New Master’s or Doctoral Data Science and Analytics Programs

Steve Pierson, ASA Director of Science Policy

The proliferation of master’s and doctoral programs in data science and analytics continues, seemingly due to the insatiable demand of employers for data scientists. Amstat News started reaching out two years ago to those in the statistical community who are involved in such programs to find out more. Given their interdisciplinary nature, we identified programs involving faculty with expertise in different disciplines—including statistics, given its foundational role in data science—to jointly reply to our questions. We have profiled many universities in Amstat News, including a few in January’s issue; here are five more.

Virginia Tech

Tom Woteki is professor of statistics at VA Tech and director of the MA program in data analysis and applied statistics in the National Capital Region. Previous positions include chief data scientist and chief technology officer for various corporations; chief information officer for the American Red Cross; and faculty appointments at Princeton and The University of Texas at San Antonio. His has a PhD in statistics and a BS and MS in mathematics, all from VA Tech.

MA Data Analysis and Applied Statistics (DAAS)

Year in which first students are expected to graduate: Summer 2021
Number of students currently enrolled: First admissions will be fall 2019
Partnering departments: Lead department is statistics
Program format: Eleven courses (33 credit hours), including a capstone project; four of 11 will be electives from partner departments emphasizing application in another discipline. Combination of online and in-person. Student profile is early to mid-career professional or government worker working full-time.

Describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
The curriculum comprises 11 three-credit courses. The core curriculum consists of seven courses that cover essential methods and tools in data analysis, applied statistics, computing, data visualization, and communication. These include three applied statistics courses that provide a solid foundation for developing, applying, and interpreting commonly used methods and a statistical computing course based on R. The four electives will include quantitative courses offered by other Virginia Tech academic departments so students can focus their data analysis skills on an area of interest to them or their employers (e.g., data science, economics, public health).

The data science option involves the four elective courses. These include two advanced courses that equip students with a wide variety of techniques considered a standard part of the data scientist’s toolkit and an advanced statistical computing course, all of which can be taught by our statistics or computer science departments. The fourth course details experimental design for standard situations such as A/B and multivariate testing, as well as providing best practices for planning cost-effective and efficient data-driven projects. We think this course is one that distinguishes our degree from other data science degrees we have seen.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?
As in many other parts of the country, there is a huge demand in the National Capital Region (NCR) for an analytically sophisticated workforce. Amazon’s selection of the NCR as one of its two HQ2 locations has heightened the demand. The DAAS degree program responds to this demand and will be part of VA Tech’s recently announced $1 billion innovation campus, which will be located in the NCR.

How do you view the relationship between statistics and data science/analytics?
In our view, contemporary applied statistics and data science are different aspects of the same discipline. The primary objective of both is to extract information and insights from data for evidence-based decision-making, predictions, and classification. Both require the ability to analyze and interpret data sets of various provenance, type, and size, and both share a wide variety of algorithms, methods, and visualization techniques, which, in turn, require significant computing and data management skills.

Grounding our degree in statistics with an emphasis on data science distinguishes it from other degrees that emphasize other aspects of the latter. Our program provides a strong statistical foundation so data scientists can be confident in their analyses. Graduates of this program will be able to analyze and mine data in all of its various forms using appropriate techniques.

What types of jobs are you preparing your graduates for?
Employers in the NCR include the federal government and companies focused on it; major corporations in the health care, financial, and consulting sectors; and a diverse array of startups in these and other sectors, including social media. Our program is designed to prepare students for successful careers with any of these employers.

What advice do you have for students considering a data science/analytics degree?
Students might consider whether they want a degree in “pure” data science/analytics or a degree program that provides them the opportunity to combine solid training in analytics with training in a field in which they would like to apply their data analytic skills. Our program, with its electives, allows them to pursue either path.

Describe the employer demand for your graduates/students.
We have reached out to prospective employers in both the public and private sectors seeking input on the design of our program. Indications are that the demand for our graduates will be very high.

Do you have any advice for institutions considering the establishment of such a degree?
We have taken the approach of building out from one of the core data science disciplines: statistics. Other common choices are mathematics and computer science. Including four elective courses in our curriculum gives us the flexibility to develop partnerships with other departments across VA Tech to tailor the degree to a wide variety of interests, both internal to the university and externally.

