Operations and Management Science (OMS)

PhD Program Handbook

January 2011

1

1. Introduction The purpose of this document is to provide background on the PhD program in Operations and Management Science (OMS). This document is intended to provide an overview for prospective applicants and to aid currently enrolled doctoral students in their academic program. The PhD program in OMS is designed to prepare candidates for a successful career as a faculty in a doctorate granting university. Each element of the OMS PhD program is designed to ensure that critical developmental milestones are achieved in a timely manner. While some exceptions exist, the PhD program in OMS typically takes four or five years of study. 1.1 What is Operations and Management Science? Operations are the processes that transform inputs into outputs that create value for society. The design and management of these processes is at the core of Operations and Management Science. Scholarship in this area is conducted at the system, organization, product, process, service, and supply chain level of analysis. The Operations and Management Science area broadly considers the acquisition, development, and utilization of resources (human and capital) that firms use to deliver goods and services to their stakeholders. Managing operations often includes decision problems that involve uncertainty, competition, risk attitudes, and multiple (often competing) objectives. Problems with these elements arise in a variety of contexts and research in Operations and Management Science is naturally interdisciplinary. Methodologically, Operations and Management Science focuses primarily on formal economic models of decision-making and methods that apply formal models to practical decision problems. Recently, this modeling and methodology focus has been complemented by econometric, experimental, and field-based empirical research. 1.3 Sample Research Topics in OMS

• Supply Chain Management and Inventory Control • Sustainable Operations • Remanufacturing, and Closed-Loop Supply Chains • Innovation, New Product Development, and R&D Processes • Technology Management • Resource Allocation and R&D Portfolio Management • Organization Design and Incentives • Complex Adaptive Systems • Revenue Management • Service Operations • Utility Theory • Financial Engineering • Decision Analysis

2

1.4 OMS Faculty Below is a current list of OMS faculty, including their doctoral granting institution and email address. Prospective students should feel free to contact any of the OMS faculty with questions about Darden or the OMS PhD program Samuel E. Bodily (MIT Sloan School of Management) Robert L. Carraway (Purdue University) Raul O. Chao (Georgia Institute of Technology) Edward W. Davis (Yale University) James R. Freeland (Georgia Institute of Technology) Yael Grushka-Cockayne (London Business School) Robert Landel (Georgia Institute of Technology) Casey Lichtendahl (Duke University) Anton Ovchinnikov (University of Toronto) Phillip E. Pfeifer (Georgia Institute of Technology) Gal Raz (Stanford University) Elliott N. Weiss (Wharton School, University of Pennsylvania) 1.5 Selected Research Abstracts Revenue Driven Resource Allocation: Funding Authority, Incentives, and New Product Development Portfolio Management (by Raul Chao, Stylianos Kavadias, and Cheryl Gaimon)

The first step in transforming strategy from a hopeful statement about the future into an operational reality is to allocate resources to innovation and new product development (NPD) programs in a portfolio. Resource allocation and NPD portfolio decisions often span multiple levels of the organization’s hierarchy, leading to questions about how much authority to bestow on managers and how to structure incentives for NPD. In this study, we explore how funding authority and incentives affect a manager’s allocation of resources between existing product improvement (relatively incremental projects) and new product development (more radical projects). Funding may be either fixed or variable depending on the extent to which the manager has the authority to use revenue derived from existing product sales to fund NPD efforts. We find that the use of variable funding drives higher effort toward improving existing products and developing new products. However, variable funding has a subtle side effect: it induces the manager to focus on existing product improvement to a greater degree than new product development, and the relative balance in the NPD portfolio shifts toward incremental innovation. In addition, we highlight a substitution effect between explicit incentives (compensation parameters) and implicit incentives (career concerns). Explicit incentives are reduced as career concerns become more salient. Preferences for Consumption Streams: Scale Invariance, Correlation Aversion, and Delay Aversion Under Mortality Risk (by Kenneth Lichtendahl, Jr. and Sam Bodily)

