Total Posts:12|Showing Posts:1-12
Jump to topic:

Correlation: Gun Ownership/Homicide Rate

JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 7:10:19 PM
Posted: 4 years ago
I thought this would be fun, as obvs lots of people are talking about guns right now. I made this chart because using the states inside the US is a much more useful comparison than trying to compare countries which have far more differences than states.

This is a chart showing the homicide rate per 100,000 citizens in each of the 50 states. The states are listed in order by gun ownership percentage, ranging from 8% to 60%.

Here's the fun part: You have to guess whether the states with high gun ownership rates are listed at the top, or the bottom.

http://i48.tinypic.com...

And just to show how correlated gun ownership and homicide rates are:

The average homicide rate for the 25 states with highest rates of gun ownership was 4.54.
The average homicide rate for the 25 states with lowest rates of gun ownership was 4.79.

So the next time someone says more guns = more crime... slap some sense into them.
twocupcakes: 15 = 13
JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 8:37:01 PM
Posted: 4 years ago
Apparently the correlation of more guns = less homicides is 0.017.

GUN CONTROL ADVOCATES WANT PEOPLE TO DIE!!!
twocupcakes: 15 = 13
DanT
Posts: 5,693
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 8:51:00 PM
Posted: 4 years ago
At 7/22/2012 7:10:19 PM, JaxsonRaine wrote:
I thought this would be fun, as obvs lots of people are talking about guns right now. I made this chart because using the states inside the US is a much more useful comparison than trying to compare countries which have far more differences than states.

This is a chart showing the homicide rate per 100,000 citizens in each of the 50 states. The states are listed in order by gun ownership percentage, ranging from 8% to 60%.

Here's the fun part: You have to guess whether the states with high gun ownership rates are listed at the top, or the bottom.

http://i48.tinypic.com...

And just to show how correlated gun ownership and homicide rates are:

The average homicide rate for the 25 states with highest rates of gun ownership was 4.54.
The average homicide rate for the 25 states with lowest rates of gun ownership was 4.79.

So the next time someone says more guns = more crime... slap some sense into them.

There is only a 41% correlation (r = 0.41) between households with loaded firearms, and violent crimes.

The sum of squared errors of prediction is 43,953% (SSE = 439.53); which means it is an extremely flawed model. Anyone who tries to say guns lead to more crime are using an extremely flawed model. It's impossible to predict crime rates based gun ownership, and is very little correlation between gun ownership and crime rates.
"Chemical weapons are no different than any other types of weapons."~Lordknukle
DanT
Posts: 5,693
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 8:51:40 PM
Posted: 4 years ago
At 7/22/2012 8:51:00 PM, DanT wrote:
At 7/22/2012 7:10:19 PM, JaxsonRaine wrote:
I thought this would be fun, as obvs lots of people are talking about guns right now. I made this chart because using the states inside the US is a much more useful comparison than trying to compare countries which have far more differences than states.

This is a chart showing the homicide rate per 100,000 citizens in each of the 50 states. The states are listed in order by gun ownership percentage, ranging from 8% to 60%.

Here's the fun part: You have to guess whether the states with high gun ownership rates are listed at the top, or the bottom.

http://i48.tinypic.com...

And just to show how correlated gun ownership and homicide rates are:

The average homicide rate for the 25 states with highest rates of gun ownership was 4.54.
The average homicide rate for the 25 states with lowest rates of gun ownership was 4.79.

So the next time someone says more guns = more crime... slap some sense into them.

There is only a 41% correlation (r = 0.41) between households with loaded firearms, and violent crimes.

The sum of squared errors of prediction is 43,953% (SSE = 439.53); which means it is an extremely flawed model. Anyone who tries to say guns lead to more crime are using an extremely flawed model. It's impossible to predict crime rates based gun ownership, and is very little correlation between gun ownership and crime rates.

almost forgot.... Here is the chart; http://www.debate.org...
"Chemical weapons are no different than any other types of weapons."~Lordknukle
JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 8:54:02 PM
Posted: 4 years ago
At 7/22/2012 8:51:00 PM, DanT wrote:
There is only a 41% correlation (r = 0.41) between households with loaded firearms, and violent crimes.

The sum of squared errors of prediction is 43,953% (SSE = 439.53); which means it is an extremely flawed model. Anyone who tries to say guns lead to more crime are using an extremely flawed model. It's impossible to predict crime rates based gun ownership, and is very little correlation between gun ownership and crime rates.

You know, I just realized something... I've forgotten most of what I learned in college.

I realized this several times last week as well, but my memory isn't so good either...

I think it's ironic that the chart I made ended up being almost perfectly chaotic.
twocupcakes: 15 = 13
DanT
Posts: 5,693
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:01:40 PM
Posted: 4 years ago
At 7/22/2012 8:54:02 PM, JaxsonRaine wrote:
At 7/22/2012 8:51:00 PM, DanT wrote:
There is only a 41% correlation (r = 0.41) between households with loaded firearms, and violent crimes.

