Beware Terrible Charts

This chart is a work of art. I hope multiple people got paid well to make it and I hope they get hit by a bus pic.twitter.com/XS6aWwwj7P
— Katelyn Gadd (@antumbral) June 24, 2021

Charts and graphs can be wonderful for breaking down information and showing trends over time, but only if they are constructed in a way that people can understand. On TV, the viewer only has a short period of time to grab that information, but when you take a screenshot, you see that there are many things wrong with this graph. Mefite Stark takes a stab at listing them.

1. The y-axis doesn't start at zero
2. The y-axis isn't labelled correctly (it should be % of Adults who report that they think violent crime is a problem)
3. The gaps between the data points on the x-axis are not representative of the gaps between the dates
4. The margin of error is not highlighted on the chart's data points
5. Straight lines between the data points imply a steady trend which may or may not be the case
6. In my opinion the most amazing weirdness is that the x-axis dates go from right to left!

Someone else mentioned that the x-axis is labeled in alphabetical order. Andy Baio took the information and made a much better chart.

Wait so it isn't even for actual violent crime? Just adults who THINK there's a violent crime problem?????
— Crim Lust (@Cr1ms0n_lus7) June 25, 2021

This is far from unique, though, as many media outlets rely on their graphics departments (or maybe an intern) to make charts instead of data scientists. You can go down the rabbit hole and see plenty of confusing or misleading information in charts here, here, and here. In fact, there are so many ways to make a bad chart that several sites are dedicated to explaining and shaming them. See more terrible charts archived at Bad Visualizations, WTF Visualizations, and the subreddit Data Is Ugly. -via Metafilter

That being said - according to fbi statistics a violent crime in the us happens roughly every 25 seconds. Might be a bad chart, but I don't think they're making up the "violent crime" thing.

Abusive comment hidden. (Show it anyway.)

They Y-axis doesn't have to start at zero and in many cases it shouldn't. The Y-axis needs to cover a range of values that makes sense for the data being displayed. For example is you wanted to show the average selling price of a home in NYC between 1990 and 2020 having the Y-axis start at zero would either require you to have a very tall chart or values that would make large differences seem nearly flat or require you to use a break line to skip the zero to some large value.

Abusive comment hidden. (Show it anyway.)

"This is far from unique, though, as many media outlets rely on their graphics departments (or maybe an intern) to make charts instead of data scientists." The idea that huge, multi Billion dollar media outlets just accidentally do this sort of thing because of a silly intern is laughable. This is by design with specific intent to manipulate the viewer while maintaining plausible deniability. They have whole teams of highly paid and highly qualified scumbags to specifically do this sort of crap. If your media group and it's advertisers are big and rich enough to be lobbying congress this is the kind of shenanigans you end up with.

Abusive comment hidden. (Show it anyway.)

Login to comment.

Click here to access all of this post's 3 comments