If you look at the colors, it seems gun violence is much worse in later years. However, if you look at the scale it tells the opposite story; gun violence was far greater in 2013. Also, 2016 isn't over yet, it should be scaled appropriately (which we can only assume is not). This chart is useless as it is.
The scale being inconsistent from year to year is terrible, it leads readers to make inferences that aren't true, and makes it very difficult to understand how gun violence has changed in the time period shown.
Any source on where this data comes from?
Also, this is very visually misleading. If you're not looking at the key closely, it's changing quite a bit between years.
Washington is, comparatively, riddled with violent crime committed with firearms; despite having some of the most restrictive firearms laws in the nation. Ironic doesn't begin to describe the kind of idiocy it takes to then recommend those same measures for the rest of the country, but alas, they do.
And I assume intentional. This is your work and submission. Exactly what statement are you trying to make? A way to deceive with diagrams?
EDIT: To explain, I haven't looked at the raw data, but reading the diagrams and taking the scale changes into account, there aren't especially big differences year to year, but you're led to believe there are. There could even be an overall nationwide decline for all I know, but the misleading rescaling makes it hard to determine by just looking at the images.
I suspect the "it is strange, isn't it" portion of the comment was meant to be a separate idea and the commenter just failed to make it a separate paragraph.
It (the strangeness) is referring to a separate idea in my post, which was the crime level in Washington DC, not the scaling.
I understand why someone would want to use a balanced distribution scale (even though you get weird boundaries sometimes), its so you get the most dynamic visualization and take full advantage of your color library.
What doesn't make sense is redefining the color scales boundaries every time the year changes. As illustrated in this visualization, it intends to lead the viewers to believe violence is getting worse, when in fact the last years maximum number is 5 times lower than the first year.
California's population in 2013 was 38.4 million. According to the map, 2 Californians per million were affected by firearm violence that led to death, so that'd be about 77 people killed by firearms.
1,225 was the number of firearm homicides recorded in 2013, according to that publication. 1,699 was the number of homicides where a weapon was "known".
But yeah, still totally off, AFAICT. I also wonder where she got the supposed 2016 data, given that 2016 hasn't happened yet.
Other interesting tidbits from that PDF: almost as many people were killed with ropes, as with rifles. About half as many were killed with knives and blunt objects as were with handguns. Hardly epidemic, if you ask me. Considering how much more convenient it is to kill somebody with a firearm, you wonder why it's not really that much more popular.
Maps and graphs are usually used to tell a story or to highlight an interesting pattern. Am I missing something here? All I see is a color coded map whose colors change volatilely because the scales are changing.
>tell a story or to highlight an interesting pattern
the story here is how visualizations are leveraged to manipulate peoples' conclusions. Especially when the conclusions inferred by the data directly contradict the authors' own.
This is the source of the data: http://www.shootingtracker.com/. Thank you all for the feedback, I'm just getting into the field and I'm sorry for the very commonplace mistakes I've made with this chart. Rest assured I will fix them and make sure that I keep these suggestions in mind in the future.
Thanks again!
