What does a Ph.D. in Data Science Look Like? And Do I Need One?

Data Science – a new frontier

In the evolution of data science, most people will reference two major events that fundamentally contributed to the establishment of the discipline. The first event was the release of the McKinsey Global Institute’s 2011 report on Big Data as the next frontier for innovation, competition and productivity – and the associated gap between the supply and the demand for deep analytical talent. At the time, the McKinsey report forecasted the “talent gap” – the number of positions requiring deep analytical talent above what the educational market was prepared to produce – as 140,000 to 190,000 positions. As it turns out, the talent gap is not only not closing, but it continues to widen. IBM reports that by 2020 the number of data science and analytics job listings is projected to continue to outpace supply by over 364,000, with over 2,720,000 positions needing deep analytical skills. The second event was Tom Davenport and D.J. Patil’s 2012 article referring to data science as the “sexiest job of the 21st century”. This article was every math and computer science geek’s high school revenge come true.

The academy responds

Academia, not typically known for its responsiveness to the needs of the market, actually did a pretty good job of developing programs in analytics and data science to address these issues. In 2006, North Carolina State University developed one of the nation’s first masters level programs in analytics. Today, there are over 200 masters level programs that are offered across the country. It is interesting to note that while most of these programs incorporate mathematics, statistics, programming and some kind of applied practicum, there is little consistency regarding where universities house their analytics and data science programs. Universities house these programs in colleges of business, science, engineering, mathematics and computing, resulting in slightly different strengths, weaknesses and specializations.

Can I get a Ph.D. in Data Science? Do I need to?

In 2015, Kennesaw State University developed the nation’s first Ph.D. program in Data Science. Today, there are dozens of Ph.D. level data science programs offered by universities across the country with hundreds of students pursuing the advanced degree. Like the masters programs, most of the Ph.D. programs integrate advanced course work in mathematics, statistics and computer science. However, unlike masters programs, Ph.D. students must also engage in research, peer-review publication and produce a dissertation. Additional information about the specifics of Ph.D. programs can be found here.

Most people who want to pursue a career in data science do not need a Ph.D. to be successful – although most do have a masters degree. So, what is driving the demand for Ph.D. programs in data science?

The first is, of course, academia. The talent gap identified by McKinsey and IBM exemplifies the need for not just deep analytical talent in the economy, but also for deep analytical talent in the classroom. The “shadow” talent gap in universities is arguably a bigger issue – universities are challenged to produce the talent needed to close the gap without having the faculty talent to teach the material. It is important to note that the 200+ masters programs offering programs in analytics and data science referenced above are currently doing so without people who have Ph.D.’s in data science. Most universities across the country have openings for faculty with Ph.D.’s in data science that they are struggling to fill. With the “oldest” Ph.D. programs in the country (those started in 2015) just starting to graduate their data science Ph.D.’s, the “shadow” talent gap in academia will likely not close any time soon.

The reason the talent gap in academia is not closing is partially because, most people who complete a Ph.D. in data science will go to work in the private sector. This is the first time in history that individuals earning Ph.D.’s are electing to NOT go into academia. While the statistics vary greatly, all indicate that over 50% of individuals who complete a Ph.D. in a STEM discipline will never teach. This is true for at least two reasons. The first is simple economics – the demand for individuals with deep analytical skills continues to elevate salaries – with most organizations paying salaries substantively above what most universities can offer. According to the 2018 Burtchworks Study on Data Science, 48% of all data scientists hold a Ph.D. and earn a median salary of $130,000 – which is about $10,000 more than their counterparts with only a masters degree. The second reason is that heavily data-centric organizations place value on cutting edge research and innovation in data science, enabling individuals who want to engage in research and scholarship, an opportunity to work outside of a university. A recent study found that over 25% of all peer reviewed articles in “academic” journals related to data science, were actually authored by individuals with Ph.D.’s but no academic affiliation.

So, while individuals interested in pursuing a career as a data scientist can be successful with a masters degree, a Ph.D. may be worth considering – even if they have no intention of teaching.

Jennifer Lewis Priestley, Ph.D. is the Associate Dean of The Graduate College at Kennesaw State University. She is director of the Analytics and Data Science Institute and launched one of the first Ph.D. programs in Data Science in the country.