How to Choose a Big Data Master’s Program

In its 2017 report Investing in America’s Data Science and Analytics Talent: The Case For Action, PricewaterhouseCoopers predicted 2.72 million Big Data job postings in the year 2020. The same report foresaw “a significant shortfall in the number of data scientists and ‘data-enabled’ professionals” to take those jobs. Average advertised salaries for these positions range from just over $69,000 (functional analysts; 770,441 job postings) to over $91,000 (data-driven decision makers; 812,099 job postings). If PwC is to be believed—and the consulting firm has a pretty reliable track record—Big Data stands to be one of the major growth areas in business employment for years to come.

Data jobs require training, which come primarily through academia (there may be a few intrepid savants out there willing to learn stochastic calculus on the job, but probably not so many employers willing to let them). Universities have established a number of disciplines to address the need: the primary fields are business analytics, business intelligence, operations research, data science, and statistics. All of these disciplines employ practices and principles of computer science, mathematics, engineering, and business in some combination; in some instances, the name of the discipline has more to do with which academic department offers the degree than with actual differences among the disciplines.

What should you look for in a quality Big Data program?

Once you decide on a Big Data discipline and a degree, you’ll want to learn more about various programs’ resources, faculty, curricular balance, and reputation. Finally, you’ll want to calculate your return on investment.

Resources
In order to study Big Data effectively, you need access to: huge, rich databases to analyze; the apps businesses typically use to sort through and make sense of all the data; and, computers powerful enough to handle the workload placed on them by the databases and apps. If you’re considering a full-time program, make sure the program you choose is outfitted with the resources you’ll need to learn. The quality and cost of the resources provisioned to a program also provide a telling indication of how strongly the university supports it.

If you’re currently employed, consider giving preference to programs that utilize the same apps you use at work. It’s not absolutely necessary—other strengths of the program you ultimately choose may be more important in your decision—but it certainly would be a nice perk. Also look to see whether the university has an institute or center dedicated to Big Data studies. If it does, that means forward-looking research is among the school’s priorities, something to consider, especially if it is among your top priorities as well.

If you’re planning to study online, your access to computing power will obviously be limited to what is available at home or at work. Still, you want to make sure that your program will expose you to the most popular analytics apps. You’ll also want to find out how you’ll access these apps, what technical requirements your setup will have to meet, and how you’ll be delivering assignments that require work with big databases and powerful apps.

Online programs can handle all of these challenges but it does complicate things, and you’ll want to make sure the program you choose has effective solutions. Sometimes this information will appear on the program website, but if not, most online programs will assign you an admissions counselor/recruiter who will answer your questions.

Faculty
Besides the resources mentioned above, you’ll also need someone to teach you how to use them. Unless you’re in a PhD program, it’s not that important that your teachers be superstars in the field. Not that there’s anything wrong with being taught by thought leaders, but unless your aim is to be on the cutting edge of theory and research, what’s more important is finding teachers who are committed to and capable of teaching.

Once you’ve narrowed your choice of programs down to a few, find the names of key faculty members online and start googling. Search for their names, restricting your search to sites like reddit.com and quora.com. These threaded-discussion sites may contain some conversations about specific faculty members’ teaching skills. Sites like ratemyprofessors.com sound promising but in fact are mostly used by undergraduates and are therefore less helpful to potential graduate students.

Curricular balance
Big Data programs typically include instruction in advanced mathematics, statistics, computer science, engineering, and business, and most are team-taught by faculty from several (if not all) of these disciplines. How these disciplines are weighted in the curriculum—e.g. whether the curriculum emphasizes computer science over mathematics; whether it includes no, some, or a lot of business instruction—can vary depending on several factors. One is the type of degree: business analytics degrees tend to include more business courses, while business intelligence, operations research, data science, and statistics degrees tend to lean more heavily on advanced mathematics and computer science. Another is the school’s strength(s): if the school is a computer science powerhouse, for example, computer science is likely to feature more prominently in the curriculum.

Academic programs list their complete curricula—required courses and electives—on their websites, nearly always with useful descriptions of each course. Review the curriculum of each program you’re considering for its balance and also to see whether the course descriptions appeal to you. Course descriptions don’t always reflect course content with 100 percent accuracy but they’re usually close to the mark and they’re certainly better than nothing at all. Course descriptions will also give you a good sense of the program’s academic approach (passive vs. hands-on learning, theory vs. practical application, etc.).

Reputation
Most students pursue advanced degrees in order to improve their career prospects. A degree from a highly regarded university—say, California Institute of Technology—is guaranteed to impress employers more than a degree from a school they’ve never heard of. That’s one reason why students are willing to spend so much more to attend CalTech.

The list of schools with national reputations is relatively short. Much longer is the list of regional/local universities that are respected in their hometown, home state, or geographic region. For example, a Big Data degree from Stony Brook University will open doors for you in the Northeast, but maybe not in the Southwest. You should be familiar with the names of the schools in your region that employers love.

Speaking of programs and employers, most program websites have a careers section that lists employers of recent graduates. These lists often tell you who recruits program graduates most aggressively, and can give you some idea of the types of job offers a degree from that institution will bring your way. Have a close look before you commit to a program.

Return on investment
Most Big Data graduate degrees will improve your earning potential in the near and long term. Some will do that better than others, but of course may cost significantly more as well. For example, an online Big Data MBA from North Carolina State University will cost you a North Carolina resident a little over $51,000 in tuition; that same student will pay over $125,000 in tuition for an online MBA from University of North Carolina-Chapel Hill 30 miles away. Both are well-respected schools in their region and beyond. Is there enough of an advantage in attending UNC over NC State to justify the cost differential? Is there any advantage at all?

The problem in answering that question is (paradoxically given the subject of this article) the lack of reliable data. Most available data about the salaries of program graduates are reported by students themselves to the schools, which in turn use the results to promote their programs. Even assuming the data are accurate, they are at best incomplete (because they represent only those students who report their income). Such data can be helpful but isn’t determinative.

In the end you’ll have to decide on your own whether the ROI justifies the expense. Forbes’ Laura Shin offers a seven-step method for calculating ROI on a graduate degree that you might find helpful. The seventh step is “intangibles,” which brings us back to our UNC-NC State example: a true Tar Heel fan might conceivably pay the extra $75,000 rather than root for the Wolfpack.

Questions or feedback? Email editor@noodle.com