of “Lab Notebook”

BGS mandates that the student’s dissertation committee review  the  student’s  “lab notebook” at each of its meetings.  The Biostatistics PhD program interprets this to mean  that the student should make available for faculty review, upon request, primary documentation of any substantial element of the dissertation. Such a review takes place at the candidacy exam and any subsequent meetings of the committee, and following the  closed session of the dissertation defense. Examples of  materials  subject  to  review include the statement and formal proof of a key theorem; the code and results of a   simulation study; or the data, code and results of a data analysis in keeping with   reproducible research best practices. Prior to the meeting, the student’s advisor, in consultation with the dissertation committee, designates a short list of such items that the student makes available in electronic or hard-copy format. The committee chair sets aside time at the meeting for the review of this material. In keeping with the BGS policy, there is   no expectation that the committee should scrutinize all such documents “in their entirety”; rather, the review should be sufficient to satisfy the committee that the student’s research records are “complete and wellmanaged”.

   Frequency of Dissertation Committee Meetings

Once a student has advanced to candidacy, his/her dissertation committee must meet at least once every six months to review goals and progress. A review of the “lab notebook” (see above) must accompany each such meeting. The advisor, working with the chair of  the committee, schedules the meetings.  The student is responsible for providing any  review materials needed by the committee in a timely fashion, typically at least two weeks before a meeting.

      1. PhD Dissertation Examination

This section provides guidelines for the content of the dissertation and the format of the defense. The Biostatistics Program follows the regulations described by BGS. The final steps toward the PhD degree are the preparation of an acceptable dissertation and the dissertation defense. Refer to the Academic Rules.

Permission to Write and Defend the Dissertation

The dissertation committee must formally grant permission to write the dissertation. Students must submit the dissertation to the graduate group within six months of receiving permission to write.

If the student will not meet the six-month deadline, s/he must meet with the committee again before the end of the six months. The committee will review the student’s progress and set a new deadline for the submission of the dissertation. Under normal circumstances, no more than one additional month will be granted. If the student does not submit the dissertation or meet with the committee during the six-month period, the graduate group will put the student on academic probation. The graduate group’s academic review committee will determine whether the student should be given permission to defend the dissertation, and under what circumstances, or whether the student should be withdrawn from the program.

After the student submits the dissertation, the committee has up to one month to review it. If the committee determines that the student must revise the dissertation prior to defending it, the student will have one month to make the revisions.

Once the committee approves the dissertation, the student will have one month in which to finalize the defense arrangements. It is expected that the student will make tentative arrangements for the defense before this point.

Students Leaving the Department Prior to Defending the Dissertation

Refer to http://www.med.upenn.edu/bgs/dissertation_policies.shtml

Dissertation Level Students Changing Mentors

Refer to http://www.med.upenn.edu/bgs/dissertation_policies.shtml

Scheduling

The dissertation defense should be scheduled when the candidate and the dissertation advisor agree that the research is near completion and the draft dissertation is in a format suitable for distribution to the committee.  As soon as a date and time are fixed, the  graduate program coordinator reserves a room (for at least two hours) and prepares the necessary public announcements. In order to accommodate space and scheduling constraints, the date of the exam should be set at least one month before it will take place. All dissertation committee members must attend the exam in person.

Content and Format of the Dissertation

A typical dissertation consists of five chapters: The first is an introduction and brief literature review; in many cases, this is similar to the dissertation proposal. The next three cover the three main topics of the dissertation; these may be written in a format suitable for submission as individual articles to peer-reviewed journals. The final chapter summarizes the dissertation findings and indicates possible future research directions. There are no upper or lower limits on the length of the document. It is the expectation that at least two of the three middle chapters will be methodological and have been submitted to peer-reviewed journals at least two weeks prior to the dissertation defense. Exceptions to this submitted paper requirement would need prior approval from the dissertation committee and the Chair of the Graduate Program.

