Steve Pierson, ASA Director of Science Policy

More universities are starting master’s programs in data science and analytics, of which statistics is foundational, due to the wide interest from students and employers. Amstat News reached out to those in the statistical community who are involved in such programs. Given their interdisciplinary nature, we identified programs involving faculty with expertise in different disciplines to jointly reply to tour questions. We profiled a few universities in our April and June issues; here are several more, plus a few PhD programs.

 

Harvard University

David C. Parkes is the George F. Colony Professor of Computer Science at the John A. Paulson School of Engineering and Applied Sciences and the co-director of the Harvard Data Science Initiative.

 
 

Rachel Schutt is lecturer in data science and education program director of Harvard’s Institute for Applied Computational Science.

 

 

Neil Shephard is professor of economics and statistics and department chair of statistics at Harvard University and the co-chair of the Data Science Education Committee of the Harvard Data Science Initiative.

 

 

Master of Science in Data Science

Year in which first students graduated/expected to graduate: 2019

Number of students enrolled: We are anticipating a class of 40–50.

Partnering departments: Institute for Applied Computational Science, Statistics, Computer Science, Applied Math

Program format: This is an in-person, full-time program. Twelve courses are required, and the degree will typically be completed over three semesters. At least one research experience is required and can be satisfied by a capstone project course or a semester-length independent study project. A final requirement is the presentation of a poster on a data science project at the annual project showcase.

Additionally, PhD students in other departments can specialize in data science as a secondary field by completing a selection of course requirements.

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.

The basic elements of our curriculum are the following:

• Four required technical courses, including our full-year “Introduction to Data Science” sequence and a course in advanced scientific computing
• One statistics elective
• One computer science elective
• One “critical thinking in data science” course
• One research experience
• Four data science electives that can be satisfied across many departments

Our data science curriculum was developed as a joint effort between the statistics department and computer science working with the Institute of Applied Computational Science (IACS) within Harvard’s Engineering School. We already had in place a master’s in computational science and engineering program (run by IACS), which is now in its fifth year, and were attracting many students who placed into data science positions, so we wanted to expand our offerings to include a second master’s program in data science.

Additionally, we already had in place a one-year Introduction to Data Science course—now in its fourth year—jointly taught by IACS, statistics, and computer science faculty. This is a survey of all topics we think are essential to data science: the data science process, machine learning, data visualization, statistical inference, algorithmic and computational thinking, experimental design, best practices in coding, and ethics and algorithmic accountability. This course maps very well to a master’s curriculum and is one of the core required elements of the program.

We have a faculty standing committee, which includes faculty from across disciplines. We have an advisory board that includes industry experts from Google, Microsoft Research, and other leading companies and national laboratories. We met with them in a full-day event to review the curriculum of the new master’s program, clarify learning objectives, and understand employer needs.

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?

Data science is a rapidly emerging field. It includes a hybrid set of skills and thinking from statistics and computer science, along with new methodologies and skills not traditionally taught in universities. Students typically graduating from statistics programs or computer science programs were not being exposed to ideas from across disciplines. We needed to train students whose thinking could transcend departmental boundaries, as well as develop a new curriculum that gives students the foundations in these new skills (statistics at scale, experimental design at scale, algorithms for large data sets, building data products in production, and algorithmic accountability and ethics). A new curriculum needed to be created to train students with the foundation and discipline to be leaders in industry and research.

How do you view the relationship between statistics and data science/analytics?

The field of statistics is foundational for data science.

What types of jobs are you preparing your graduates for?

We are preparing students with computational and analytical skills to get jobs across sectors and disciplines. Students are able to work at technology companies like Facebook and Google; media companies like BuzzFeed and The New York Times; and finance companies like Morgan Stanley, Two Sigma, and Citadel. They also are able to work in start-ups, ed-tech, fin-tech, marketing, and consulting.

A few students each year go on to PhD programs in robotics, computer science, statistics, neuroscience, and business.

What advice do you have for students considering a data science/analytics degree?

If you want to build a computational and statistical foundation that can be applied across multiple application areas (e.g., finance, genetics, marketing), this is a good area to get into. If there is already an application area you are interested in, you should examine whether it may be better for you to deeply specialize in that area. However, we try to construct the program so as to build a strong foundation while giving students the flexibility to go deeper into their areas of interest through the electives.