University of Notre Dame

Roger Woodard joined the University of Notre Dame in 2017 as the inaugural director of its new online MS-ACMS data science graduate program after spending 14 years at North Carolina State University. At NCSU, Woodard was a teaching professor in the department of statistics and, since 2013, director of online programs in the department of statistics. Woodard graduated cum laude from Culver Stockton College with a BA in mathematics in 1992. He earned a PhD in statistics from the University of Missouri.

Master of Science in Applied and Computational Mathematics and Statistics: Data Science Specialization

Year in which first students are expected to graduate: 2019
Number of students currently enrolled: 73
Partnering departments: Applied and computational mathematics and statistics (with additional faculty from the Mendoza College of Business and the Office of Information Technology)
Program format: This is an online program with 30 credit hours. The students in our program are part-time and most have employment and other commitments on top of that.

Describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
The online master’s program in data science at the University of Notre Dame was created with a great deal of input and advice from our industry partners. Several partners were surveyed about the desired content of the program as it relates to real-world application.

Prior to beginning the program, we suggest students have basic coding abilities; however, some students begin with little or no coding experience and succeed.

The curriculum was built to help students develop three major skill sets: technical expertise; communications; and ethics. The curriculum is designed to be lock-step so students take all courses with their cohort. This promotes cohort bonding and comradery that we find to be unique in an online program.

Programming in R and Python is woven into a majority of the courses in the program and supplemented by exposure to technologies such as SQL, Hadoop, and Tableau. Students are taught to have a strong conceptual understanding of statistical methods as they progress from basic probability and statistics through linear modeling, behavioral data science, generalized linear modeling, and time series courses. The students also complete a series of courses that expose them to machine learning, image analysis, and deep learning. The capstone course challenges students to use the skills they have developed to help a company with a real-world data problem. We think this is an excellent way for students to have a hands-on experience with the data science skills they have developed.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?
When Notre Dame’s senior leadership decided on the online data science degree, they identified it as an exciting multidisciplinary field growing exponentially in social relevance and industry impact. Also identified was a need for highly skilled and multidimensional professionals. The senior leadership thought Notre Dame could provide a distinctive learning experience rooted in statistical and computational methods. In addition, and even more crucial, they thought Notre Dame could provide the ethical training and communication strategies necessary for long-term success in today’s workforce.

Having our leadership working closely with visionary industry leaders such as AT&T’s John Donovan enabled Notre Dame to build an online curriculum and program structure to serve the needs of working professional students located around the world.

How do you view the relationship between statistics and data science/analytics?
Statistics is one of the many disciplines that have come together in our program. We see that statistics brings in a vital set of tools for thinking about data. We work hard to help our students develop a strong conceptual understanding of sources of variability and their impact on models.

In the realm of big data, with millions of observations, traditional techniques like hypothesis testing may be less interesting, but thinking about sampling, sampling variability, and model fitting are still relevant.

In our program, we work hard to bring together statistics, computer science, behavioral sciences, communication, and ethics in a program that stresses conceptual understanding, not just techniques. By building this deeper understanding, we try to develop data scientists who can adapt to changes in technologies and methodologies.

What types of jobs are you preparing your graduates for?
We train our students to be “data scientists,” though we realize this role will look very different depending on the industry and employer. We want our students to be able to apply various methodologies and consider the ethical impact their decisions have. We want them to be able to collaborate with different subject-matter experts and tell the story of the data they are working with.

The students in our program tend to be mid-career and thus have more corporate and industry experience than students in many other programs. This experience means our students are more likely to be able to obtain more senior and managerial positions once they graduate with their new skill sets.

Our program stresses flexibility and the collaborative nature of data science. It is preparing our students to adapt in any situation and use critical thinking in understanding data science methodologies. Data science is an ever-changing and always-evolving field, and we want our students to be prepared to handle any situation that comes their way.

What advice do you have for students considering a data science/analytics degree?
When considering a program, it is always beneficial to talk to the students who are in the program or have gone through the program. They can offer insight into what the daily life and culture is and how they balance it with other responsibilities.