Lifetime financial decisions often require a decision analyst to elicit a decision-maker’s preferences for consumption streams. In assessing such preferences, the analyst might look for a set of reasonable conditions to check when selecting a utility form. We provide such a set of conditions and show that they lead to the multiplicative-expo-power (MEP) utility form. Some of our conditions involve tradeoffs under certainty and others relate to choices under uncertainty. In the deterministic setting, we invoke

3

increasingness, continuous differentiability, mutual preferential independence, and preferential scale invariance. In the uncertainty setting, we invoke componentwise risk aversion, a utility independence condition, and correlation aversion. We apply the MEP utility in a life-cycle consumption planning problem under mortality risk. In this con- text, we find that the correlation averse MEP utility is more realistic than and as tractable as the popular correlation neutral additive-power and additive-exponential utilities. We show that the correlation averse decision-maker prefers to consume more in life and leave less behind in death. In addition, we demonstrate that the correlation averse decision-maker is more averse to delaying consumption in the face of mortality risk.

Revenue and Cost Management for Remanufactured Products (by Anton Ovchinnikov) Demand cannibalization is a major concern for firms that consider offering lower-priced remanufactured (refurbished) versions of their products together with the higher-priced new products. This paper presents a new model for analyzing demand cannibalization and supports it with a study of consumer behavior. This study shows that there exists a segment of consumers who use the price of the remanufactured product to infer about its quality, and thus pricing the remanufactured product below a certain threshold can, in fact, decrease cannibalization and persuade such quality-cautious customers to purchase the higher-priced new products, while still attracting new price-cautious consumers who will purchase the refurbished products. The paper then embeds this logic into the firm’s price and quantity optimization and shows that as a result the firm remanufactures under broader conditions, charges a much lower price, and typically remanufactures more units – leading to an increase of profits from remanufacturing by up to a factor of two as compared with making decisions based on standard approaches. A Fractiles Perspective to the Joint Price/Quantity Newsvendor Model (by Gal Raz and Evan Porteus)

Pricing and quantity decisions are critical to many firms across different industries. We study the joint price/quantity newsvendor model where only a single quantity and price decision is made, such as a fashion or holiday product that cannot be replenished and where the price is advertised nationally and cannot be changed. Demand is uncertain and sensitive to price. We develop a method for easily finding the optimal price and quantity that applies to more general cases than the usual one in which uncertainty is either additive, multiplicative, or a combination of the two. We represent a quantity by its fractile of the probability distribution of demand for a given price. We use a standard approach to approximating a given distribution with a finite number of representative fractiles and assume that these fractile functions are piecewise linear functions of the price. We identify effects that are not usually seen in a joint price/quantity newsvendor model. For example, although the optimal quantity is a decreasing function of the unit cost, the optimal price can be nonmonotone in the unit cost and we shed insight into why. We illustrate that using a simplified structure of demand uncertainty can result in substantially lower profits An Integrated Decision-Making Approach for Improving European Air Traffic Management (by Yael Grushka-Cockayne, Bert De Reyck, and Zeger Degraeve)

We develop a multistakeholder, multicriteria decision-making framework for Eurocontrol, the European air traffic management organization, for evaluating and selecting operational improvements to the air traffic management system. The selected set of improvements will form the master plan of the Single European Sky initiative for harmonizing air traffic, in an effort to cope with the forecasted increase in air traffic, while maintaining safety, protecting the environment, and improving predictability and efficiency. The challenge is to select the set of enhancements such that the required performance targets are met and all key stakeholders are committed to the decisions. In this paper, we develop and implement a model to identify a preferred set of improvements to the arrival and departure procedures to and from