Describe the employer demand for your graduates/students.

We don’t yet have any graduates for the master’s in data science. But, for our master’s in computational science, all our students get jobs before graduation. They tend to have multiple offers to consider. We’re highly selective in who we admit in the first place. Many students tend to have formed relationships with potential employers during the course of the program through internships and capstone projects.

Do you have any advice for institutions considering the establishment of such a degree?

Collaboration and a shared vision between the statistics department and computer science was essential. When you start with the premise that you want to create the best educational experience that you can for students, then organizational politics becomes secondary. While there sometimes are additional bureaucratic hoops when more than one department is involved, these aren’t insurmountable.

Part of the success of the program is due to having the Institute for Applied Computational Science already in place, with a full-time staff dedicated to administering these master’s programs and working to create a holistic experience for students, including advising, mentoring, community-building, job talks, seminars, and conferences for faculty and students.

The University of British Columbia

Milad Maymay earned his BS at The University of British Columbia. He has more than 17 years of experience managing projects and programs in both the nonprofit and public sectors.

 

Giuseppe Carenini has been teaching artificial intelligence, machine learning, and natural language processing for more than 10 years. He is also collaborating with companies that aim to make data more useful in supporting complex decisions (Compass) and for public engagement (Metroquest).

 

Paul Gustafson earned his BS in mathematics and master’s in statistics from The University of British Columbia and his PhD in statistics from Carnegie Mellon University. Gustafson also won the CRM-SSC Prize and is a Fellow of the ASA.

 

 

Master of Data Science

Year in which first students graduated/expected to graduate: 2017

Number of students enrolled: 45

Partnering departments: Departments of Statistics and Computer Science

Program format: 10-month professional master’s program, in-person, 30 credits, two-month capstone project

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.
The UBC Master of Data Science is a professional program harnessing the combined expertise of the UBC departments of computer science and statistics. It helps meet the growing need for people who can apply computational and statistical techniques to data and then effectively communicate results from analyses to various stakeholders.

Using descriptive and prescriptive techniques, students extract and analyze data from both unstructured and structured forms and then communicate the findings of their analyses in ways that promote informed decisions based on data. Multidisciplinary in nature, the Master of Data Science program enables graduates to span both the statistical and computational perspectives. The curriculum, informed by consultations with local industry, takes a scientific approach to use data to explore different hypotheses.

The program includes 24 one-credit courses offered four at a time, in four-week segments. Courses are lab oriented and delivered in-person. Graduates can appropriately select and tailor data science methods to deal with diverse data types (numerical, categorical, text, dates, graphs, etc.) across diverse subject-area domains.

The program also includes a two-month (six-credit) capstone project, allowing students to work alongside their peers with real-life data sets. In this project, students determine questions of interest for the data in conjunction with mentors drawn from academia, industry, and nonprofits. Students experience the complete data science value chain, applying techniques they have learned to investigate the questions relevant to the mentors.

Although professional experience is desired, it is not a mandatory requirement.

Students have backgrounds across a wide range of fields, including biology, business, engineering, and the social sciences.

Prerequisites include the following:

• One course in programming
• One course in probability and/or statistics
• One course in calculus or one course in linear algebra (completion of a course in both calculus and linear algebra is recommended)

What was your primary motivation(s) for developing a master’s data science/analytics program? What’s been the reaction from students so far?

In every domain, from health care and e-commerce to utilities and gaming, a staggering amount of data has led to a new field and an unprecedented demand. There is a growing need in many fields (especially in western Canada, the Pacific North West, and Silicon Valley) for people who can apply computational and statistical techniques to data and then effectively communicate results from analyses to various stakeholders.

Students’ reactions to the program have been extremely positive. Almost all the students in our first cohort—which completed in June 2017—are employed, and more than 500 candidates applied for September 2017 entry to the program.

How do you view the relationship between statistics and data science/analytics?

In this program, and generally on this campus, data science is viewed as being built upon a solid foundation in statistics and a solid foundation in computer science.

What types of jobs are you preparing your graduates for?

The extraordinary thing about data science is that our graduates can work in almost every industry and sector—government, education, health care, consulting, tourism, and technology. Depending on the background and experience of the students prior to the program, our graduates get jobs in entry- to mid-level positions as data scientists, data architects, or data analysts. Our graduates are working for organizations such as Microsoft, Electronic Arts, Visier, The University of British Columbia, and Translink (Metro Vancouver’s public transportation authority).