Content and Format of the Defense

At least four weeks prior to the exam, the student should provide each committee member with a copy of the full dissertation. The committee members review the dissertation and prepare exam questions based on it. The defense consists of two parts:

  1. Open session. The advisor introduces the student and describes the process to all attendees. The candidate then presents his/her research in the style of a departmental colloquium. Typically, the candidate presents one chapter in depth,   with a very brief overview of the others. This presentation should not exceed 45 minutes. At the close of the formal presentation the candidate takes questions from the audience. To leave sufficient time for the closed portion of the exam, the chair  has the right to terminate the open session if it goes onbeyond one hour.
  1. Closed session. In this part of the exam, attended only by the student and the committee, committee members ask specific questions related to the dissertation. Because the committee members have read the entire dissertation, this is their opportunity to ask questions about any part of it, including chapters not presented in detail in the open session. Once all committee members are satisfied that the questioning is complete, the student is asked to leave the room. The committee then discusses the exam and votes an outcome (see below). The student is then asked to return to the examination room and informed of the outcome. The committee chair sees that necessary forms are signed and returned to the graduate program office.

Dissertation Examination Outcomes

The dissertation exam has three potential outcomes:

  1. Pass. The student has completed the dissertation requirements for a PhD in biostatistics. The student then works with the graduate program coordinator to ensure that all other requirements are met prior to deadlines for the proposed graduation date.
  2. Conditional pass. The defense was satisfactory but additional requirements must be met. Commonly, the student is asked to address specific questions raised at the defense, or to incorporate edits proposed by committee members. The dissertation advisor typically oversees these changes, but other committee members may also review changes at their discretion. Once the additional requirements are met, the student is considered to have completed the dissertation. The student is not required to defend the dissertation again.
  3. Fail. The student must defend the dissertation again.

 

    1. Teaching Assistants
      1. Courses That Receive Teaching Assistants

The program assigns TAs to courses in the biostatistics graduate program, the Masters of Science in Clinical Epidemiology, and programs outside of the department such as the Biomedical Graduate Studies program, or the Masters of Science in Health Policy. The Program Chair assigns TAs to courses based on course needs and student qualifications.

      1. Students Who Serve as Teaching Assistants

All doctoral students are required to serve as a teaching assistant for a minimum of two but typically three semesters as assigned by the graduate program chair.  Each TA will spend  the equivalent of six hours per week for each semester course. Some students may be assigned to TA a full semester course that requires additional hours (12-14). Beyond the   first course, students may be eligible to receive supplementary compensation for additional teaching.

Benefits and duties of Serving as a Teaching Assistant

In addition to being a degree requirement for all doctoral students, the teaching experience  is an opportunity to work closely with a faculty member in the department, review and  deepen understanding of the material being taught, and acquire and sharpen  teaching  skills.

 

TA duties typically include some or all of the following:

    • Attend regular meetings with the course instructor
    • Attend lectures
    • Hold office hours
    • Assist in or teach lab sessions
    • Assist in the preparation of written course materials, exams and solution sets
    • Grade assignments or exams and record the grades
    • Coordinate access to computing facilities, online data sets, and web applications

Timeliness in the completion of these duties is essential. The course instructor and TA should communicate regularly to discuss duties, to share feedback from the students, and to ensure that the TA’s time is being used efficiently. A student who feels s/he is spending on average more than the designated number of hours on teaching activities should speak to the instructor or, if concerns remain, with the Program Chair. Both instructors and TAs should recognize that time pressures can vary greatly over the course of a semester. Open communication is the key to a successful teaching experience.

In cases where TA duties include assisting in the grading of exams, course instructors should provide the TA with clear guidance on how to assign points. Instructors should also recognize that some students may feel awkward evaluating their peers.

TAs should share their e-mail addresses and mailbox locations with their students. TAs are not on call for their students; nevertheless, students should expect reasonable access to TAs, particularly in the days leading up to exams and project due dates.

TAs are encouraged to speak with the instructor, their academic advisor and the Program Chair about their teaching experiences, particularly if difficulties arise. A potential conflict can often be avoided if its warning signs are recognized early and the situation is handled thoughtfully.

    1. Other Policies
      1. Student Travel

BGS allows doctoral students to apply for partial reimbursement (currently, up to

$1,000/year) for travel to professional meetings if they are making a presentation.   Applicants must justify the expenses prior to attending the meetings. Dissertation advisors sometimes are able to augment these travel funds.   In addition, some training grants  provide funds for student travel. Request for Travel Funds here.

