Anyone can make a chart. There’s enough how-to’s on the internet to twist data in hundreds of useless shapes that will depict data however you see fit. Here is an example of a beautifully made chart from Charles Joseph Minard depicting Napolean’s winter march to Moscow and the number of troops over the campaign: The chart, for those of you that cannot read French, is drawn as a map. The thickness of the line is relative to the number of troops in Napoleon’s army. The left side of the chart represents France and the right Moscow. There is no question that, based off the size of the line, the number of troops Napoleon starts with (tan line) is tenfold larger than the number he ended with (black line). There is even a space that shows Napoleon’s army attempting to cross a frozen river. The most important piece, and the one that got Napoleon in trouble, is the starting/ending numbers. When concept and context meet, the representation can be devastatingly impactful. The number of deaths becomes better represented by this chart than any series of numbers could allow.

“If only one man dies of hunger, that is a tragedy. If millions die, that’s only statistics.” Joseph Stalin (1947)

Above is the exact same chart of Minard’s. Yup, terrible. It doesn’t bring the emotional weight of the information and lacks the details that make the data meaningful. There are a ton of rules being broken for-the-worse. It makes me sick to look at it and I created it to serve a point. Ignorance surrounds charts, but it shouldn’t. Take a look at a more recent pie chart that appeared on a well known news network.

Hopefully, the first thing you notice is that the percentages add up to 128%. Among other issues, the chart also claims the MOE “Margin of Error” is the source. This was all done, I believe, intentionally. It brings attention to the chart which brings people to the site and drives up unique user engagement. In other words, “any news is good news”. That all being said, there are instances in which, a information designer can be creative and ethical. Below is a fantastic picture by one of Prof. Sackey’s University of Texas students:

So how is representing numbers effectively accomplished? I’ll give you a hint, it’s not just about knowing design. Knowledge makers and translators need to know the science of information design. In the case of representative number comparison through graphics, we rely on Steven’s Power Law. Steve’s Power Law effectively written as:

You don’t need to know the math, but you do need to know what’s happening. For all sensory stimuli (vision, taste, sound, etc), the closer in relation that two objects appear based heavily on ratio, the more difficult it is to distinguish the difference in the size. Here is an example of a graph that I came across that plays around with this concept: Without the change from a stein to a cup, the difference in caffeine represented would be difficult to distinguish. Seriously, how often can you tell the difference in squiggly lines? It’s not a common comparison. The same goes for charts and graphs. Two seemingly recognizable objects formatted for comparison will tend to have issues with scale-ability. This is not only a concept that distorts data, but is commonly used as a convincing point. Think of it like Stalin’s quote. Imagine a chart that only has two numbers 1 and 1,000,000 with a scale of 1,000,000. The 1 looks like a thin line by comparison. Now add another number, let’s say 10. The 10 and the 1 look nearly identical. Change that 10 to 100 and I doubt a random person would be able to tell the difference, especially if the chart is not labeled. Those differences can influence decision makers and, more importantly, influence the average person. The ethical results vary wildly with the editing of the chart. Before I get into too much of a rant with ethics, just consider the amount of knowledge it takes to properly make something as seemingly simple as a chart or graph. Question the the heck out of one when you see it. Steven’s law comes into place more than you think. Some examples include: bubble charts, 3d rendering of a pie chart or column graph, and line graphs that fail to express correct X-Y scaling. The list really goes on and on. The purpose of these images is to make information more accessible, not deceive. So, if you see an image that has taken data from literally anywhere and recreated it with some form of chart/graph, scrutinize it. Don’t take it for face value. If you find that you are the one creating the chart/graph, try your best to follow as many best practices as possible.