What advice do you have for students considering a data science/analytics degree?

Data scientists need to be familiar and comfortable with a variety of skills and tools. That is why our program is almost an equal combination of core computer science concepts and tools (about one-third of the courses), core statistics concepts and tools (another third), and more emerging and data science–specific topics (the final third). Students should be aware of this prior to choosing a career in data science. Those who prefer more programming and are less interested in interacting with domain experts should consider big data programs or computer science programs. But those who prefer variety in their day-to-day work or prefer to be generalists can consider a career in data science.

Describe the employer demand for your graduates/students.

The demand for our graduates/students is significant. We organized more than 15 employer/industry talks for our first student cohort. Furthermore, we received numerous job postings and requests from industry partners to participate in our capstone course. The demand seems to be increasing even further this year.

Do you have any advice for institutions considering the establishment of such a degree?

The program benefits from great collaboration between the department of computer science and the department of statistics. Drawing a program co-director from each department and having about a 50-50 split in resources and responsibilities has proved effective. Each department has pride and “skin in the game” when it comes to making this program successful.

 

Texas A&M University

Simon Sheather is the academic director of the MS (analytics) program at Texas A&M University. From 2005–2014, he served as professor and head of the department of statistics at Texas A&M University.

 

 

Jon (Sean) Jasperson earned his PhD in business administration with an emphasis in information management science from Florida State University.

 
 
 

Master of Science in Analytics

Year in which first graduated: 2015

Number of students enrolled: 110. Class of 2018 = 45. Class of 2019 = 65.

Partnering Departments: Texas A&M Department of Statistics and Mays Business School

Program format: The program is 36 credit hours, five semesters long, and taught live online to working professionals on a part-time basis. Full-time employment with at least three years of work experience is required. The program has no thesis option; rather, students bring data from their organization and build predictive models as a graduation requirement. Our work-based capstone project is what makes the program unique. Admissions are cohort-based and only offered in August.

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.

Taught in partnership with Texas A&M Mays Business School, the part-time five-semester curriculum in the program consists of 67% statistics and 33% business. The program is taught by distinguished faculty credited with receiving numerous teaching awards at the university level in addition to having countless citations in their field. The lectures use relevant case studies and real-world applications designed to teach students how to apply what they learn immediately. The curriculum was designed to teach students how to apply statistical methods using big data to solve business problems.

What was your primary motivation(s) for developing a master’s data science/analytics program?

To narrow the gap. The main motivation for developing the program was to help students and companies get a competitive advantage in analytics. The extensively quoted U.S. News and World Report analysis about the growing gap between people with skills to analyze data and those who don’t inspired us to get started. Wanting to be different and still help the cause, our program focused on working professionals and partnering with companies to train and develop their own.

To offer a unique experience with a work-based capstone project. The program graduated its first class in 2015, and the reaction since has been strong. Students have valued their experience in the course as first-rate—a true testament to the innovative curriculum, the distinguished faculty who teach, and the strong impact of the work-based capstone project. A significant number of the graduates have received generous promotions from their employers and been asked to expand and optimize analytics in their organizations.

Companies also value the program in this aspect, as they have sent more employees from different divisions to enroll after graduating employees in previous cohorts.

How do you view the relationship between statistics and data science/analytics?

The “buzz” word frenzy. Today, it’s data science. Tomorrow, it’s machine learning. The day after? Who knows? It’s difficult to pin an exact definition of what data science is given its broad reach—and compare it to a set discipline. Statistics/statistician is a “sexy” field/job in itself. The ability to predict with impressive accuracy, using sophisticated modeling and a touch of genius, is a unique skill that not many have. What should the relationship between statistics and whatever the buzz word be? Strong. A firm grasp on one yields astonishing results when used with the other.

What types of jobs are you preparing your graduates for?

The leaders in analytics. The MS analytics program prepares our students to be leaders and pioneers of analytics in their organizations. Designed for the working professional, our students are not active job seekers—rather, they receive generous promotions. Some have shared that they received promotions to chief data scientist, senior data scientist, and lead performance analyst. Our students are also asked to optimize and expand analytics within their organizations.

What advice do you have for students considering a data science/analytics degree?