You also want to look at the faculty and student interaction. We pride ourselves on the faculty and student interaction in our program. We know that, in an online program, there is the potential to feel isolated. We counter that with building close cohorts and fast response times from faculty and staff. It is important to learn from your faculty member, but it is also important to learn from your fellow classmates, as everyone brings something unique to the table.

Describe the employer demand for your graduates/students.
The field of data science is a quickly growing one. Currently, there is a high demand for data scientists. We graduate our first class this upcoming May and have already seen students accelerate in their careers. One of our corporate partners told us they have hundreds of job openings for data scientists and, once they fill them, they will have even more.

Do you have any advice for institutions considering the establishment of such a degree?
When we created this program, we knew we wanted it to be unique and we wanted it to offer the Notre Dame experience, even though the students are not on the main campus. We start their program with an orientation on campus so they can meet their classmates, faculty, and staff in person before they start classes. We think this sets the students up for success because they know they are part of a family. They know that, when they face adversity, they can reach out to classmates, their faculty, and our program staff for help.

We designed this curriculum specifically for this program, and it was created in a way that weaves together our mission throughout the courses. We custom built each and every course with our Office of Digital Learning and our faculty so the students would have a holistic experience. By creating a digital campus with our NeXus platform and other tools such as Zoom, we have been able to create a program that brings the Notre Dame tradition and experience to our students all over the world.

Rutgers

Rong Chen is distinguished professor of statistics and director of the master’s in data science (statistics track) program in the department of statistics at Rutgers University in New Brunswick. He specializes in nonlinear, nonparametric time series analysis; Monte Carlo methods; and statistical applications in bioinformatics, business, and engineering. He is a fellow of the ASA and Institute of Mathematical Statistics.

Master of Statistics (Data Science)

Year in which first students graduated: 2017
Number of students currently enrolled: ~100
Partnering departments: Statistics and computer science
Program format: 10 in-person courses (30 credit hours), 8 required and 2 electives

Describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
The Rutgers Master’s in Data Science (MSDS) program was developed by the statistics department in collaboration with the computer science department, resulting in two structurally separated subprograms under a unified Rutgers master’s in data science program: the statistics track and computer science track. The statistics track, which awards the Master of Statistics degree, directs its focus toward the data analysis aspect of the data science field.

The MSDS (statistics) program is designed for highly motivated students with a bachelor’s degree in mathematics, statistics, engineering, computer science, economics, finance, or a related quantitative field. Multivariate calculus, linear algebra, introduction to probability, theory of statistics, and statistical computing or advanced programming courses are prerequisites.

The program requires students to successfully complete 10 courses (30 credit hours), of which eight are required courses and two are electives. A comprehensive project report, which may be an extension of one or two course projects the student has done during the study, is also required for graduation. The program also contains a practical training requirement, which can be fulfilled via an internship, full-time employment in the field, on-campus research, or participation in either the MSDS practitioner’s seminar series or the MSDS career workshop series.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?
With increasing data collection and advanced computational abilities, the analysis of massive data and the extraction of useful information from it have become major emerging frontiers of computer science, statistics, and related fields. Analytical tools have become indispensable to industry and society in general, creating a strong surge in demand for professional data scientists.

The statistics faculty at Rutgers are actively engaged in state-of-the-art research in data science in both theoretical development and applications.

Rutgers is located in the tri-state area, with a vast number of companies seeking talent in data analytics. Rutgers is ideally positioned to offer a data science program that produces high-quality data science professionals to meet local and national demand. The number of applicants to the program has grown many folds since the inception of the program, and we expect the trend to continue.

What types of jobs are you preparing your graduates for?
The Rutgers MSDS (statistics) program trains students to become professional data scientists working in all industries, consulting firms, federal and local governments, and various organizations. They are highly trained in data collection and survey, data wrangling, statistical analysis, machine learning, information communication and visualization, and informed decision-making.

What advice do you have for students considering a data science/analytics degree?
An advanced degree is desired for a student who wishes to become a professional data scientist due to the interdisciplinary nature of the field and the vast range of tools needed to master. A good program should provide a mix of theoretical foundation, methods, and practical experiences. The program should not be a training program that only offers tools. A deep understanding of the state-of-the-art methods with extensive practice experiments on how the methods are used in practice and how the results are interpreted is vital. The program should also offer extensive career development services, with professional staff experienced with industry relationships and networking skills. Rutgers’ MSDS program was designed under these considerations.

