MAPD maps out the future of analytics education

Analytics education best practices

Inaugural INFORMS Meeting of Analytics Program Directors meets the needs of fast-growing academic interest area.

By Melissa R. Bowers

The number of master of science (MS) in analytics degree programs increased from five in 2007 to 192 in 2016. Prompted by the rapid proliferation of analytics programs, INFORMS launched the first annual Meeting of Analytics Program Directors (MAPD) on April 1, the Saturday preceding the INFORMS Business Analytics Conference in Las Vegas. Sixty-three program representatives attended. The day-long meeting provided a forum for current and future directors of analytics programs to freely discuss issues and challenges common among such programs, share best practices and learn about industry’s view of graduate analytics programs.
Growth in Analytics Programs

In 2011, the McKinsey Global Institute forecast a shortage in the supply of deep analytical talent in the United States on the order of 140,000 people by 2018. In December 2016, in a follow-on report, McKinsey reported that while data and analytics capabilities have advanced significantly, the acquisition of analytics talent remains a challenge, and that “most companies are capturing only a fraction of the potential value of data and analytics” [1, 2]. To fill the analytics talent gap, universities continue to add graduate programs in analytics, primarily MS programs in analytics, business analytics and data science through a mixture of part-time, online and full-time formats as depicted in Figure 1. In fact, the MS program market experienced 18 percent and 29 percent increases in U.S. full-time MS programs in analytics and business analytics programs, respectively, from 2015 to 2016 as shown in Figures 2 and 3, while the number of full-time MS programs in data science remained unchanged from 2015-2016 as displayed in Figure 4. The data sources for Figures 1-4 were the North Carolina State Institute for Advanced Analytics (fall 2016) website and individual program websites (fall 2016-spring 2017). The North Carolina State website included full-time, part-time, online, other and blended MS programs in the United States.

Connection Between Industry and Analytics

Melissa Bowers leads a session at the 2017 MAPD in Las Vegas.

Because an interaction with industry is central to the success of any graduate analytics program, the meeting kicked off with an industry presentation entitled “An Industry View of Graduate Analytics Programs” by David Dittmann, director of Business Intelligence & Analytics at Procter & Gamble. Drawing from 20 years of experience in operations research and analytics at P&G, Dittmann presented an overview of analytics at Procter & Gamble followed by an insightful assessment of the industry needs of MS in analytics programs.

Dittmann pointed out that to be successful analytics professionals in industry, students must possess strong technical, programming and model-building skills along with strong database management skills, since, based on his experience, a significant portion of analytics time is spent dealing with messy data. However, he emphasized that students must not rely simply on garnering a command of analytics tools, but must gain a fundamental understanding of analytics methods and business principles to accompany their expertise in computing and database systems.

According to Dittmann, to prepare graduates to succeed in industry, analytics programs must emphasize the importance of adapting and continuously learning new tools necessary to solve new and existing analytics problems rather than learning a specific tool. As such, it is critical that students gain experience working on real-world analytics problems as a part of their curriculum. Through that experience, it is imperative that students learn the necessary skills to communicate their ideas, insights and analytic results clearly to sell their analytic recommendations. Finally, Dittmann emphasized the value that a demonstrated passion for analytics along with the ability to think creatively and critically contribute toward a successful industry career in analytics.

In summary, analytics programs must maintain strong industry ties to make certain that their graduates meet the evolving needs of employers to promote placement, but more importantly, to provide analytics students the opportunity to solve real-world analytics problems as a part of their curriculum to ensure relevance.

Student Recruiting and Placement

Figure 1: Growth in master of science programs in analytics (A), business analytics (BA), data science (DS) and other (O).

The success of any analytics program also depends on both student recruiting and placement as they form synergistic “bookends.” Recruiting the right students into a program directly impacts the placement outcome. Likewise, a successful placement history positively impacts student recruiting.

Michael Rappa of North Carolina State University provided specific insight during a session entitled “Recruiting the Right Students and Placing them in Internships and Permanent Positions,” which was followed by a panel discussion facilitated by Terry Harrison (Penn State University), with panelists De Liu (University of Minnesota) and Betsey Voorhees (Michigan State University).

Rappa provided detailed admissions and placement data on North Carolina State’s MS in analytics program. At NC State, each student applicant participates in an in-person panel interview to help identify the set of incoming students best suited for the program that focuses on producing team players well-versed in analytics methodologies and tools with top-notch communication skills.

