The Power of Color in Building Data Visualizations

Screen Shot 2014-04-22 at 1.04.14 PM  






Color may seem an activity reserved for comparing curtain samples or choosing between pairs of shoes. In terms of data, science and research, you may think that color hardly factors into the equation. If you want to be professional you use black and white. If you want to be "fun" you use bright colors. Right?

WRONG! Science is no longer as cold and corporate as it once was. An effective visualization is driven by design principles and is created by way of a process in which we make careful aesthetic aesthetic choices that will optimize the cognitive processing of the data.

Colors elicit powerful psychological responses, and so we must choose them thoughtfully and deliberately. Certain colors help us concentrate, others evoke certain emotions, and yet others lead us to look at certain pieces of information and not others. Here are a few principles to keep in mind when designing a visualization:

Be neutral by default The color palette you choose should, above all else, be gentle and easy on the eyes. If you want to display multiple pieces of information, it’s useful to use color variation to distinguish them from each other. However, too many colors side by side looks like a total mess. It’s best to start with a base of soft, standard colors that are easily distinguishable from each other yet don’t demand your visual attention or detract from the information displayed.

Screen Shot 2014-04-22 at 2.03.00 PM

Use for emphasis Brightness, shade, and hue can be manipulated to differentiate and highlight certain pieces of information. “Hot” colors like red and yellow demand our attention whereas cooler colors like green and blue stay in the background and set a neutral base. Similarly, brighter shades  stand out amongst a palette of softer, weaker colors. Keep in mind that if everything stands out, nothing stands out, so use moderation when emphasizing information with color.

Screen Shot 2014-04-22 at 1.04.14 PM

Organize Information We tend to group like colors in our minds and thus we typically perceive similar colors as related. Keep this in mind if you’re trying to group together metrics or dimension values. It's an interesting idea to color-code related pieces of information across a visualization. The eye will logically view the similar colors together and mentally compartmentalize them as a group which increases simplicity and readability.

Screen Shot 2014-04-22 at 2.03.07 PM

Keep the variation between colors distinguishable yet subtle in order to maintain the cohesiveness of the tile pattern or the visualization as a whole.

and finally the Golden Rule… Moderation is Key.