There’s no denying the exponential growth of data science as both a field of study and career choice. In 2012, the Harvard Business Review went as far as to call the role of data scientist “the sexiest job of the 21st century.”

While there’s been much said and written about the actual staying power of data science, it’s difficult to deny the current buzz. At the time of this writing, there are roughly 30,000 job postings on LinkedIn and about 22,000 postings on Indeed for “data scientist.”

However, the explosion of data scientist jobs in the private sector is a double-edged sword. While it has increased interest in the field, it has also harmed the academic community around data science. It is this reason that Jennifer Priestley, a professor of statistics and data science at Kennesaw State University, argues that it’s time to get serious about a Ph.D. in data science.

SEE: Quick glossary: Big data (Tech Pro Research)

Data science really began emerging at the master’s degree level, Priestley said, as a response to the demands of the private sector marketplace. Master’s degree programs, while more focused than a bachelor’s degree, tend to be relatively short.

Now, Priestley said, we are beginning to see an emergence of undergrad programs in data science. However, from what she’s seen, it’s not usually a bachelor of science in data science, it’s usually a minor, a concentration, or track. So, the degree may be in computer science with a minor or focus in data science, for example. But, that could be a good thing.

“My opinion is, at the undergraduate level, it doesn’t make sense to major in data science because you have to have some domain knowledge,” Priestley said.

In other words, as a data scientist you can’t just translate data into information, you have to be able to know what it means.

The talent gap for data science is well-documented, and that creates a lot of hype around the field. Right now, data science is a buzzword, and Priestley isn’t afraid to call it that. But all that buzz can have a negative effect on the field itself.

“They take one course in–name it–data mining or ‘data science’ in their MBA program and all of a sudden on their LinkedIn profile they’re a data scientist,” Priestley said. “I think that does the discipline a huge disservice.”

This is where the Ph.D. comes in. To be clear, Priestley isn’t arguing for the first Ph.D. program to come about–there are already a few in existence. Kennesaw State University, New York University, the University of Maryland, Georgia State, and others all have Ph.D. programs in data science. However, an argument is being made for their importance and for more, similar programs to arise.

The growth of Ph.D.-level study in data science could help solve two core problems. First, it could continue to address the talent gap in the private sector by training people in a deep and meaningful way to be able to solve difficult, complex problems, likely on the R&D side of things. This would feed the pipeline for higher-level talent.

Secondly, the rise of the Ph.D. could also help with what Priestley called the “shadow gap of talent in academia.” In essence, she said, people from the private sector are coming to the academic community asking for more data scientists, but there aren’t enough Ph.D.s in data science working in universities to teach and train these individuals.

“We have this misalignment between what the private sector is asking for from academia and what we’re ultimately able to provide, which ultimately is fueling the gap,” Priestley said.

By producing more people with Ph.D.s in data science, those graduates could then go back into universities to teach next generation of data scientists. In starting to build out a program, universities have to understand the nature of data science.

“Data science, in many ways, is like tofu,” Priestley said. “What I mean by that is, if you embed data science in a business school, then it’s going to take on all the flavors and the culture of the business school education. And, that means something.”

It’s the same thing if you embed it in the college of computer science or mathematics. None of these approaches are wrong, Priestley said, and it’s actually good for the discipline at this stage, because it provides “labs” for data sciences study within those other fields.

However, Priestley said she believes an interdisciplinary approach is best. But, we don’t need standardization, because data science is still so much in its infancy that making all the programs the same could harm potential innovation.

“If we try to standardize it too quickly, then we’re going to end up suppressing a lot of the organic findings and organic development that’s happening because so many universities are taking different approaches to the discipline,” Priestley said.

Right now, the landscape is very diverse. Kennesaw State, for example, is interdisciplinary, but it relies on statistics. Georgia State, though, puts data science in the business school. There’s so much need for this talent, and she believes all these different approaches are helping close the talent gap.

After building the program, the universities will have to offer compensation that is competitive in both the academic realm and the private sector if they want these individuals to come back as professors.

“You can’t expect somebody that has a Ph.D. at this intersection of skills to walk into a math department and be happy making $50,000 a year,” Priestley said. “That’s not going to happen, that’s unrealistic.”

SEE: Can IT keep up with big data? (TechRepublic)

A Ph.D. program will help advance the data science field, but not in the ways that people may think. It’s not so much about new techniques, it’s more about looking at data differently–understanding that all content is data and how it can be broken down and harnessed in a predictive model.

If you’re considering a career in data science, of course you need to be enumerate, with basic skills in mathematics and programming, but it’s more complex than that.

“The best data scientists are coming out of psychology, and music, and history, because they’re creative and they’re not necessarily restricted to thinking in a very linear fashion,” Priestley said.

The 3 big takeaways for TechRepublic readers

  1. The massive growth of data science in the private sector has increased interest in the field, but it has led to a huge talent gap in both academia and business.
  2. More Ph.D. programs in data science could help solve the talent gap in the private sector with high-talent candidates, but it could also solve the talent gap in academia with more data science professors to teach.
  3. Ph.D. programs will also help the field of data science by encouraging more people to look at data differently and see how it can be leveraged in new ways.