Entering a program so heavy on technical skills without a statistics degree is intimidating. It can feel like you are behind on day one and that your peers have significantly more experience in the world of data than you do. But data and statistical techniques have a way of sneaking into nearly every discipline, even where you wouldn’t expect them.
Approaching data science with an interdisciplinary lens has allowed me to maximize my abilities and find creative intersections between my skills and my ability to communicate complex data.
Growing up, I was definitely an art kid. I took art classes in undergrad and love to craft and work with my hands. On the surface, you probably wouldn’t expect that to contribute to being a successful data scientist, and you definitely wouldn’t expect it to help me with math. However, some lessons I’ve learned from my numerous crafty hobbies have helped me become more resilient, resourceful, and adaptable in my work at the IAA.
Printmaking is a tedious process. You learn to be adaptable when you make a mistake, because there are no do-overs.
Visualization
The low-hanging fruit when you consider the connection between art and data science is, of course, visualization. Having an eye for what is visually compelling and appealing, which can be developed by practicing and studying art in detail, allows you to communicate analytic findings in a way that captures audience attention and communicates the most important findings. Having a good understanding of color, form, and space allows you to make your visualizations accessible and interpretable.
Pattern and Noise
Finding and using patterns to your advantage is a key part of both artistic processes and data analysis. In fact, pattern (or repetition) is one of the basic principles of design. Knowing when to use pattern and when to be comfortable with random variation is one key to visually cohesive work. The human brain is constantly searching for patterns to make meaning out of what it sees. Sounds like the work of a data analyst, right? Finding signal in the noise is a skill that is honed over time, creatively and technically.
Pattern and random variation are as much a part of design as they are of model-building.
Storytelling
Perhaps more important than knowing your data is knowing how to tell a story with it. The same is true in art. Sure, you can have the most technically beautiful drawing in the world, but without context (both within the piece and in the broader world in which the piece was made), it’s lifeless. The choices you make in the composition make all the difference in how an audience interprets your work, even if the subject is exactly the same. The same is true for data. You must use data to tell a compelling story, using real-world context to inform how you communicate it.
Attention to Detail
Creating art and analyzing data also require significant attention to detail. Regardless of medium or style, the biggest differentiator between those who produce “good” art and those who produce “bad” art is in the details. The same is true in data analysis. Truly understanding your medium, its possibilities, and its constraints allows you to make decisions about your work that elevate your final product.
Attention to detail is often time-consuming and difficult, but can be very rewarding.
Iteration
Developing art, especially pieces that require sketching or painting, is an iterative process. You conduct your initial work, refine it, make decisions based on where you are and where you’re going, and repeat. This is reminiscent of the data lifecycle: clean, analyze, evaluate, repeat. In both, you must be patient and adaptive. Progress comes not from perfection in the first attempt, but from the willingness to refine and reimagine.
All of this is to say, maybe art and data aren’t so different. Finding patterns, telling a story, and sharing it with the world are ultimately the goals of both. And if you think about it, finding meaning in raw data is an art all of its own, right?
Columnist: Delaney Black