4

airports. We provide an integrated approach for valuing a large number of alternatives, while considering interactions among them. The model combines quantitative and qualitative expert assessments of the possible enhancements and identifies commonalities and differences in the stakeholders’ perspectives, ultimately recommending a preferred course of action. The model is currently being adopted by Eurocontrol as the formal trade-off analysis methodology supporting all enhancements’ decision-making discussions throughout the construction of the master plan. 2. Required Background PhD students in OMS must demonstrate a basic knowledge of management, mathematics, probability, and statistics. This requirement can be fulfilled from the student's prior course work (for example, a completed MBA and an undergraduate degree in Mathematics, Science, or Engineering). Students should obtain waivers from the OMS PhD Coordinator for those portions of the background requirements fulfilled prior to entering the doctoral program. 2.2 Management Breadth PhD students must demonstrate a breadth of knowledge in two areas of management outside of OMS. A PhD student who has already completed an MBA has met this requirement. Otherwise, the student must complete two graduate level courses at Darden offered outside of OMS. A student should take courses to satisfy the breath of knowledge requirement during the second or third year of study after developing a better understanding of those areas of management that best support particular research interests. 2.3 Mathematics, Probability, and Statistics Students are required to demonstrate knowledge of calculus, linear algebra, probability, and statistics. This basic knowledge in mathematics serves as a prerequisite for most of the required coursework. Any mathematics deficiency must be remedied during the first year in the doctoral program. In addition, students should consider the doctoral prep course offered by the UVA Department of Economics (ECON 509). This course is offered prior to the beginning of the fall semester. Note that the credits hours taken in this category may not be used to meet the PhD credit hour requirement since they represent prerequisites to the program 3. Required Coursework

A PhD student in OMS must complete a minimum of 59 credit hours in the following areas: (i) Economics, (ii) Operations Research and Management Science, (iii) PhD Seminars, (iv) other selected courses and independent study.

3.1 Economics (8 credits)

PhD students in OMS must complete two courses in Economics offered at the graduate level during the first year in the doctoral program:

• Microeconomic Theory I (ECON 7010) • Econometrics I (ECON 7710)

5

3.2 Operations Research and Management Science (24 credits)

To successfully undertake independent research, PhD students in OMS must demonstrate rigorous understanding of research methodologies. Research tools in the field of Operations and Management Science are both descriptive (empirically driven techniques related to the collection and analysis of data) and normative (modeling driven related to optimization). The following four courses (12 credits) must be completed, earning an average grade of B or better:

• Mathematical Programming (SYS 6003) • Stochastic Systems (SYS 6005) • Applied Multivariate Statistics (SYS 6013) • Linear Statistical Models (SYS 6021)

Each student must select four additional research tools courses (12 credits) from the following list, earning an average grade of B or better. These courses should be selected to support particular research interests (such as whether the student intends to complete a normative or empirical dissertation). Students should seek input from their research faculty advisor to identify the appropriate courses to support their research program. Moreover, for planning purposes, students should refer to the Department of Systems and Information Engineering for a tentative schedule of courses. A student may take a course that does not appear in this list subject to faculty approval of the substitution. Any four of the following courses (12 credits) must be completed, earning an average grade of B or better:

• Dynamic Systems (SYS 6012) • Decision Analysis (SYS 6014) • Discrete Event Simulation (SYS 6034) • Agent-Based Modeling of Complex Systems (6035) • Risk Analysis (SYS 6050) • Financial Engineering (SYS 6054) • System and Decision Sciences (SYS 7001) • Advanced Stochastic Processes (SYS 7005) • Time Series Analysis and Forecasting (SYS 7030) • Advanced System Simulation (SYS 7034) • Heuristic Search (SYS 7042) • Risk Analysis (SYS 7050) • Sequential Decision Processes (SYS 7052) • Multi-Objective Optimization (SYS 7054) • Response Surface Methods (SYS 7063) • Bayesian Forecast-Decision Theory (SYS 7075)

3.3 PhD Seminars (12 credits) Four PhD level Operations Management and Management Science seminars are required for letter grade, earning a B or better. This is the minimum requirement and many students will complete more than the four required, as described below. Typically, two PhD seminar courses are offered each academic year. Grades of A are not guaranteed in PhD seminars. Faculty will provide students with feedback during the course so that performance expectations are well understood. Also, students should request performance feedback from the faculty during the semester. Each PhD seminar will require some form of a research exercise. Examples of research exercises include a project, a literature review, suggestion of extensions to a set of papers, or student development of a topic suggested by the faculty. The research exercise is not