Describe the employer demand for your graduates/students.
Data scientist has been named the best job in America for three years running, according to Glassdoor’s 2018 rankings. All our graduates have found high-paying jobs, with a small portion going to PhD programs. We see demand coming from financial services, technology, consulting, government, and pharmaceuticals, just to name a few.

Do you provide career development services?
Yes. The MSDS (statistics) program at Rutgers is a professional master’s degree program. A dedicated professional services office manages career development activities, industry relationship activities, and outreach activities. It assists students with résumé and cover letter writing, conducts mock interviews, and provides career advice. We have also established long-term strategic partnerships with companies in terms of internship positions and research collaborations.

Do you admit part-time students?
Yes. The program is designed to fully accommodate part-time students. All classes are evening classes and meet once a week, so students can attend classes after work. Most of our part-time students complete the program in five or six semesters (including summer sessions).

University of Alabama

Nickolas Freeman is an assistant professor of operations management in the department of information systems, statistics, and management science in the Culverhouse College of Business. Freeman’s research interests include health care operations management, supply chain management, and applied analytics.

 
 

John Mittenthal earned his doctorate in industrial and operations engineering at the University of Michigan – Ann Arbor. He applies his expertise in optimization to practical-based problems and is a member of The University of Alabama Business Analytics Symposium organizing committee.

 
 

Denise McManus earned her PhD from Auburn University. Prior to embarking on an academic career, McManus gained professional experience through development and delivery of information systems at Motorola, Boeing Computer Support Services, and International Business Machines (IBM).

 
 

Jan Jones, director of specialized master’s programs, joined Culverhouse College of Business in 2005. She has served in both the management and marketing departments and Manderson Graduate School of Business. Jones earned her BBA with a minor in economics from the University of Mississippi.

 
 

Sharif Melouk is the associate dean of the Manderson Graduate School of Business and a professor of operations management at The University of Alabama. He earned his PhD from Texas A&M University and currently conducts research in health care operations and behavioral operations management.

 
 

Master of Science in Business Analytics (MSBA)

Year in which first students are expected to graduate: 2020
Number of students currently enrolled: Anticipate 20–25 in first cohort
Partnering departments: The MSBA is a 36-credit hour, full-time program in which students progress as a cohort. All instruction is in person on The University of Alabama campus in Tuscaloosa. All classes include project work, with some projects spanning multiple classes.

Describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
The University of Alabama’s MSBA program was developed with the needs of companies looking for analytics talent in mind. In particular, our curriculum planning was informed by survey responses from professionals involved in the assessment and acquisition of analytics talent at well-known companies. Based on insights regarding essential skills and trends garnered from the survey responses, our curriculum focuses on the following:

  • Teaching students the principles of acquiring and managing data
  • Providing students with hands-on experience in the use of cutting-edge analytics techniques by using data from, or inspired by, real-world problems
  • Teaching students how to use software packages and programming languages commonly employed for analytics projects, including SAS, SQL, Python, Tableau, and R
  • Teaching students how to interpret results, communicate results effectively, and use tools that promote reproducibility
  • Making students aware of the ethical implications of analytics tasks
  • Educating students about the landscape of the analytics community and the various opportunities for continued skill development throughout their career

Our program also includes 90 contact hours of seminars that will cover important new topics not covered in the core courses and provide more in-depth coverage of topics from the core courses. Faculty and experienced analytics professionals will teach these seminars. Furthermore, graduates from the program will be given access to new seminar materials that have been developed after they have graduated from the program to enable their continued professional development.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?
Although we have seen an increase in the number of business analytics programs offered by degree-granting organizations, our survey of industry professionals—along with online sources of job market information—suggest there is still a substantial gap between the supply and demand for analytics talent. Moreover, we believe the long tradition of analytics education and research at the Culverhouse College of Business has provided experience and insight that allow us to offer an exceptional analytics educational program.

For example, through our Institute of Business Analytics, we host an annual symposium (bit.ly/2HqKL0u) that has featured speakers from General Motors, the US Centers for Disease Control and Prevention, the Public Broadcasting System, ESPN, Regions Bank, Deloitte, Proctor & Gamble, USAA, Booz Allen Hamilton, Overstock.com, SAS, Lockheed Martin, and others.

