Michael Rappa, a Distinguished Professor and Director of the Institute for Advanced Analytics at North Carolina State University in Raleigh
” The five main qualities of a data scientist who wants to be employable are:
Technical Skills: “They have to have the mathematics, statistics and computer science skills in order to really work with data and be able to analyze it,” says Rappa. “You surely have to have a solid foundation in statistics and mathematics, but you don’t have to have PhD level knowledge.”
Teamwork Skills: Almost all of the data science roles at data-driven organizations are team-based. Importantly, these are cross-disciplinary teams. “You can go through 16-plus years of education and never really develop teamwork skills, so it’s something that we take very seriously,” Rappa says. “In fact, we’ve developed an entire curriculum around teamwork. We’ve had really good conversations with employers who will tell us if a student doesn’t have specific knowledge in a particular technical area, we can kind of fix that. We can send them to a program. But if we hire someone who can’t work in a team, we can’t really fix that. It might never be fixed.”
Communication Skills: This isn’t just about giving a presentation. This is about briding the gap between hard analytics and an MBA’s view of the business. It’s about “applying complex areas of knowledge and deriving insights from data, as part of a process of making decisions about what to do with that data,” Rappa says. That can mean, in order to communicate the value of a data insight, a data scientist might have to show, rather than tell. She might have to campaign for, and spend some money to build out a product, so that someone who has little insight into the data or technical knowledge, but makes decisions for the business, can see why it’s useful and learn to trust the data scientist. “The communication process has really got an added layer of burden to it,” Rappa says. “The data scientists have to really be able to communicate in a very credible way with decision makers so that they…can trust what they’re [being told].”
Business Skills (with a little bit of empathy): The ideal data scientist candidate will be able to analyze data in a business context, and empathize with the specific process and values of the business. “The notion that you take a statistician, put them in a cubicle and dump data on them and have logistic regressions popping out isn’t really useful for most businesses,” Rappa says. “They need individuals who can bring an understanding of the business to the analysis, and by doing so bring value to that work that they’re doing. They know that we’re a retailer, or we’re a bank, or we’re a social network…and can bring value to the data analysis, and not just let it be passive data.” Rappa says his program focuses heavily on this issue, where conventional statistics programs are fundamentally focused on data itself.
Tool Mastery: There are a great many analytical tools on the market now, and Rappa makes sure his program is replete with training around the latest offerings.”