Voorhees described placement success at Michigan State through career fairs to facilitate both internship and permanent job placement. On the recruiting side, the panelists also endorsed methodologies to reach out directly to potential applicants via undergrad career fairs or through undergraduate quantitatively-oriented clubs, since it is critical to reach potential student applicants early to ensure enrollment in appropriate prerequisite courses. For students who have demonstrated strong quantitative abilities but lack the appropriate prerequisite classes, programs may choose to provide online or in-house resources to deliver the prerequisite knowledge base.

Capstone/Practicum Experience and the Curriculum

Parallel breakout sessions covered a variety of program approaches to two vital components of every analytics program: a capstone/practicum experience and the analytics curriculum.

Figure 2: Growth in MS analytics programs.

The capstone/practicum session provided brief overviews of two different MS in business analytics capstone experiences presented by Mike Fry (University of Cincinnati) and Melissa Bowers (University of Tennessee, Knoxville), along with an overview of the practicum experience in the MS in business analytics program at Wake Forest University presented by Jeff Camm.

The practicum experience at Wake Forest differs slightly from the capstone experience at the University of Tennessee, with the course structure and timing being the primary difference. The Wake Forest practicum is rolled out as three half-semester courses at the end of the students’ academic program, while at the University of Tennessee, the capstone is a semester-long course delivered in students’ final semester of study. In both, students work on real-world industry projects in teams, as the coursework guides them through standard project task requirements such as determining project scope, data acquisition and cleaning, analysis, generating insights and recommendations, and implementation. Both require an oral presentation to the industry client, along with a written technical report and software with documentation as applicable. Similarly, both also focus on soft skills development in the areas of leadership, teamwork and project management.

The capstone project at the University of Cincinnati is not limited to one standard prescribed format. Instead, the capstone is an individual student effort and may take one of three different forms. The student capstone project may focus on a research project sponsored by the Center for Business Analytics, an internship project or the extension of a course case analysis or project. For all capstone project options at the University of Cincinnati, the deliverable is a written technical report.

Important lessons learned include: standardize industry nondisclosure agreements, obtain project data well in advance, institute a standard formal statement of work with the industry partner, be prepared for a project requiring big data technologies, and let the students learn by doing.

The curriculum session provided overviews of an established MS in business analytics program, an MS in business analytics program ready to launch in the fall of 2017 and one in the early stages of development set to launch in the 2018-2019 academic year. Don Kleinmuntz, University of Notre Dame Chicago, described the curriculum of the university’s established MS in business analytics (MSBA) program with a course format similar to that of an executive MBA program. Harish Krishnan, University of British Columbia (UBC), and James Cochran, University of Alabama, both detailed developing MSBA programs.

Figure 3: Growth in MS business analytics programs.

Regardless of program format, in general, the analytics curricula include content in the areas of statistics, operations research, data mining, database management and computing taught in the context of descriptive, predictive and prescriptive modeling. However, the format, including credit hours, duration and mode of delivery, along with specific required and elective content, often differs. Because the students are employed while attending, the Notre Dame Chicago MSBA program is a weekend program taught in Chicago requiring a trio of four- or five-day residency periods on the South Bend campus. The program requires 30 credit hours of analytics coursework delivered over a 12-month period. While the program initially had an online component, it is now delivered entirely in a face-to-face environment. It is a lock-step program with no electives.

UBC’s MS in management and operations research program is transitioning to a MSBA program this fall. The new, 30-credit-hour program is taught in residence at UBC over a nine-month period. Students may choose an optional internship/capstone project that extends the program to one year. The curriculum is also a lock-step program.

The University of Alabama’s MSBA program, in the early stage of development, is designed to be flexible to suit the needs of students. Many topical areas are presented as a set of two courses so that students can earn a certificate after completing both classes in a series. The class structure is designed for coursework in the mornings Monday through Thursday, leaving afternoons and Fridays open for seminars, project work and case analysis. A seminar series will allow for flexibility to incorporate continuously evolving cutting-edge content.

Industry View of Academic Programs in Analytics

Pooja Dewan, chief data scientist at BNSF Railway, facilitated an industry panel that included Timothy Merkle of Steelcase, Inc., Zahir Balaporia of FICO, Cara Curtland of Hewlett-Packard, Catherine Gihlstorf of SAS Institute and Janine Kamath of the Mayo Clinic.