At this point, the reaction from students has been extremely positive.

How do you view the relationship between statistics and data science/analytics?
Statistics provides the theoretical foundation upon which data science and analytics build. Essentially, statistics provides the mathematical raw materials necessary for collecting, organizing, and analyzing data. Using the mathematical toolkit provided by statistics, data science and analytics seek to identify meaningful combinations and implementations of the tools that can solve important business problems at the scale required by today’s companies.

What types of jobs are you preparing your graduates for?
We aim to produce graduates suited for entry-level analytics positions in manufacturing and service companies, as well as government institutions.

What advice do you have for students considering a data science/analytics degree?
First, we would advise students to use the many resources available online to fill any gaps in their mathematics and statistics knowledge.

Second, similar resources can be used to gain some level of familiarity with an open-source programming language applicable to statistical computations such as R or Python.

Third, students should seek out opportunities in which they can learn about the types of problems that can be addressed using data science/analytics. Several good podcasts such as Linear Digressions, the O’Reilly Data Show Podcast, and Not So Standard Deviations provide such exposure.

Also, competitions hosted by Kaggle provide a good perspective on problem types and possible data sources/formats. The posted kernels also provide insight into the skills students will need to gain to be competitive upon graduation.

Do you have any advice for institutions considering the establishment of such a degree?
Concerning the challenge of putting together an interdisciplinary program, the first step is clearly defining the mission of the program, which must be consistent with some existing or developing societal need. Once the mission is defined, it can be broken into steps that represent a road map.

Next, faculty who can contribute to the various milestones on the roadmap should be identified. It is worth noting at this point that the composition of the faculty team should emphasize not only expertise but also the potential for collaboration. The emphasis on collaborative potential is necessary because a successful analytics program will expose students to the end-to-end analytics process, from data acquisition to the communication of results and implementation. Faculty should design their courses and structure coursework and assignments in such a way that this end-to-end analytics process is clearly articulated and reproducible by the student upon graduation.

In preparation for program launch, we held several meetings with faculty to devise a program outline that streamlined coursework to minimize redundancies and identified potential data sets and projects that could be used across multiple courses.

Colorado State University

Jana Anderson is an associate professor and the director for the online learning and Master of Applied Statistics programs in the department of statistics at Colorado State University.

 
 
 

Master of Applied Statistics – Data Science Specialization

Year in which first students are expected to graduate: 2019
Number of students currently enrolled: 77
Program format: Offered both fully on campus and fully online. 30-31 credits are required. Consulting project required in place of thesis. Student type traditional, non-traditional, full-time, part-time, and continuing education. Internship program being developed. We currently require a consulting project with real client and real data. No assistantships are provided.

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
We cover probability and statistical inference, both with applications, regression, generalized regression, design of experiments, quantitative reasoning, statistical learning, data mining, machine learning, programming, database management, and linear algebra. Elective courses include mixed models, multivariate analysis, Bayesian analysis, business intelligence, and data exploration. The final class is a consulting project.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?
We have seen an increasing need for our MAS graduates to have a solid background in programming and database skills. Exposure to SQL and Python in the master’s program increases the number of options for students after graduation.

How do you view the relationship between statistics and data science/analytics?
Statistics provides a strong basis for studying data science. However, without coursework from other areas, the student’s background for working in data science will be incomplete.

What types of jobs are you preparing your graduates for?
Data analyst, business analyst, statistical analyst

What advice do you have for students considering a data science/analytics degree?
Recognize the need for a solid background in calculus and linear algebra, and look for a program that includes courses on statistical inference, regression, experimental methods and statistical/machine learning. Also essential are data visualization, database, and programming skills.

Describe the employer demand for your graduates/students.
Based on prior experience with our Master of Applied Statistics – Statistical Science Specialization, we expect to have interest from industry, financial investment firms, internet service providers, and marketing firms, among others.

Do you have any advice for institutions considering the establishment of such a degree?
Consider the strengths of your department and other departments in your institution. Having a good working relationship with, for example, math and computer science departments, can make a big difference in developing a quality program, as will the ability to work with your dean or associate dean in developing a cross-disciplinary program. Try to avoid turning out students who are able to “plug and chug” through procedures, without understanding how those procedures work.