6

the equivalent of a summer research paper in terms of length, depth, or quality. However, the exercise can be part of a portfolio from which to start and/or continue the student’s research program. The faculty has developed the following policy regarding the doctoral seminars. The key driver of the policy is recognition that an academic career reflects a commitment to continuous investment in knowledge creation. First, while in residence and prior to successfully completing the comprehensive exam, a student should take any newly offered PhD seminar for letter grade. Second, while in residence and prior to successfully completing the comprehensive exam, a student should take any repeated PhD seminar topic on a pass/fail basis. We believe that knowledge is both reinforced and further developed when PhD seminars are repeated. Once a student completes the written comprehensive exam, newly offered PhD seminars may be taken pass/fail and repeated seminars may be audited. Lastly, if taking a PhD seminar pass/fail, a student is not typically expected to complete the research exercise (project).

• R&D and Innovation • New Product Development and Technology Management • Supply Chain Management and Inventory Control • Sustainable Operations, Remanufacturing, and Closed-Loop Supply Chains • Revenue Management • Quality Management and Six Sigma • Dynamic Programming and Optimal Control

3.4 Other Selected Courses and Independent Study (15 credits) To fulfill this 15-credit category, a student should select courses that support his/her career objectives. The PhD student should meet with the research advisor to jointly select these courses. Application oriented (as opposed to methodology oriented) courses that relate to OMS may be particularly appropriate for this 15-credit category. A student should register for an independent study course with an OMS faculty while undertaking research that involves direct and substantial guidance from that faculty. In addition, a student may take an independent study course as a directed reading in support of his/her research interests. The 15 credit hours in this category are NOT part of the PhD Comprehensive Exam content. 4. First and Second Year Research Requirement Prior to embarking on the PhD dissertation, faculty will help students develop the capability to create original research of publishable quality. Immediately upon entry into the program, students should begin working on research under the direct supervision of one or more OMS faculty. Initially, the faculty largely directs the research. Over time, however, the balance shifts until ultimately the student drives the research activity with faculty supervision and input. In this section, we describe the research activities students undertake prior to the doctoral dissertation. 4.1 First Year Research Requirement Throughout the first year of the program, each PhD student works with a faculty on research. The purpose of the first year research activity is to afford students a concrete research experience. The faculty member will design a research activity that best integrates his/her research program with the student's interests and background.

7

The first year research project may develop from a project in a PhD seminar, as a result of coursework, or as a result of discussion with faculty. Under the direction of at least one OMS faculty, a student devotes extensive effort to a selected research problem. The purpose of the first year research project is to develop the student’s research skills (including creativity, rigor, written, and oral communication). BY THE END OF OCTOBER OF THE SECOND YEAR, STUDENTS MUST PRESENT THEIR RESEARCH EFFORTS. In addition, we require the following: (i) A written document describing the work must be available at least two weeks prior to the presentation. (ii) The seminar must be announced two weeks in advance to the entire management faculty. (iii) The presentation must be scheduled during an academic session (e.g., between the first day of class and the last day of exams of fall or spring semester). 4.2 Second Year Research Requirement During their second year of study, students should obtain faculty approval of a research problem they wish to pursue as their second year research activity. It is the student’s responsibility to interact with faculty to obtain approval of the research problem and a commitment from the faculty to supervise the research activity. The purpose of the second year research project is to further strengthen the student's research skills (including creativity, rigor, written and oral communication). The student should demonstrate that she/he is developing the necessary skills to define and analyze a research concept, and to write-up the research content in a clear and concise manner, representative of papers published in major OMS journals such as those covered in PhD seminars. Relative to the first year research project, limited faculty participation will occur. BY THE END OF OCTOBER OF THE THIRD YEAR, STUDENTS MUST PRESENT THEIR RESEARCH EFFORTS. The policy described for the first year research presentation applies here as well. 5. Teaching Requirement Prior to their first teaching experience (which typically occurs in the third or fourth year of the program), a PhD student must develop a comprehensive understanding of the material and the teaching approach used for that course. During the first year in the PhD program, the student will be assigned a teaching mentor. The student should attend courses offered by their teaching mentor on a regular basis. While not a full teaching assistant (TA) for the course, the PhD student may be asked to help the faculty with classroom preparation, teaching, or grading. During their third or fourth year in the doctoral program, students will be responsible for teaching a course on their own. 6. Written Comprehensive Examination Students must complete all coursework in Section 3.1, 3.2, and 3.3 (44 credit hours) before taking the comprehensive exam. A student may complete all of the 15 credits referred to as "other selected courses and independent study" (Section 3.4) after taking the comprehensive exam. Typically, the earliest a student is ready to take the written comprehensive exam is after they complete their fourth semester in the doctoral program. The exam provides students with an opportunity to synthesize various research topics to develop a holistic view of OMS. The exam reflects the broad content of the PhD program in OMS with particular emphasis on the OMS PhD seminars. The exam also reflects methodological know-how as presented either in OMS PhD seminars or in required methodological courses.