The sum of squared errors of prediction is 43,953% (SSE = 439.53); which means it is an extremely flawed model. Anyone who tries to say guns lead to more crime are using an extremely flawed model. It's impossible to predict crime rates based gun ownership, and is very little correlation between gun ownership and crime rates.

You know, I just realized something... I've forgotten most of what I learned in college.

I realized this several times last week as well, but my memory isn't so good either...

I think it's ironic that the chart I made ended up being almost perfectly chaotic.

There are no labels on your chart, so it's hard to read.
"Chemical weapons are no different than any other types of weapons."~Lordknukle
JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:04:31 PM
Posted: 4 years ago
At 7/22/2012 9:01:40 PM, DanT wrote:
At 7/22/2012 8:54:02 PM, JaxsonRaine wrote:
At 7/22/2012 8:51:00 PM, DanT wrote:
There is only a 41% correlation (r = 0.41) between households with loaded firearms, and violent crimes.

The sum of squared errors of prediction is 43,953% (SSE = 439.53); which means it is an extremely flawed model. Anyone who tries to say guns lead to more crime are using an extremely flawed model. It's impossible to predict crime rates based gun ownership, and is very little correlation between gun ownership and crime rates.

You know, I just realized something... I've forgotten most of what I learned in college.

I realized this several times last week as well, but my memory isn't so good either...

I think it's ironic that the chart I made ended up being almost perfectly chaotic.

There are no labels on your chart, so it's hard to read.

I purposefully left them off. I wanted to see if someone could tell which end of the chart was states with lots of guns, and which end was states with few guns, just by looking at it.
twocupcakes: 15 = 13
16kadams
Posts: 10,497
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:05:34 PM
Posted: 4 years ago
Bad way to compare unless you are using regressions and other statidtical models to avoid statistical problems. In other words, even though I agree, your point falls into the endogenity problem.
https://www.youtube.com...
https://rekonomics.wordpress.com...
"A trend is a trend, but the question is, will it bend? Will it alter its course through some unforeseen force and come to a premature end?" -- Alec Cairncross
DanT
Posts: 5,693
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:07:51 PM
Posted: 4 years ago
At 7/22/2012 8:37:01 PM, JaxsonRaine wrote:
Apparently the correlation of more guns = less homicides is 0.017.

GUN CONTROL ADVOCATES WANT PEOPLE TO DIE!!!

A 1% correlation is not very strong. I think your math may be wrong. If r = -0.017 when measuring guns to homicides, than more guns = 1% less homicides.

According to the SSE, models that try to predict violent crime rates based on gun ownership are severely flawed.
"Chemical weapons are no different than any other types of weapons."~Lordknukle
JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:10:49 PM
Posted: 4 years ago
At 7/22/2012 9:05:34 PM, 16kadams wrote:
Bad way to compare unless you are using regressions and other statidtical models to avoid statistical problems. In other words, even though I agree, your point falls into the endogenity problem.

It's not meant to be a serious study. It's meant to shut up people who say 'Durr, but the UK don't have ur gun murders'.

Certainly, if comparing country to country with no controls, then state to state should be fine too :D
twocupcakes: 15 = 13
JaxsonRaine
Posts: 3,606
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:11:32 PM
Posted: 4 years ago
At 7/22/2012 9:07:51 PM, DanT wrote:
At 7/22/2012 8:37:01 PM, JaxsonRaine wrote:
Apparently the correlation of more guns = less homicides is 0.017.

GUN CONTROL ADVOCATES WANT PEOPLE TO DIE!!!

A 1% correlation is not very strong. I think your math may be wrong. If r = -0.017 when measuring guns to homicides, than more guns = 1% less homicides.

According to the SSE, models that try to predict violent crime rates based on gun ownership are severely flawed.

I wasn't being serious :)
twocupcakes: 15 = 13
16kadams
Posts: 10,497
Add as Friend
Challenge to a Debate
Send a Message
7/22/2012 9:17:21 PM
Posted: 4 years ago
At 7/22/2012 9:10:49 PM, JaxsonRaine wrote:
At 7/22/2012 9:05:34 PM, 16kadams wrote:
Bad way to compare unless you are using regressions and other statidtical models to avoid statistical problems. In other words, even though I agree, your point falls into the endogenity problem.

It's not meant to be a serious study. It's meant to shut up people who say 'Durr, but the UK don't have ur gun murders'.

Certainly, if comparing country to country with no controls, then state to state should be fine too :D

Comparing country to country is not fine, I wrote a long post no one bothered to read on these forums yesterday. As the only way to get good trends relies on comparisons, as long as the proper steps are taken.
http://www.debate.org...
https://www.youtube.com...
https://rekonomics.wordpress.com...
"A trend is a trend, but the question is, will it bend? Will it alter its course through some unforeseen force and come to a premature end?" -- Alec Cairncross