Does a Data Scientist need a PhD?
I see many job specifications in my line of work, many of them give a pretty good insight into the type of person a company is after, and with added info over the phone, they can help to bring a job to life. Sometimes, however, it’s like somebody has chewed up and swallowed the Big Data page on Wikipedia and spit it out onto a piece of paper. Written by those who don’t understand what the role is and what they will actually be doing, taking every single buzzword and phrase which relates to Big Data and putting it on the page as a necessary requirement.
It is these types of job specifications that can maybe seem off-putting for candidates, they don’t make sense and seem unrealistic. However, one requirement over all the others seems to be the most popular topic of discussion; education. Many job specs will require a Data Scientist to have a PhD as a minimum requirement, but do they actually need one? Some roles may be much more R&D focused and therefore will require somebody with a research background and skillset, typically exemplified by doing a PhD. However this can’t be the case for all Data Scientist roles, can it?
What Does the Expert Say?
To get an industry viewpoint on this I asked Billy Wong, Lead Data Scientist at TES Global, for his thoughts on the matter:
Let’s first define what we mean by Data Science. For me it is the overlap between domain knowledge (so that you can ask the right questions and interpret the answers), IT skills (so that you can find, collate and transform the data) and analytical skills (so that you can turn the data into answers).
Having an advanced degree is certainly useful. The experience in conducting research prepares the data scientists to structure questions, critically examine assumptions and use various analytical and statistical techniques. However, PhDs don’t have a monopoly on these skills, and these skills alone don’t make a good data scientist.
For me, a job specification should be a clear way of identifying what this person will actually be doing and therefore what skill set is needed to do this well. Education plays a big part, but maybe it’s not everything. What are your thoughts?
Big thank you to Billy Wong of TES Global for collaboration on this piece. Andy Driver leads the Big Data Analytics function at RedCat Digital and is recruiting across the UK, Europe and the USA, get in touch on firstname.lastname@example.org to find out more.