8

7. Dissertation Proposal and Defense Upon successfully completing the Comprehensive Exam, a student is admitted to PhD Candidacy. At this time, a PhD student is formerly permitted to begin work on the PhD dissertation. Typically, a student has already begun work that will become a part of the PhD dissertation prior to candidacy, e.g., the first and second research papers may ultimately become part of the PhD dissertation. During the third year in the doctoral program, students should continue to build their research expertise. Over time, additional emphasis is placed on the student’s initiative in identifying and solving research problems. Similarly, written and presentation skills are further developed. Although students may be working somewhat independently, we urge students to continue to seek input and direction from faculty on a regular basis. Since students should have already completed their research tools courses and should have sufficient teaching experience, they are expected to undertake research activities even while they are teaching. With guidance from the faculty, students will develop a Proposal for the PhD Dissertation. The Proposal should demonstrate that a student has the ability to (i) think independently, (ii) take initiative developing and structuring the research problem, rigorous solution, and analysis, (iii) write-up the research in a clear, organized and concise manner representative of papers published in major journals, and (iv) offer a research seminar presentation close to the quality necessary for successful job interviews. THE PROPOSAL PRESENTATION SHOULD BE MADE EARLY IN THE FOURTH YEAR. Throughout the fourth and fifth year of study, students should complete their PhD Dissertation. During the preparation of dissertation, students are urged to have timely and consistent communication with their dissertation committee. The research is presented at an oral defense open to the entire faculty of the Darden School of Business. 8. Faculty Advisors and Dissertation Committee Prior to arriving on campus, newly admitted students should seek information on faculty research interests. If possible, incoming students can provide input to the OMS PhD Coordinator indicating areas of overlapping interest with the faculty. Based on student input and availability of the faculty, the OMS PhD Coordinator will make an initial assignment of each incoming student to a team of faculty advisors (one of which must be a tenured faculty at the Darden School) for the purpose of first year research activities. The initial faculty-student assignment will hold for at least one semester; however, the assignments may be shifted as better fits are identified. Once a student has successfully passed the comprehensive exam, they must form a dissertation committee. The committee consists of 5 people: an advisor or co-advisors (at least one of which must be a tenured faculty at the Darden School) that are faculty in OMS, and at least one member that is external to the OMS faculty. 9. Performance Milestones and Feedback To help guide students and ensure the successful and timely completion of the PhD program, several opportunities have been identified for formal feedback from the faculty. The outcomes of the milestone reviews, as indicated below, are critical to determine the status and continuity of the student in the PhD