Research. There are more than 150 analytics/data science programs offering degrees, advanced degrees, and certificates. Each one offers unique classes tailored to fit the wants and needs of prospective students.

What are your needs? Answer the following questions:

• What am I looking for in an analytics program?
• How will that specific program meet my needs, both personal and professional—short term/long term?

A degree in an area like data science/analytics offers a much broader skill set than traditional computer science/engineering or statistics. With such programs, students receive a healthy dose of both fields, with application capabilities in many areas across many industries.

Describe the employer demand for your graduates/students?

Strong demand. Since our program is for working professionals, we don’t track placement. But, we do receive a significant amount of inquiries with internship/job opportunities for our students that we share with them.

Do you have any advice for institutions considering the establishment of such a degree?

Be unique. With more than 150 programs out there and more being created, start with this question: What are the elements that will make your program unique?

Build strong partnerships. Survey your corporate contacts who seek to employ graduates and who would collaborate with your department.

Keep it fresh. Find ways to keep material fresh in class, materials that keep students engaged with relevant coursework.

 

University of Colorado, Denver

Katerina Kechris is an associate professor in the department of biostatistics and informatics at the University of Colorado Anschutz Medical Campus. Her research focuses on the development and application of statistical methods for analyzing high-throughput omics data.

 

Farnoush Banaei-Kashani is an assistant professor in the department of computer science and engineering at the University of Colorado, Denver. His research focuses on developing novel machine learning, as well as data mining and management techniques that enable big data lifecycle.

 

Biostatistics MS Emphasis in Data Science Analytics

Computer Science MS Track in Data Science in Biomedicine

Year in which first students graduated/expected to graduate: 2019

Number of students currently enrolled: 2–3/year under each track

Partnering departments: Biostatistics and Informatics, Computer Science and Engineering

Program format: The biostatistics MS is an in-person degree with 36 credit hours and a culminating thesis or publishable paper. Students are typically full time and participate in 1–2 years of a research assistantship. The emphasis in data science analytics is a focus area within the MS degree where students take three elective courses (total nine credits) from a list of courses related to data science, of which six credit hours will count toward their MS electives and the other three credit hours are required additional credit hours for the emphasis; hence, a total of 39 credit hours is required for graduation from the emphasis. Students in the emphasis will also write a thesis (or publishable paper) with a focus on data science.

The computer science MS is an in-person degree with 30 credit hours, including nine, 15, and six credit hours for core courses, elective courses, and thesis, respectively. Students are typically full time and participate in two years of a research assistantship. The data science in biomedicine track is a focus area within the computer science MS degree where students take three elective courses (total of nine credits) from a list of courses related to biomedicine, of which three credit hours will count toward their MS electives and the other six credit hours are required additional credit hours for the track; hence, a total of 36 credit hours is required for graduation from the data science in biomedicine track. The track graduates will write a thesis (or publishable paper) with a focus on data science in biomedicine.

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.

We believe a challenge of interdisciplinary education is finding the right balance of “breadth versus depth.” That is, depth in any one topic may be sacrificed by covering multiple disciplines. On the other hand, the interdisciplinary nature of the program may be compromised if the curriculum is focused too much on one topic. We designed our programs to be parallel tracks within existing degrees so students have depth in one focus area (biostatistics or computer science), and then diversify their skills and expertise in complementary areas with additional classes and thesis research.

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 developed biomedical-related data science tracks within the existing biostatistics MS and computer science MS degree programs in response to the changing landscape of biomedical research and technology, which is relying more on the generation, archiving, querying, analysis, and interpretation of large data sets. By implementing a new track, students pursuing this training within the respective MS program will have an official designation within their degree. This designation will help with employment and other opportunities such as internships and fellowships, where it would benefit students to show documented training and experience in data science. These tracks were developed in parallel to promote synergy between the departments of biostatistics and informatics and computer science and engineering.

How do you view the relationship between statistics and data science/analytics?

This relationship has been discussed in great detail on many blogs, editorials, and websites. Without getting into those details, we adapted a popular quote that is a pithy description of data scientists: “It’s the person who is better at statistics than any computer scientist and better at computer science than any statistician” (adapted quote from Josh Wills, director of data engineering at Slack.

What types of jobs are you preparing your graduates for?

