I collected some data on homocide vs firearms ownership by country, and I found that there was almost no correlation between the two.
Pic related is a graph of gun ownership per 100 persons (data taken from here en.wikipedia.org
This first graph is one in which outliers (many of the south american countries, as well as the US, are left in).
Now, The above data is highly affected by outliers that were intentionally left in, but it does show a NEGATIVE relationship between firearms per capita and homicides per capita. Pic related is the linear regression data. As you can see, it has a negative trendline, but the correlation is very, very low.
On this chart, R-square is the correlation coefficient; basically, it is the percentage of the variation in a y variable that can be explained by the chosen x variable (in this case, 0.79% of the variation in the homicide rate can be explained by variation in guns per capita). The chosen regression line has an equation of y= -0.08x + 8.1, (xvaraible 1 coefficient and intercept coefficient in the image), which would seem to suggest that every gun per 100 persons reduces the homicide rate by 0.08. (The correlation between x and y is very low, so take the regression line with a grain of salt)
This is the graph of the above, with influential points like the US (a country with very high gun ownership and moderately high homicide rates) and many latin american countries (which have very low amounts of guns per capita, but extremely high homicide rates) removed. As you can see, there is still a negative relationship between guns per capita and homicides per capita.
Gun ownership doesn't correlate with suicide either.
What's important in the second graph isn't the negative trendline; what's important is the Low correlation coefficient. Even with outliers removed (most of which actually aid the argument that more guns doesn't mean more crimes, such as the south american countries), correlation is still a measly 0.01, indicating there is no correlation between homicides and guns per capita.
Gun policy doesn't correlate with homicide due to a replacement effect with other methods of killing people.
There's strong evidence that fatherlessness is a major factor in homicide rates.
drive.google.com
Link to the file I used.
bumping
Bump
Excellent OP!
Of course if you really want to take a bottle of red pills explore the relationship between gun crime and race...
good research
Also, this goes to show that you shouldn't trust those images people always post about guns vs homicide rate, because the end up leaving out a lot of countries with low gun ownership and high homicide rates, and including outliers like the US.
great post OP
Scientific inquiry is strengthened by peer review. If any of you have any holes you want to try and poke in my numbers (even if its obvious bait), please go for it.
I would but I only took one data mangement/stats course and that was in high school, a while ago. Forgotten most stuff but these sheets are bringing it back.
Both quantitative and qualitative criticisms would be helpful.
not criticism, but this may be of interest to you
Interesting. I wonder what the correlation is here.
I woudln't care if a million people died from guns in the US daily, you're not taking my fuckin guns commie bastards
I was going to attempt to graph homicide rates vs the economic freedom index to show that there's more of a correlation between those two than guns and homicide, but the VBA macro i wrote is being a bitch and i honestly can't be arsed.
bumping
Nice thread op.
Thanks for not being a shit.
As a token of my appreciation please have this rare photo of Hillary Clinton appearing on stage with Hitler
who is that on the right
looks like lady gaga
...
Dude, that's just a proxy measurement of niggers.
It would be interesting to see if there's any estimates of illegal firearm ownership.
An illegal firearms clearly implies higher homicide rate, but how do the people respond to illegal firearms.
The statistics I used include the number of registered firearms in the country, as well as an added estimate of the number of illegal firearms in the country.
...
It's hard to be critical of someone doing God's work...
Thanks user
Great work OP
All in a day's work
...
...
would be interesting to see how fatherlessness affects white (vs how it affects blacks)
...
A woman is shot by her partner every 16 hours.
That's less than 500 people per year.
most murder victims and suicide victims are male afaik.
meant for
Wow. A few hundred a year in a country of 350 million people.
Then she should get back in the fucking kitchen.
bamp
Sup Forums pol maybe isn't the best chan for level headed. too many shills and shitposters
If you run a regression you should include controls for the variation among states, like sociodemographic controls, race, income, education etc.
Also a simple regression doesn't mean causality since you have omitted variable bias (lacking controls) or didn't rule out reverse causality (simultaneity).
But you still did a good job. I'm very pleased to see some serious stats / econonetrics.
Use R (rstudio)
Is it free?
Yes
getting all those data would be a mission. we know about MAO-A anyway
>en.wikipedia.org
race and gun murder
Bump for actual legitimate content.
How in god's name do I use this thing
Or rather, what does R do that excel can't? I'm not keen on learning an entirely new programming language just for one thing
The single most delicious part of that is that places with high illegal gun ownership still do not correlate to violence -- if white!
Every suburban home outside Chicago has a nine millimeter pistol in the underwear drawer of the parent's bedroom. It is functionally used just like a legal weapon -- there to protect the house, not knock over gas stations -- so it never shows up.
>using excel for analysis and visualization
Nevermind, i'm figuring it out. Still though, what advantages are there to using R besides just using excel?
...
Wow
...
Multiple linear regression using these variables.
Could be problems with the gun estimates. If they are only picking up legal, by the book guns held by law abiding people and missing all the guns held by criminals in secret, you would expect this result.
I respect you for doing this user. You're one of the good ones.
Doesn't seem like there's a significant relationship between gun ownership and firearm homicides, user. Or is that what you're trying to show? I'm confused
He's showing you niggers are niggers has nothing to do with gun ownership
Army knife bro is correct. R is the most based stats package.
See
What does it let me do that excel can't?
The best thing about R in comparison to excel is its not microshit proprietary software.
Good thread, have a bump.
Ok, but assuming i'm not richard stallman, whats the point of me learning R?
It has vastly more statistical functions and visualization features. It can handle larger datasets than excel as well.
It also has an extremely steep learning curve stick to excel
So race is the single most important factor, who would have known.
Don't bother with R. There are some good statistical packages in Python or Octave if you prefer the MATLAB style.
Can I just use MATLAB? I get it for free thru my uni
I've got a bunch. I save more as they show up.
Anyone who's taken a serious look at gun crime and firearms could've told you there's nothing there over two decades ago. There is absolutely zero basis for banning firearms. This is really important to get across. All these years of anti-gun rhetoric - why?
These people KNOW it doesn't stop murders or violence. So why ban guns?
Because they don't want the general public armed.
Why not?
They don't want you to defend yourself.
And that puts everything into a very serious perspective.
...
VERY interested in income. Wonder if that might have a stronger correlation than race.
As an aside, iPython (or the newer Jupyter) notebooks are great for presenting results in a much nicer format than in Excel. It's really easy to get started with too.
Yeah if you get MATLAB for free then that's a great tool to use. Personally I prefer Python for most graphing and statistical analysis tasks but MATLAB is a pretty powerful piece of software if you get used to it.
Bumping for gun control
You're not going to see anything with a higher correlation for murders than race.
6% of US (black male population) commit ~40-45% of all murders in the US. It's obscene.
I make charts too bro
It's fun
I'm more looking globally for this one, but i'll probably do the US next
It's strong but not as strong as race
It's basically true globally as well. There's a huge disparity in black criminality vs white criminality.