      1. International Students Travelling Abroad

Students are entitled to two weeks of vacation per year. Unfortunately, it is sometimes difficult for international students leaving the US to re-enter in a timely fashion, often due to visa issues.  While the program understands that these issues are largely out of the control  of our international students, it also puts a strain on financial sponsors when students are unavailable to carry out theobligations of their research assistantships in person.

International students who leave the country for vacation must make a plan with their  sponsor inadvance to address the possibility of delayed re-entry. The written plan must be approved by both the research sponsor and the Program Chair.  Sponsors are not required  to allow students to work remotely and, in particular, are not expected to fund students to work remotely for indefinite periods of time.  International students should expect that  funding will be suspended if suitable arrangements are not made and/or if students are absent for more than two weeks. Details on travel requirements for international students  can be found at https://global.upenn.edu/isss/travel

      1. Registration in Biostatistics Courses

Any student who does not have dissertation status and who wishes to take a course in Biostatistics must formally register for the course. The University reserves the term  “auditor” for a student who registers for a course without the intention of receiving a grade and academic credit. BGS PhD students are not allowed to audit a course. MS students may request a waiver of this policy from the Chair of the GGEB.

A student with dissertation status (i.e., who is paying dissertation fees and therefore does not pay for course units) may sit in on a course without registering for it. Students who intend to participate in a course in this way must agree to participate in the course in a manner defined by the instructor, and must obtain the instructor’s prior approval.

    1. Committees

Six committees provide governance and administrative leadership to the Biostatistics graduate programs. Membership is subject to change annually.

The Admissions Committee is responsible for the application and admission process. Responsibilities include developing admission policies; identifying qualified students; reviewing applications; selecting students to interview;  ranking  students  for  admission; and reviewing applications from students in other BGS programs who seek to transfer into Biostatistics. Admission decisions are subject to approval by the GGEB Chair, the GGEB Executive Committee and the BGS Admission Committee. The chair of the Biostatistics Admissions Committee, together with the chair of the Epidemiology  Admissions Committee, represents the GGEB in the BGS AdmissionCommittee.

The Curriculum Committee is responsible for all rules and policies related to courses, MS theses and doctoral dissertations. Responsibilities include developing policies related to course content; reviewing requirements for MS theses and PhD dissertations; approving proposals for the creation of new courses; and reviewing student course evaluations.

The Academic Advising Committee was created to facilitate the assignment of academic advisors, lab rotation mentors and to insure that each student is meeting program requirements in a timely fashion. This committee works closely with the curriculum committee to insure that all faculty mentors are up-to-date on the curriculum and all requirements.

The Qualifying Examination Committee conducts the Program’s written qualifications exam. Responsibilities include developing guidelines, policies and procedures for the exam; soliciting questions for the exam; reviewing and selecting questions; creating the exam itself; managing its grading; presenting results to the faculty; and evaluating the merits of appeals of exam results. Decisions on the outcome of the exam are made by the Biostatistics program faculty assembled as a committee of the whole.

The Student Recruitment Committee conducts outreach to establish and maintain a pipeline of talented undergraduate students to apply to the GGEB MS and PhD programs in biostatistics. Activities include informational presentations to undergraduate departments of mathematics and statistics in Greater Philadelphia and additional targeted areas, attendance and recruitment at national undergraduate research conferences, creation and national circulation of program announcements, communication with prospective applicants, and hosting an informational open-house each fall. The Student Recruitment Committee also aims to increase awareness and enrollment in Penn's summer undergraduate research opportunities in biostatistics.

The Student Awards Committee explores opportunities and supports applications for student awards. The Committee chooses the award winner for the Saul Winegrad Award for Best Dissertation. In addition, along with students in the program, the Committee nominates an outstanding faculty member for the Jane M. Glick Graduate Student Teaching Award. Additionally, there are three awards exclusive to the GGEB: the Tom Tenhave Student Leadership Award, the Graduate Student Teaching Award, and the Faculty Teaching Award. These are awarded at the end of every academic year.