This is the first year of our tracks, so we do not have job placement data yet. But the curricula are designed so graduates will be prepared to work in biomedical sectors (academic, public, and private) that require the analysis of large and diverse data sets from high-throughput omics, imaging, biomedical sensors, and electronic health records.

What advice do you have for students considering a data science/analytics degree?

We believe it is important to find a program that stresses interdisciplinary education, where students gain knowledge and skills in multiple domains from experts in the respective areas. However, a challenge of interdisciplinary education is finding that right balance of “breadth versus depth,” as discussed previously. When researching programs, students should ask faculty about their philosophy for striking that balance. This would give more information about how that interdisciplinary degree is different than pursuing related degrees such as computer science or (bio)statistics.

Describe the employer demand for your graduates/students.

In Colorado and nationwide, data science skills within the biomedical field are in high demand in industry (e.g., biotechnology and pharmaceuticals), the public sector (e.g., NIH), and academia.

Do you have any advice for institutions considering the establishment of such a degree?

Even if there is mutual interest among different departments, creating an interdisciplinary program has practical challenges. In our case, the two departments are on different campuses (~ 8 miles apart), have different tuition structures, and follow different academic calendars. Despite these differences—with support from department and school leadership and good communication between the programs—we were able to overcome these practical challenges.

 

As the ASA reached out to the statistical community for their involvement in starting master’s programs in data science and analytics, we learned about a few doctoral programs we wanted you to know about. We’re grateful to these universities for telling us about their innovative programs.

 

University of Wisconsin-Madison

Karl Broman, a professor in the department of biostatistics and medical informatics at the University of Wisconsin-Madison, is an applied statistician focusing on the genetic dissection of complex diseases in model organisms. He was named an ASA Fellow in 2016.

 

PhD Data Science Analytics, Biomedical Data Science PhD

Year in which first students graduated/expected to graduate (or year of first class for PhD program): 2018

Number of students enrolled: Expect 6–10 per year

Partnering departments: Biostatistics and Medical Informatics

Program format: In-person; 51 credits; thesis; traditional full time

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.

Our biomedical data science PhD program integrates the two related disciplines of biostatistics and biomedical informatics. Students select three year-long sequences from a set of core topics, including at least one biostatistics sequence (for example, biostatistics methods) and at least one computer science/informatics sequence (for example, artificial intelligence/machine learning). Students will also complete at least six credits of biology coursework (such as genetics and genomics) and three semester-long research rotations concerning substantive problems in biomedical data science and advised by a computational faculty member in collaboration with a faculty member from the biological, biomedical, or population health sciences. Cohort cohesion is developed through an in-depth second-year reading course—in which a selection of seminal articles in (bio)statistics, computer sciences, and biology is deconstructed—and a third-year professional skills course. Prerequisites include courses in calculus, linear algebra, programming, and data structures, but students will be admitted with a portion of these and can complete the remainder during their first year in the program.

What was your primary motivation(s) for developing a doctoral data science/analytics program? What’s been the reaction from students so far?

The primary motivation for our program was to synthesize training in biostatistics and biomedical informatics (which our department has done at the faculty level for 20 years) to provide students with broad computational skills, along with knowledge in an area of biomedical science, so they can make sense of the complex, high-dimensional data that is now the norm in biology, biomedical research, and public health policy. We are recruiting our first class of PhD students, who will enroll in fall 2018.

How do you view the relationship between statistics and data science/analytics?

Statistics, as a field, is sometimes viewed as being restricted to the use of probability theory for formal inductive inference and the quantification of uncertainty in such inference. But applied statisticians’ work has always included a broad set of activities, including data management, data cleaning, data visualization, statistical computing, software development, and communication about data. In some ways, the term data science does a better job of capturing these activities, including other techniques such as machine learning, that have been developed by computer scientists. I would like statistics to be viewed more broadly, to include the full spectrum of problems that one must confront when seeking to make sense of data. However, it is important to both recognize and take advantage of the many important data science ideas that have arisen outside of statistics.

What types of jobs are you preparing your graduates for?

Students graduating from our program will be prepared for academic positions and positions in industry and government. Our program includes a professional skills seminar through which students will explore and prepare for the range of employment opportunities so they can make informed career choices and be ready to carry out a successful job search.

What advice do you have for students considering a data science/analytics degree?