Panelists provided insight on the desired attributes for analytics hires. While keen technical skills are a requirement, they all agreed that a degree in analytics is sufficient to demonstrate technical ability. Thus, they seek students with business acumen, soft skills, an innate curiosity about data, good communication and presentation skills, domain knowledge, creativity, resourcefulness, the ability to deal with ambiguity, the ability to learn from failure, and students who are team players, quick learners and “big picture thinkers.”

Figure 4: Growth in MS data science programs.

To turn out graduates with these desired characteristics, the panelists indicated that analytics programs should broaden their academic programs to include instruction on creative thinking, business problem framing, soft skills and the elevator pitch, as well as an internship or work experience.

INFORMS Service to Analytics Programs

INFORMS created a centralized online tool and data repository to help analytics program constituents learn more about the growing population of analytics programs. Sharif Melouk (University of Alabama) took the audience for a test-drive of the INFORMS site. Melouk pointed out many features on the site including:

  • resources of interest to analytics faculty members such as course syllabi, cases and data sets, along with access to INFORMS analytics publications and Edelman and Wagner Prize competitions;
  • a prototype academic program database populated with a static, limited set of program data assembled during the last several years;
  • relevant resources for current and future program directors, as well as administrators seeking to start a program in analytics;
  • information for those seeking either the CAP® (Certified Analytics Professional) or aCAP (Associate Certified Analytics Professional) certification; and
  • information for industry executives who wish to partner with an analytics program or hire analytics professionals.

A Call to Action

As the number of analytics programs continues to grow, a data-driven formal program ranking is inevitable. As such, Terry Harrison facilitated a session entitled, “Making Analytics Program Data Available – A Design Discussion,” to propose that analytics program directors collectively design an analytics program database provided by and curated directly from the source – directors of analytics programs themselves. As Harrison pointed out in his presentation, the database will serve many purposes, the most important of which is to provide comprehensive, transparent, accurate and timely data for use by potential analytics students in identifying the analytics program for which they are best suited. Thus, the programs will be incentivized to continuously update their program data to ensure that potential students always have access to the most up-to-date program data.

In addition, the repository will serve as a source of benchmarking data for program directors, establish a cooperative environment for the exchange of best practices, and deliver value to the discipline as the data will be created and distributed by the analytics programs themselves instead of a third party. It will require time, effort and collaboration among analytics program directors to accomplish this task. To start the development process, we have established a forum for discussion as an INFORMS Connect site. For more information on the site, contact Bill Griffin (bgriffin@informs.org) at INFORMS.

Future Vision

MAPD is committed to providing a forum to share best practices, spawn new ideas, discuss common concerns and provide networking opportunities to benefit analytics program directors. As such, we envision the MAPD meeting agenda evolving each year to meet the dynamic needs of both current and future analytics program directors in the rapidly changing analytics discipline.

The 2017 MAPD survey of attendees showed that 100 percent of those responding indicated that they were either very likely or somewhat likely to attend this event frequently. Thus, we are proud to say that the first annual Meeting of Analytics Program Directors was a success.

The 2017 meeting committee, chaired by Melissa Bowers, included Jeff Camm, Pooja Dewan, Mike Fry, Bill Griffin, Terry Harrison, Sharif Melouk and Jill Wilson. Thanks to all.

The second MAPD meeting is scheduled for April 14, 2018, the day before the 2018 INFORMS Business Analytics Conference in Baltimore. Make plans now to attend.

Melissa Bowers (mrbowers@utk.edu), associate professor and Beaman Professor of Business, is the director of the MS in Business Analytics program in the Haslam College of Business at the University of Tennessee Knoxville. She chaired the inaugural 2017 Meeting of Analytics Program Directors in Las Vegas.
References

  1. James Manyika, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles Roxburgh and Angela Hung Byers, “Big Data: The Next Frontier for Innovation, Competition and Productivity,” McKinsey Global Institute, May 2011.
  2. Nicolaus Henke, Jacques Bughin, Michael Chui, James Manyika, Tamim Saleh, Bill Wiseman and Guru Sethupathy, “The Age of Analytics: Competing in a Data-Driven World,” McKinsey Global Institute, December 2016.