On Twitter I came across this tweet by the New York Times Graphic Desk, featuring an article about a new, yet unpublished survey of 4,000 people on gun-ownership.
A couple of users already mentioned concern that this visual representation might be misleading. For one, because it doesn’t put the data into relation with the overall population demographics. And secondly, because with a stacked bar graph, you’d expect figures to add up to 100% which they mostly don’t.
Now, this might be due to rounding, and usually a small note on that front can solve confusion.
But it might also be due to a misunderstanding on my side how to read the graphic. Unfortunately it cites figures that are not publicly available, so I can’t check back with the original source and look for hints.
As I understand it, the graphic is meant to read „Of female gun owners, 42% possess one handgun, 17% possess a long gun, and another 42% possess both types of weapons“.
Of course there is a chance that people of the „both“ group are also within one of the previous groups, but then the sum of percentage shares would most likely be a lot greater than 100 percent and not only a few percentage points.
The other point of criticism relates more to the racial breakdown of gun ownership, as black people make up a smaller share of the overall population. Right now it looks as if there are far more „armed black people“ than there are „armed white people“.
That’s misleading, because in fact the visual doesn’t show that a black person is more likely to have a handgun than a white person. Instead, it shows that — among gun owners — the majority of black people only owns one handgun; whereas the majority of white gun owners has both a hand gun and a long gun.
Contextualization: Gun owners vs non-gun owners
Moreover, as a non-US-citizen, I do only have a rough idea, how many people at all have guns vs those who don’t. At least the number of gun owners is mentioned in the article „Today, about 55 million Americans own 265 million guns.“ So approximately 17% of Americans own a gun.
To do that thoroughly, you’d also need a racial breakdown of the overall gun owners.
Of course this can’t solve the uneven add-up of the gun owner-breakdown into categories, to fix that you’d need the original source. Also, I’m not 100% happy with the color scheme, but I hope it’s at least a little less misleading.
Would be glad to hear about your thoughts how to further improve this graphic!