Join Bill Shander for an in-depth discussion in this video Ask what makes a good data visualization, part of Learning Data Visualization.
So, in some ways, this might be the most important movie in the course. How do you make a good visualization? You might ask yourself that. And I actually use that as an acronym to help me remember. Really, what it comes down to, is accurate story knowledge. If you can create visualizations that are accurate, and that tell a good story, even if it's not a linear story. And provide real knowledge to your audience. Then that's a good target to shoot for. So let's talk about accuracy. This was published by a national news network, and as you can see it's, tracking the unemployment rate under President Obama.
And, I'd like you take a look at it for you know, about five or ten seconds, 30, however long you have. You can pause the video if you really want to think about it before I tell you the answer. And I'd like you to tell me, what's wrong with it. So you might have noticed. That data point on the far right in November, the most recent one, is showing us 8.6% and yet look at where it is visually, it's on the same line as 9%, and if you compare to the 8.8% data point in March, it's higher than that one. So, this is just wrong.
This is what it should have looked like. Can you see any other issues with this chart that might qualify it as being inaccurate? Take a look at the scale. The highest data point here is 9.2%, and the lowest is 8.6%. The scale should have been set from 8.6% to 9.2%, right? From the very highest to the very lowest, or maybe 8.5% to 9.5%, like these red dotted lines show, or 8.7 to 9.3. You can play around with it a little bit, but showing all the way from 7.5 to 10.5, is exaggerating to show a flatter line.
It's accurate technically, but, it's misleading. This data comes from the Bureau of Labour Statistics, and this is how they provided it. Now, if you look at one more issue here, take a look at the headline, and the x axis label. This is supposed to reflect the unemployment rate under President Obama, that's what this is saying, which sounds like, this is for the entire length of his term. But of course this isn't. This is just for the year of 2011. So the data really isn't providing the unemployment rate under President Obama. It's the unemployment rate for 2011.
So that's accuracy. Let's talk a little bit about story. In the last movie, So What is Data Visualization, I pointed our Charles Menard's visualization of Napoleon's march on Moscow. This is very frequently referred to by Edward Tufte and others, as one of these great, classic examples of visualization. And, one of the reasons is because it tells such a compelling story. Napoleon's army, marched into Moscow, and then retreated from Moscow. And lost something like 98% of its army, 400,000 men.
And, while you can read that in a sentence, and maybe you can see a bar chart that sort of reflects it, something about this really tells the story. You can see them moving to the east, and the army just getting smaller, and smaller, and smaller. They hit the Berezina River on the way back, and it shrinks by 50% yet again, and look at how thin that line is by the time it gets back. Meets up with a few more of its forces, but what a compelling story. And you can also track the temperature and how cold it was, which helps explain what was going on. Finally, knowledge. Another classic example of data visualization.
This is a map created by John Snow, who was a doctor in the 19th century and, he had a theory that cholera was a waterborne disease. At the time people thought it was an airborne disease. And, when there was an epidemic of it, he mapped out where all the deaths were in. See the little black lines next to the homes where people were dying from cholera. And he was able to prove that they were clustered around this one water pump here on Broad Street in London. He took this data, and transformed the understanding of cholera.
For the entire world going forward. This imparted real knowledge, in fact, his work including this map, was part of the origin of the fields of epidemiology and public health. If you're careful and accurate, and you tell a good story, and you impart real knowledge to your audience, you can transform them in ways that just informing them will never do.
- Channeling your audience
- Understanding your data
- Determining the information hierarchy
- Sketching and wireframing your ideas
- Defining your narrative
- Using typography, color, contrast, and shape to convey meaning
- Making your visualization interactive