Important Tips for Getting Into Data Science — If You’re Coming From a Non-Technical Background
Although there is still much to learn, you may already have a head start you weren’t aware of.
In the modern age of big data, data science is becoming an increasingly popular subject, consistently attracting people from all disciplines. By its very nature, data science is an interdisciplinary field, and so it benefits immensely from this multitude of perspectives.
There are still some foundational topics that everyone studying data science should learn, but the missing pieces of your knowledge base will differ depending on what background you come from.
While many hold the misconception that you must have a technical background in order to excel as a data scientist, this is far from the truth. Increasingly, data science requires the expertise of those with a qualitative skill set to aid technical-minded folks. That said, if this is you, there are still a few tricks you should pick up in order to contribute to a data science team.
Let’s get into it.
Learn to program
I know, I know — I just said you don’t have to come from a technical background, and now I’m telling you to learn programming. But hear me out.
At its core, data science is not based in computer science, but in statistics (which you should also probably get a little more comfortable with). That’s where the majority of the theory for things like data analysis, machine learning, etc. come from.
Programming, on the other hand, is a tool which makes it feasible to collect, store, and statistically analyze the ever-growing amount of data available to us in the modern age. Strictly speaking, you don’t need knowledge of super advanced programming topics like compilers and Turing Machines in order to be an effective data scientist. Is it useful for some data scientists (i.e. those who work more in the design and implementation of systems for data)? Certainly. But that doesn’t have to be you.
Rather, being able to write and interpret at least simple programs will aid you in two ways:
- When you’re exploring data in an introductory stage, you will be able to conduct simple processing and manipulation tasks to better understand the data. This puts you in a better place in terms of now discussing research questions or company goals with the team as a whole.
- If you’re assisting a team of statisticians and programmers as the resident domain expert, a basic understanding of programming will influence the how you contextualize the data and present possible approaches for analyzing it (if you know nothing about how programs work, you might give advice that is computationally infeasible).
As such, while you need not be an expert, you will definitely benefit from at least a foundational understanding of programming.
My recommendation: start with a basic Python course. It reads very similarly to English and doesn’t have many of the syntactical quirks and idiosyncrasies of some other programming languages. As a result, it’s perfect for people without prior programming experience, and it has the additional advantage of being one of the primary languages used in the realm of data science.
Learning it will serve you well in the long run.
Lean into your qualitative skill set
There are two primary reasons you can contribute immensely to data science if you come from a non-technical background.
For starters, while tech folks might be great at building systems and developing models, they often lack the people skills needed to actually push these artifacts out into the world. A huge part of data science is communicating insights in a clear way to the public, ideally in an effort to combat misinformation while earnestly educating.
This is where you — the non-technical superstar — come in. If you’re a talented writer or communicator (this doesn’t necessarily mean you give Ted Talks on the daily — it might just mean you’re a charismatic salesperson who loves interacting with people), then there’s a perfect niche for you in data science. Own your skill set, and market yourself as the bridge between programmers and customers.
Secondly, non-technical research skills are becoming increasingly more important to effective data science. I’ve said it before, and I’ll say it again: data science is about more than just the numbers. A blind fascination with quantitative data is a perfect recipe for unethical models and has dangerous societal implications.
If you are trained in social sciences research, modern data science doesn’t simply want you — it desperately needs you. You have the potential to polish data collection techniques and improve existing analysis algorithms in a way that better accommodates how humans actually operate in society. In the words of my PhD advisor,
“You shouldn’t write a machine learning algorithm and then retrospectively consider how it might be made ethical. From the very beginning of the data collection process, take into account the human aspect of the data, and build it into the core model itself.”
To do this, data science demands a deep understanding of qualitative research that you just might be able to offer.
Learn to appreciate STEM
In my previous article, I discussed how many tech people view the social sciences with a subtle disdain, feeling themselves to be superior on account of the popular attention their field receives. This viewpoint is forged early on in the hallways of university computer science departments, forwarded by ever-larger groups of students drowning in computers and code.
However — the reverse is just as true. It’s not uncommon for humanities students (and eventually, graduates) to adopt one or more of the following judgmental attitudes toward STEM folks:
- “They can’t write an essay to save their lives.”
- “Look at them, obsessed with code and working all day and night without any time for relaxation — what kind of life is that?”
- “So full of themselves and their high-end tech companies.”
To some extent, such comments are harmless, a precipitant of camaraderie among groups of like-minded students just messing around. However, just as with the judgments from the technical side presented in my last article, they cloak a hint of truth within them.
I don’t have anything against the occasional joke — but I do recommend you leave any actual stereotypes you hold about STEM majors back in those university halls. It won’t do you any good in the industry.
Data science is built on the combination of scientific and humanistic fields; thus, it requires the help of both to thrive.
Learn to work together.
Bonus tip: don’t underestimate yourself
I’ve hinted at this point briefly above, but I wanted to explicitly underscore it because of its immense importance.
Many non-technical people turn away from data science because they feel they aren’t “cut out” for a technical field. Get rid of this misconception: as I said above, data science is not purely technical.
You may run into people with the outdated idea that data science doesn’t need the help of anyone other than mathematicians and programmers. Don’t let them get to you — instead, show them just how much of a difference you can make.
In their obsessive and false superiority over you, they’re blind to the fact that they themselves are the ones getting left behind.
Final Thoughts and Recap
Here’s a quick round-up of the tips above:
- Learn to program — some knowledge of code will round out your skill set and help you contribute to data science teams more effectively.
- Lean into your qualitative skill set — data science needs those who can combat unethical practices and convey insights clearly.
- Learn to appreciate STEM — a sense of superiority will do you no good; we need to work together.
- Bonus: Don’t underestimate yourself.
As you make your foray into data science, I hope you find the above tips to be helpful. Best of luck, and until next time!