Jeff Holt is the Program Director for the MS in Data Science program at the University of Virginia. He previously was Chair of the Department of Statistics, and has a joint appointment in the Department of Mathematics. His statistical research interests focus on sampling methods applied to ecological settings, and he recently published a linear algebra textbook. He teaches courses for MS in Data Science students on statistical computing and linear models.
We reached out to Dr. Holt to get his insights on what it takes to succeed in data science, the difference between data science and related disciplines, and the key skills data science students need to master.
There isn’t really a clear definition of the boundary between data science and analytics, but I think that data science programs tend to be somewhat heavier on computation. There are employment opportunities for students coming out of both types of programs, so I don’t think that students should choose based solely on the possibility of employment (though I understand that employment is a completely legitimate and important concern). Instead, I think students should try to select a program that best matches their interests and strengths. A data science program may well benefit a student who is interested in a greater blend of computation and analysis.
In conversations with employers, they consistently mention the importance of effective communication skills. There is a tendency for students to think that the technical side of data science is all that matters, but that is definitely not true. Beginning the process of developing effective communication skills requires that we recognize their importance.
Communicating technical results effectively is challenging. It can be hard to strike the right balance between technical accuracy and being able to reach a nontechnical audience. It takes practice, so the more time students spend interacting with other students and instructors, the better. In addition, learning about effective uses of the graphical displays of data (which is often underemphasized) can greatly enhance technical communication, especially for a nontechnical audience.
On one hand, terrific work has been done that has made it possible for some types of data work to be automated. I expect that advancements in this direction will continue, displacing some human workers. It’s hard to say how fast this will happen, and to what extent, but I can’t imagine it not occurring.
On the other hand, there is huge growth in the types of data being collected, and in the types of questions being considered related to data. As things stand currently, humans are required to recognize that a type of data is worth collecting, to determine the types of questions that the data might help in answering, and to develop methods for extracting the needed information from the data. These are all complicated tasks that aren’t suited to automation that is currently available.
In the near term, I think that the creation of jobs due to growth in data work will greatly outpace the elimination of jobs due to automation, and I’m not sure that automation will ever eliminate all jobs. I don’t precisely know what I mean by “near term” but it seems like we’re much closer to the start of the data boom than we are to the end, so I’m not worried about the possibility of mass unemployment of data scientists happening anytime soon. However, I’m not an expert on AI — maybe Skynet will take over tomorrow and put everyone out of work.
Learn as much as you can before you start applying to programs. There is a lot of information available, but data science is not well-defined so it is challenging to sort out the field. Once you’ve decided to take the plunge into a data science program, be ready to keep learning. (That may actually be two pieces of advice.)
Regardless of one’s definition of data science, one thing everyone seems to agree on is that data science evolves at a fast pace, and I don’t see any reason for that to change. This makes data science a terrific field for people who are inquisitive, like to continually expand their horizons, and aren’t intimidated by the prospect of learning something new.