Hacker News Reader

Comment by davinic | Hacker News Reader

There are several studies on this that conclude otherwise. For example[1]:

"Firearm prevalence is significantly related to total violent crime (B = .600, p < .05). With each unit increase in firearm prevalence, the expected count of the violent crime index increases by .600. This also indicates that the percent change in the total violent crime is an 82% increase for every unit increase in firearm ownership. The prevalence of guns does significantly increase the violent crime in the county. This finding is consistent with previous research on firearm prevalence and crime both in the United States and internationally."

[1]: http://www.cjcj.org/uploads/cjcj/documents/jpj_firearm_owner...

View on HN

So the study found a correlation between suicide by firearm and violent crime, not firearm prevalence and violent crime. They try to justify using suicide by firearm as a stand in for gun ownership, but if they don't have good stats on gun ownership to use for this study, then how could the previous studies have been able to make the connection between suicide by firearm and prevalence since they would need the same data.

>Firearm prevalence in the United States is difficult to determine because there is no database that collects information on firearm ownership and prevalence. Thus, analyses that study firearm prevalence have had to develop proxies for firearm ownership. As a proxy for firearm ownership, the current analysis used the percentage of suicides by a firearm from 2000 to 2010.

Little tricks like this are why people lose faith in science.

Proxies are used all of the time in studies and analysis of information that can't be directly collected. To claim that proxies are why people lose faith in science is absurd.

I view that study with suspicion as there are also some which show a very poor link or no link at all. But in either case: prohibiting the possession, sale, and consumption of alcohol would decrease disease and increase productivity. Why should we not do that?

That's an interesting correlation matrix on page 13.

Reply on news.ycombinator.com