Gun Violence in the US from 2013-2016(deborah-digges.github.io)
deborah-digges.github.io
Gun Violence in the US from 2013-2016
http://deborah-digges.github.io/shooting-viz/index.html
25 comments
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.
Yeah that is pretty ridiculous, it is non-obvious that the scale changed.
Makes me wonder if it is purposely misleading.
I've updated the visualization to correct this
I've updated the visualization to correct all the mistakes it was making
I wish the scale didn't change per year.
Does anyone else find the gun crime in Washington DC strange and/or ironic?
Does anyone else find the gun crime in Washington DC strange and/or ironic?
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.
You're right. The scales changing don't allow changes in states over time to be seen. It is strange isn't it?
> It is strange isn't it?
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.
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.
It (the strangeness) is referring to a separate idea in my post, which was the crime level in Washington DC, not the scaling.
Its not strange, its lying with statistics.
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.
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.
I get a blank page with just the title and the drop-down boxes at the bottom.
Perhaps the d3 library isn't loading. What's the error you see in your network tab?
Mixed Content: The page at 'https://deborah-digges.github.io/shooting-viz/index.html' was loaded over HTTPS, but requested an insecure script 'http://d3js.org/d3.v3.min.js'. This request has been blocked; the content must be served over HTTPS.
go to the http version of the page, and it works fine. http://deborah-digges.github.io/shooting-viz/index.html
Hi, I get 304 errors when opening the page. Chrome on windows 7.
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.
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.
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.
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.
According to "Homicide in California 2013" (https://oag.ca.gov/sites/all/files/agweb/pdfs/cjsc/publicati...) Table 21, in 2013, there were 1,699 homicide crimes where the type of weapon used was a firearm.
That's just crime data, it doesn't include suicides.
Am I reading this wrong or is the data way off or is it indicating some other kind of firearm violence?
According to "Homicide in California 2013" (https://oag.ca.gov/sites/all/files/agweb/pdfs/cjsc/publicati...) Table 21, in 2013, there were 1,699 homicide crimes where the type of weapon used was a firearm.
That's just crime data, it doesn't include suicides.
Am I reading this wrong or is the data way off or is it indicating some other kind of firearm violence?
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.
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.
Answering my own question - the graph is about mass shootings. The data is from http://www.shootingtracker.com.
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!