Find some data—perhaps a friend’s data—and dig in! Focus first on data visualization, but always have specific questions you’re seeking to address. It’s easier to acquire and develop your computing and data analysis skills if you have specific challenges in mind.

Describe the employer demand for your graduates/students.

It’s clear there’s great demand for skilled data scientists. It also seems clear that employers focus not so much on skills in statistical theory as on communication, ability to work in an interdisciplinary team, and more general problem-solving abilities.

 

University of Central Florida

Liqiang Ni is an associate professor of statistics in the department of statistics at the University of Central Florida. His main research interests include dimension reduction, multivariate analysis, actuarial science, and business intelligence. He has served as the graduate coordinator since August 2017.

 

Shunpu Zhang is a professor of statistics and chair of the department of statistics at the University of Central Florida. Under his leadership, he and his colleagues created this new PhD program in big data analytics to spearhead UCF’s effort to meet the big data challenge in 2017.

 

PhD in Big Data Analytics

Year in which first students graduated/expected to graduate (or year of first class for PhD program): Fall 2018

Program format: In-person; 72 credit hours beyond a bachelor’s degree, with up to 30 hours transferable from a completed master’s program in statistics, computer sciences, or mathematics; dissertation; full-time students; graduate teaching/research assistantship available

Please describe the basic elements of your data science/analytics curriculum and how the curriculum was developed.

We require the incoming students for our PhD program in big data analytics to have a bachelor’s degree in statistics, mathematics, and computer science with experience in at least one programming language. The curriculum includes advanced statistics, big data architecture such as distributed storage and processing, data mining, machine learning, and other recent developments in big data analytics. These courses will be taught (or jointly taught) by statisticians, computer scientists, and experts from the industry. We expect our students to possess the following core skills: data management; algorithm development; programming in R and Python; statistical inference; and communication and data visualization.

Traditional PhD programs in statistics aim to train students to analyze small to medium size, structured data. Our curriculum will focus on big data analytics. It is to train researchers with a statistics background to analyze massive structured or unstructured data to uncover hidden patterns; interesting, actionable associations; and other useful information for better decision-making. In addition to statistical inference and software, the new program has an interdisciplinary component that combines the strength of statistics and computer sciences.

What was your primary motivation(s) for developing a doctoral data science/analytics program? What’s been the reaction from students so far?

Started in 2001, the department’s data mining program (MS level) is the oldest such program in the United States. In recent years, we have seen a growing need for an educated and talented workforce of data scientists above the MS level who can contribute to industry, government, and academia through innovative applications of data analysis methodologies. The PhD program we are developing pursues high levels of community and business engagement. Our PhD Industry Advisory Board consists of member-participants from CFE Federal Credit Union, Sodexo, UCF Institute for Simulation and Training, the Walt Disney Company, Johnson & Johnson, CitiGroup Inc., iCube Consultancy Services Inc., UniKey Technologies, SAS, and Health First. The board continuously provides feedback on industry-driven competencies.

How do you view the relationship between statistics and data science/analytics?

We believe statistical science is an integral part of data science/analytics. A good data scientist/data analyst must be a good statistician, regardless of the label or title.

What types of jobs are you preparing your graduates for?

We are preparing graduates for academic/research institutions, industry, and government.

What advice do you have for students considering a data science/analytics degree?
Be open-minded and always ready to learn something new. You do not have to choose between hiking the mountain (statistics) and swimming the ocean (computer science, etc.). You can do both.

Describe the employer demand for your graduates/students.

Data scientists are in short supply, and the compensation for data scientists is very high. A McKinsey study predicts that, by 2018, the number of data science jobs in the United States alone will exceed 490,000, but there will be fewer than 200,000 available data scientists to fill these positions. Data scientists can earn a base pay of $116,840 (Glassdoor, 2016) and an average base salary of $113,436 (Forbes, 2016).

Do you have any advice for institutions considering the establishment of such a degree?

A big data analyst needs to have the ability to use existing methods or develop new methods to uncover true information from enormous amounts of data. To do so, a program in data science needs to provide students rigorous training in big data structure, programming skills, statistical methodologies, algorithm development, and interpreting and communicating the information discovered in the data. We believe such a goal can only be achieved by developing a cross-department joint program, instead of an “add-on” program that only requires students to take a few courses from other disciplines.