How-To: Create meaningful data in the era of information overload
From social media to news applications and online books, there has been a shift in the way that we as a species interact and think. With the world in our smartphones, there is an endless supply of data on pretty much everything. We have the ability to find the answers to our questions with a few simple searches…that is, until we use up our data plans (sorry mom!). But cellular data isn’t the only data we should be focused on.
Big data, big data, and more big data. As an intern for a Big Data & Analytics team, data is a recurring buzzword. I am currently working on a project where I extract data from a database and load it into a table. The question is, how do I transform this table into meaningful data that tells a story? My AP statistics teacher would be proud to hear the answer — graphs and charts.
Tableau Software is a company that provides users with a platform to create aesthetically pleasing and compelling data visualizations. They have a simple “show me” function that allows the user to click and drag data categories into the function which then automatically creates a graph based on the type of data that is being used. For example, if the user chooses to use the data for average sales versus profit, Tableau will most likely create a scatter plot with default settings.
The fact of the matter is, default settings are boring. Our minds work too quickly for boring. Data has to appear worthwhile in the first five seconds that we view it, otherwise our brains move on to something more compelling. Tableau actually refers to this phenomenon as the 5 second rule, for which they have a “5 Second Test” that all data visualizations should follow.
Here are some points I’ve gathered to create meaningful data in the era of information overload:
Leave the pie charts to kindergartners. Why? Because they’re evil, that’s why! But on a serious note, the underlying goal is to convey a message and tell a story. Charts that only show one piece of information, with no trends or correlations, have no value because conclusions can’t be drawn from a simple statement. Sure, pie charts can show that more handbags than belts were sold. But why? In which regions? For which times of the year? How long has this been a trend? TL;DR? Pie charts don’t provide meaningful information, so we’ll leave them behind in elementary schools.
As my soccer fans would say, upper 90°! Or for those of you who don’t watch soccer, start with the top, left-hand corner. If you’re going to put more than one visualization on a single worksheet, put the most important information in the top left-hand corner, and make that the largest visualization on the page. Visual provided below. Shoot for the star(s)!
You’re in control of how viewers understand your data. Make them want to look, then make the data easy to analyze. Reel ’em in! That being said, this isn’t art class. Colors and size are important, shapes aren’t as imperative. Like I said, you’re in control, so choose wisely. Humans can only process roughly 8 different colors at a time. Personally, I’d stick to three different colors. If you must use more, stick to a theme. If you’re still unsure, follow these guidelines: 1. decide what the viewer should conclude from your visualization. 2. don’t make them work for it!
Let’s talk about Comic Sans. Don’t. Just, don’t. Consider your audience and choose fonts and text sizes that are easy to read. This is the era of information overload after all…and if your data is difficult to read, it’ll pop right off of a viewer’s interest stack frame. Here are some examples of appropriate fonts: Arial, Trebuchet MS, Verdana, Times New Roman, Lucida sans. Pro tip: Times New Roman is said to be the fastest font to read.
Congratulations! You’re four steps closer to being a data visualization pro.
