Why Visualization

This article was also published in Visual.ly.

Visualization is hot right now; all of a sudden there are dozens of companies and products that want to help you visually analyze your data, build your own visualizations, and visually display interesting data sets of all kinds. So, why is visualization interesting? Why is it desirable? And most importantly, why is it useful?

To answer these questions, we need to go all the way back to biology. The answer is, fundamentally, our visual system is extremely well built for visual analysis. There’s a huge amount of data coming into your brain through your eyes; the optic nerve is a very big pipe, and it sends data to your brain very quickly (one study estimates the transmission speed of the optic nerve at around 9Mb/sec). Once that data arrives at the brain it’s rapidly processed by sophisticated software that’s extremely good at tasks such as edge detection, shape recognition, and pattern matching.

That last factor, pattern matching, is the key when it comes to discussing the benefits of presenting information visually. Typically, the important messages in data are represented in the patterns and pattern violations: trends, gaps, and outliers. This is the interesting stuff. This is meaning. This is what we go to the data hoping to find.

Let’s look at the classic instructive example, Anscombe’s Quartet, devised by statistician Francis Anscombe to demonstrate this very issue. The four following data sets share many characteristics: mean, variance, correlation, and regression.

If you were looking at them in a spreadsheet, you wouldn’t find much of interest to discuss. The magic happens when these data sets are graphed:

All of a sudden, the data sets look very different from each other, and you can’t help but see and understand much more about the underlying numbers; that’s the magic. This is exactly why we use visualization: it casts data into a format that can be grasped and understood much more quickly and easily than the raw numbers alone. Visualizations are so capable and powerful at conveying knowledge that they can be more effective than words at changing people’s minds. ( “Graphical corrections are also found to successfully reduce incorrect beliefs among potentially resistant subjects and to perform better than an equivalent textual correction.”)

So we know that visualization is effective at conveying knowledge. Besides propaganda, what is it good for?

One of the most important benefits of visualization is that it allows us access to huge amounts of data in ways that would not be otherwise possible. There are thousands of examples of visualizations of big data, from fun, to current and historic, to financial and mortal.

The knowledge encompassed in these various data sets would be nearly inaccessible to the casual, or even moderately interested viewer, if it was not visualized. But a good visualization gives us access to that knowledge, and does so quickly, efficiently, and effectively. (Naturally deeper insight comes from deeper study, but there’s a lot to be said for quick access and reader engagement.)

Finally, and perhaps most importantly, visualizations give us access to actionable insight. It’s not until we have access to the knowledge within the data that we can actually act on it. Consider situations where pricing changes over time, such as real estate or used cars. With visualizations, we can quickly understand if we should rent or buy, or if that used car is a good deal.

Without that insight, we’re effectively lost, and the odds of us making a good decision are considerably reduced.

So there it is. We’re wired for visualization, it allows access to challenging data sets, it allows exploration, it can be fun, and provides useful information in an efficient way. That’s why visualization is hot right now, and why this is just the beginning.