txblackout
Member
- Joined
- Jan 29, 2014
- Messages
- 339
When discussing guns and gun violence here are some straightforward numbers:
High positive or negative number indicates strong correlation or anti correlation
Results:
Correlation of guns per/capita to all murders by state -.339 No correlation
Correlation of guns per/capita to gun murders by state -.328 No correlation
Correlation of murders to % african american population by state +.788 positive correlation
Correation of gun murders to % african american population by state +.811 positive correlation
Methodology:
Data on gun murders is from wikipedia (from the census bureau)
http://en.wikipedia.org/wiki/Gun_violence_in_the_United_States_by_state#cite_note-fbi2010-1
Data on african american populations is from wikipedia
http://en.wikipedia.org/wiki/List_of_U.S._states_by_African-American_population
Data on african american populations from district of columbia
http://en.wikipedia.org/wiki/Demographics_of_Washington,_D.C.
Used the correl function in excel to compare columns of state data
Key flaws/assumptions/caveats:
This is just a quick and dirty back of the envelope analysis.
The data is not necessarily controlled for exact years. So the gun violence data does not necessarily match the exact year of the demographic data. The assumption is that the demographics dont change substantially within a few years.
Correlation is not the same as causation.
I just used the most convenient excel function. There are probably better statistical methods to use.
Im assuming guns per capita is legal guns per capita.
Crimes committed with guns would be another way to look at this. Murders are at the extreme end of this. My assumption is that murder rates are somewhat correlated with violent crime rates committed with guns. This might not be true. Even if it is not true, it doesn't matter for this analysis about guns and relationship to murder. Another analysis of guns related to violent crimes would be worth doing.
Discussion
Murders and gun murders are somewhat correlated with having a high african american population and not correlated at all with guns per capita.
I was surprised that gun murders did not have more of a correlation with per capita (legal) gun ownership.
One reasonable, underlying assumption by gun control advocates might be that murders are highly correlated with illegal gun ownership and that legal gun ownership is correlated with illegal gun ownership. They in essence have created a logic chain that looks like [more legal guns -> more illegal guns -> more murders].
Like many logic chains, it seems reasonable. However, because there is no correlation between per capita legal gun ownership and murders, there is either no correlation between legal and illegal gun ownership OR there is no correlation between illegal gun ownership and murders. One of the links in the logic chain does not hold.
why is the african american component not discussed, ever?
I have heard people suggest this on various news programs and they are typically shot down vehemently. But not with data.
It is political/media suicide to discuss this or to push the issue. The sad thing is, the answer to murders is probably to find a fix to violence in the african american community. Neither liberals nor conservatives want to talk about the problem which means it cant be solved.
Liberals believe that by removing the tools for murder that they can reduce the murder rate. That way they dont have to address the uncomfortable problems in the inner city african american community. This doesnt make them bad or evil people, just in denial.
Liberals usually try to lump suicide in as "gun deaths" or "gun violence". In any discussion of data, this is the first thing to look for.
Ultimately I think murders are the cleanest way to look at this.
High positive or negative number indicates strong correlation or anti correlation
Results:
Correlation of guns per/capita to all murders by state -.339 No correlation
Correlation of guns per/capita to gun murders by state -.328 No correlation
Correlation of murders to % african american population by state +.788 positive correlation
Correation of gun murders to % african american population by state +.811 positive correlation
Methodology:
Data on gun murders is from wikipedia (from the census bureau)
http://en.wikipedia.org/wiki/Gun_violence_in_the_United_States_by_state#cite_note-fbi2010-1
Data on african american populations is from wikipedia
http://en.wikipedia.org/wiki/List_of_U.S._states_by_African-American_population
Data on african american populations from district of columbia
http://en.wikipedia.org/wiki/Demographics_of_Washington,_D.C.
Used the correl function in excel to compare columns of state data
Key flaws/assumptions/caveats:
This is just a quick and dirty back of the envelope analysis.
The data is not necessarily controlled for exact years. So the gun violence data does not necessarily match the exact year of the demographic data. The assumption is that the demographics dont change substantially within a few years.
Correlation is not the same as causation.
I just used the most convenient excel function. There are probably better statistical methods to use.
Im assuming guns per capita is legal guns per capita.
Crimes committed with guns would be another way to look at this. Murders are at the extreme end of this. My assumption is that murder rates are somewhat correlated with violent crime rates committed with guns. This might not be true. Even if it is not true, it doesn't matter for this analysis about guns and relationship to murder. Another analysis of guns related to violent crimes would be worth doing.
Discussion
Murders and gun murders are somewhat correlated with having a high african american population and not correlated at all with guns per capita.
I was surprised that gun murders did not have more of a correlation with per capita (legal) gun ownership.
One reasonable, underlying assumption by gun control advocates might be that murders are highly correlated with illegal gun ownership and that legal gun ownership is correlated with illegal gun ownership. They in essence have created a logic chain that looks like [more legal guns -> more illegal guns -> more murders].
Like many logic chains, it seems reasonable. However, because there is no correlation between per capita legal gun ownership and murders, there is either no correlation between legal and illegal gun ownership OR there is no correlation between illegal gun ownership and murders. One of the links in the logic chain does not hold.
why is the african american component not discussed, ever?
I have heard people suggest this on various news programs and they are typically shot down vehemently. But not with data.
It is political/media suicide to discuss this or to push the issue. The sad thing is, the answer to murders is probably to find a fix to violence in the african american community. Neither liberals nor conservatives want to talk about the problem which means it cant be solved.
Liberals believe that by removing the tools for murder that they can reduce the murder rate. That way they dont have to address the uncomfortable problems in the inner city african american community. This doesnt make them bad or evil people, just in denial.
Liberals usually try to lump suicide in as "gun deaths" or "gun violence". In any discussion of data, this is the first thing to look for.
Ultimately I think murders are the cleanest way to look at this.