From Liberal Arts to Data Science: What to Expect on Your Journey

When I began my journey as a liberal arts major, data science wasn’t on my radar. As a political science student, I was much more interested in the qualitative side of things: the complexities of governance, the dynamics of power, and the narratives that shape societies. While I enjoyed these intellectual explorations, the job market was highly competitive, and I struggled to stand out in a sea of candidates. In the background, I developed a growing respect for the power of data—its ability to uncover hidden patterns, challenge preconceived notions, and offer insights that qualitative analysis often lacks.

This set me on a path I hadn’t anticipated: transitioning into data science. Today, I am a data scientist at Elder Research, and although my journey hasn’t been straightforward, it’s been rewarding from the very beginning.

To those in a similar position contemplating a career change, a liberal arts degree doesn’t have to be a barrier. Anyone can become a data scientist, and the learning curve is closer to two years than ten.

The key skills aren’t necessarily technical prowess from the start but rather diligence, curiosity, and an open mind. The field is expanding rapidly, and the entry requirements aren’t as daunting as you might think.

However, it’s important not to underestimate the time and emotional energy required to upskill. No one will do the learning for you. Drawing from my own experience, here’s a practical guide on what it takes, what to expect, and how to land a job as a data scientist coming from the humanities.

If you’re contemplating a shift from a liberal arts background to a more technical field like data science, recognize that this is a significant transition, not just in skills but in mindset. If your motivation is solely financial, you may find the journey particularly challenging. However, this often isn’t the driving force for most liberal arts students, who are typically drawn to fields like literature, history, philosophy, and the arts out of a deep-seated passion and curiosity.

My advice is to find a way to connect this passion with data science. For example, if you’re fascinated by literature or philosophy, you might explore how neural networks and natural language processing models can analyze and generate text, revealing new insights about language, meaning, and human communication.

Sociologists and historians can delve into areas of data science that uncover patterns in social behavior or historical trends, using data to enhance our understanding of societies and cultures. Those with a background in the visual arts can apply their creativity to data visualization, making complex datasets accessible and engaging, or even explore machine learning techniques in art generation. This approach not only makes the learning process more enjoyable but also helps you maintain your integrity, avoiding the temptation to cut corners along the way.

Start by taking a few programming courses in Python and SQL. Tools like CoPilot and ChatGPT are excellent for helping intermediate coders advance to the next level, but they will not take you from zero to one. Imagine a software engineer claiming expertise in your field by simply relying on the output of large language models. Learning how to program involves more than just getting a computer to perform tasks; it’s about understanding what a computer is and how it operates.

On the job, you will need to be able to understand your colleagues’ code and debug your own, and a foundational knowledge of programming is irreplaceable. There are many free resources designed for beginners in programming. Be ready to dedicate not just time but effort. If you’re not finding yourself struggling or contemplating giving up, you might not be pushing hard enough.

Think of it like learning a foreign language. It’s not just memorizing and mapping new vocabulary; there needs to be a more holistic understanding. And be wary of programs marketed with guaranteed quick results. In my experience these courses advance slowly enough to create an illusion of competence but too fast to impart actual understanding.

There will be areas of data science that will not interest you. For me, this was software engineering. After I got the hang of working with Python notebooks and CSV files, I found myself standing proudly atop the peak of “Mount Stupid” on the Dunning-Kruger Effect curve.

My models could predict fatality on the Titanic and categorize three species of iris flowers. But then came the reality check: navigating Python modules, packages, and environments; tackling object-oriented programming; working with containers and reproducible builds; and facing the thorny task of testing. These were the less glamorous but essential parts of the journey.

While areas like this aren’t as exciting, they’re crucial for becoming a well-rounded data scientist. It will be tempting to rush through and get back to the fun stuff like building machine learning models, but it’s crucial to persevere through these challenges. At least tread water.

The best way to make the most of your newly acquired knowledge is to put it into practice. Step away from YouTube tutorials and classes that come with built-in guardrails, and dive into working with raw datasets that genuinely interest you.

Make the experience enjoyable and relevant to your passions. If you’re into sports, build a prediction model for fantasy teams or explore open-ended research questions (e.g., Has defense actually declined in the NBA? What characteristics World Series winners have in common?). If finance excites you, explore time series analysis on stock data or replicate market-basket analysis techniques.

An elevated technical understanding will actually fill in conceptual gaps. That stock market model you are so proud of? I wouldn’t put too much money behind it. It will likely fail to perform well on new data. Maybe some patterns that seemed meaningful were actually just random chance or influenced by unrelated factors in the original data. But through this process, you’ll gain a deeper appreciation of what kinds of problems data science is well equipped to solve—like optimizing sales or detecting fraud—offering lessons that transcend the traditional approaches you’ll encounter in business school classes.

After a few months of dedicated effort, you will find yourself browsing job postings. It’s normal to feel paralyzed by imposter syndrome, especially after realizing just how much of the data science world you were unaware of just weeks ago. Resist the urge to retreat into more studying. Start applying for entry-level roles in data analysis, engineering, or data science.

At this point, you’ll learn more on the job than on your own. The worst that can happen is that you hit a technical question in an interview that exposes some gaps in your knowledge. But even then, you’ll gain valuable feedback on the skills you need to develop. Many companies keep your resume on file and may invite you back for future opportunities. It’s a no-risk situation, so seize it.

As far as interview preparation itself, rather than getting that extra hour of Python practice or memorizing facts on the newest trends in generative AI, my advice is to revisit introductory courses on data structures and algorithms. With your newfound experience, these concepts may click more easily and help break through any remaining mental blocks.

Alternatively, survey some project management courses to learn about Agile principles and the dynamics of working in cross-functional teams. Remember, no one expects you to be an expert at this stage—let your curiosity guide you. Take a methodical approach, noting things that capture your interest. Pay attention to the precision with which experts speak, and start connecting the dots between the technology, the theory, and your own journey.

My last piece of advice is to make sure that wherever you end up, your relationship with your company and colleagues is built on shared interests, not just a paycheck. I was fortunate to join Elder Research, a company that genuinely values its employees and cultivates a growth mindset. While many firms make similar claims, Elder stands out by truly investing in and trusting its people, even if it means sacrificing their bottom line.

Elder offers more than just on-the-job training—it fosters an environment that feels closer to a tight-knit academic circle than a traditional consulting firm. I’ve been encouraged to dedicate time to independent research so long as I eventually share my insights with my colleagues. And senior data scientists frequently lead collaborative study sessions, which promote continuous learning and team growth.

Accordingly, Elder’s hiring philosophy goes beyond conventional data science pathways, prioritizing positive attitudes and passion over hard technical skills, which can be developed through their platforms like Statistics.com. They took a chance on me and set me up for success.

Once you gain a foothold in the industry, I recommend seeking out the right culture and putting down professional roots in a place where you can learn, thrive, and make meaningful contributions.