9

program. Please note that the OMS faculty are also interested in input from the doctoral students aimed at improving the quality of the PhD program. Program Approval. Each semester prior to registration, a PhD student must have his/her course schedule approved. First year students should seek approval from the OMS PhD Coordinator. Thereafter, students should obtain approval from their faculty research advisor. Students must submit an updated course plan and transcript to the OMS PhD coordinator each semester. Semester Grade Report. At the end of each semester, students must share their grades with the OMS PhD Coordinator and their faculty advisor. The faculty will assess the level and quality of progress being made in course work (as well as the balance between course work and research) and provide any needed guidance. First-Year Student Review (May of the first year). This milestone review reflects course work (including PhD seminars), research activities, and any teaching experiences. The assessment reflects the input of all OMS faculty. The student's research advisor and at least one other OMS faculty member will meet with the student to convey the milestone feedback. In addition, a written assessment will be placed in the student's file. First Research Paper Presentation (October of second year). This assessment is a milestone review that takes place in a meeting following the student's first research paper presentation. The assessment reflects the views of all OMS faculty. The OMS faculty will issue a pass or fail for the research presentation. It is also possible to be given a conditional pass. If this occurs, the conditions required for passing the first research paper will be clearly spelled out. Second-Year Student Review (May of the second year). This milestone review reflects course work (including PhD seminars), research activities, and any teaching experiences. The assessment reflects the input of all OMS faculty. The student's research advisor and at least one other OM faculty member will meet with the student to convey the milestone feedback. In addition, a written assessment will be placed in the student's file and shared with the student Second Research Paper Presentation (October of the third year). This assessment is a milestone review that takes place in a meeting following the student's second research paper presentation. The assessment reflects the views of all OMS faculty. While the structure of this review will be similar to the first year paper presentation, the bar for passing will be higher as students are expected to be improving their research capabilities. PhD Comprehensive Exam and Candidacy. Students should make every effort to complete the written exam no later than summer following the third year of study. Three outcomes are possible: (i) A student may pass the written exam. (ii) A student may be asked to complete a follow-up exam or take additional course work focused on specific topics where the original written response was deficient. If the outcome of this follow-up activity is satisfactory, the student has passed the exam and continues in good standing in the PhD program. (iii) A student may be asked to leave the doctoral program. Since the faculty is committed to providing feedback and working closely with doctoral students throughout their program of study, it is highly unusual for a student to reach this stage in the program and be terminated unless they are or have been on probation (see below). Third-Year and Subsequent Annual Student Reviews. The milestone review of students at the end of the third year reflects performance on the Comprehensive Exam (assuming this has been completed), course work, research activities, and teaching experiences. The assessment reflects the input of all OM faculty. The student's research advisor and at least one other OM faculty member will meet with the student to

10

convey the milestone feedback. In addition, a written assessment will be placed in the student's file and shared with the student. Students will be provided with a clear message regarding their progress and development in the doctoral program following each milestone review. The aim of this feedback is to keep students on track to successfully complete the program within a four year timetable. The faculty will assess a student's status from the following categories.

(i) A student may be given a pass and deemed in good standing.

(ii) A student may be given a conditional pass (such as on a first year research paper). If this occurs, the student will work with her/his research advisor (and possibly one additional OMS faculty) to remedy the deficiency and remove the conditional status. If the conditional status is not resolved in a reasonable amount of time, the student will be placed on probation (see (iii), below).

(iii) A student may be placed on probation. If this occurs, a clear understanding of how the probation

may be removed (including a timetable) will be given. An additional review is required as indicated in the terms of the probation.

(iv) A student may be terminated from the program. It is highly unlikely that termination will occur

unless a student is already on probation. Two consecutive terms of probation are likely to result in termination.

10. Additional Information and Administrative Details Research Assistant (RA) work. Each funded student is expected to do 10 hours per week of RA work for an approved faculty member as a condition of their funding. The assignments of RAs will be made each semester and will typically vary, both to ensure students are exposed to a variety of research approaches and to give faculty the opportunity to work with students in the program. Efforts will be made to see that RA assignments provide a “fit” for both student and faculty. University of Virginia Registrar: http://www.virginia.edu/registrar/index.html. The complete course catalog is available under the Student Information System (SIS) menu. University of Virginia Department of Economics: http://artsandsciences.virginia.edu/economics/index.html University of Virginia Department of Systems and Information Engineering: http://web.sys.virginia.edu/ Professional Societies:

• Institute for Operations Research and Management Science (INFORMS) • Production and Operations Management Society (POMS)

11

OMS Journals:

• Management Science • Operations Research • Production and Operations Management • Manufacturing and Service Operations Management • Journal of Operations Management • Decision Analysis • Decision Sciences • Interfaces (practice-oriented journal) • European Journal of Operations Research