There are an abundance of articles, on numerous platforms, that explain why "You don’t need a master’s degree to get into Data Science!" This is often followed by some kind of a review of the resources available, perhaps the author’s story of how they did it, and lots of encouragement to give it a go. Sometimes, there are not-so-subtle jabs at universities. This can feel very liberating to a future data scientist!
Let’s explore this topic as data scientists.
If "No! You don’t need a degree!" is the answer, what is the question? The question here is as follows: "Is it possible for someone to get a job in the field of Data Science without a master’s degree by instead following a regimen of self-study and networking?"
This is a great example of one of the first problems we see among failed Data Science projects: coming up with a thorough and detailed answer to the wrong question. I talk about this regularly in my classes and have experienced it myself in my professional career.
The question being answered is one of existence. I’ll restate it: Does there exist, in the entire field of Data Science, someone who has a job as a Data Scientist and did not get a master’s degree? Hopefully now you see the issue. Of course there is. There is probably someone who did not even get an undergraduate degree. There is probably someone who never took a formal programming course, in person or online.
Instead, let’s think about the real question someone is asking when they click an article title like that. It is probably more like this:
What is the probability that someone with my background, temperament, and self-discipline can build a successful career as a Data Scientist through self-study, online certificates, and professional networking ?
All these articles can say, assuming your background, temperament, and self-discipline are the same as the author, is "The probability is not zero." That probably feels less liberating.
And really, there is probably an even more precise question that a hopeful data scientist is asking:
Given my current employment status, academic background, personal family situation, temperament and self-discipline, what path should I choose if I want to have a successful career in data science?
This is a much, much harder question to answer, but it is much close to the real question. There are numerous resources out there, and I’ve written a couple my self. Here is one:
Should you get a Master’s Degree in Data Science? A faculty perspective.
It is true, however, that a Master’s Degree is not for everyone. Here is my take on thinking through if it isn’t for you:
As you are reading, though, consider this statistic from Gartner Research that I found recently: 71% of current Data Scientists hold a master’s degree or Ph.D. (Trends in Data Science and Machine Learning Talent by Peter Krensky, 16 November 2020, ID G00730094).
This data point also cannot answer your questions, though. It does not say, for example, that you only have a 29% chance of becoming a data scientist if you do not have an advanced degree. We know that many of the early data scientists were advanced degree holders from an array of quantitative fields. This does not mean the same will hold for new entrants into the field, and given the demand for data scientists, I expect that 71% to drop in the coming years, perhaps substantially.
The good news is that you do not need a precise estimate of these probabilities. However, the 71% statistic should influence your choice of next step in your career path. It is fair to assume that your chances of becoming a Data Scientist are greater with a master’s than without one.
There are huge numbers of data science projects, within large, established firms, that never go anywhere because teams spent hundreds or even thousands of hours answering the wrong question. As you are embarking on your career, be sure you are thinking very precisely about what questions you are asking and answering, and be sure that the two align. Even slight changes of words matter a lot: "Is it possible…" is very different from "Is it probable…"
If you do decide to apply to graduate school, take a look at this, it will save both you (and me) some time and energy:
Good luck!
