拥枪率和犯罪率的相关性

turbopascal

这个研究说,拥枪率高,谋杀率高啊.包括除美国外的其他发达国家，包括美国的州：

https://www.hsph.harvard.edu/hicrc/firearms-research/guns-and-death/

大家有什么看法？

jasongxg

只能说这个报告是给控枪人士需要的

东区小刀

jasongxg 发表于 2017-1-10 18:54
只能说这个报告是给控枪人士需要的

这标牌做的真好

deserteagle

jasongxg 发表于 2017-1-10 18:54
只能说这个报告是给控枪人士需要的

这个竖在门前牌子的内容，潜台词是：禁枪的邻居没枪，去抢他们吧！而且我不会帮他们打你……

turbopascal

搜索更可信的Google Scholar找到：http://link.springer.com/article/10.1007/BF01064462

简单说：没有证据说明枪多造成暴力犯罪多：

What effects do gun control restrictions and gun prevalence have on rates of violence and crime? Data were gathered for all 170 U.S. cities with a 1980 population of at least 100,000. The cities were coded for the presence of 19 major categories of firearms restriction, including both state- and city-level restrictions. Multiple indirect indicators of gun prevalence levels were measured and models of city violence rates were estimated using two-stage least-squares methods. The models covered all major categories of intentional violence and crime which frequently involve guns: homicide, suicide, fatal gun accidents, robbery, and aggravated assaults, as well as rape. Findings indicate that (1) gun prevalence levels generally have no net positive effect on total violence rates, (2) homicide, gun assault, and rape rates increase gun prevalence, (3) gun control restrictions have no net effect on gun prevalence levels, and (4) most gun control restrictions generally have no net effect on violence rates. There were, however, some possible exceptions to this last conclusion—of 108 assessments of effects of different gun laws on different types of violence, 7 indicated good support, and another 11 partial support, for the hypothesis of gun control efficacy.

turbopascal

这篇也是来自Google Scholar，说过去关于枪拥有量的数据的假设几乎全无效，并且仍然没有证据说明枪多造成暴力犯罪多

http://journals.sagepub.com/doi/abs/10.1177/0022427803256229

Valid measures of macro-level gun levels are essential to assessing the impact of gun levels on crime and violence rates, yet almost all prior research on this topic uses proxies that are either invalid or whose validity has been assumed rather than demonstrated. The present study uses city, state, cross-national, and time series data to assess the criterion validity of over two dozen gun indicators. The criterion measures used are primarily direct survey measures of household gun prevalence. The results indicate that (1) most measures used in past research have poor validity, making past findings uninterpretable, (2) the best measure for cross-sectional research is the percentage of suicides committed with guns, and (3) there are no known measures that are valid indicators of trends in gun levels, making credible longitudinal research on the subject impossible at present.

zacvoid

jasongxg 发表于 2017-1-10 18:54
只能说这个报告是给控枪人士需要的

我笑抽了

zacvoid

deserteagle 发表于 2017-1-10 19:18
这个竖在门前牌子的内容，潜台词是：禁枪的邻居没枪，去抢他们吧！而且我不会帮他们打你……
...

这个完全翻译才有幽默感

我的隔壁邻居想要禁掉所有的枪
他们家是没有武装的(放大的那一行字)
出于对他们的尊重
我保证不会用我的枪去保护他们

dcx

zacvoid 发表于 2017-1-10 19:42
我笑抽了

非常尊重禁枪的邻居。

dcx

zacvoid 发表于 2017-1-10 19:50
这个完全翻译才有幽默感

我的隔壁邻居想要禁掉所有的枪

多么善良的人啊。虽然自己有强力的武器, 但对与自己意见不同的人是多么的尊重啊。
反观禁枪派, 从来不尊重别人的权力, 就想着怎么把自己的意见强加于人。

gongzi-Z

政客调研 —— 先设定，再论证。

无语

zacvoid 发表于 2017-1-10 19:50
这个完全翻译才有幽默感

我的隔壁邻居想要禁掉所有的枪

308

milit2014

这种研究毫无意义， 只能证明枪具有更大的能量。一个拿刀的一个拿枪的当然不一样。 要看犯罪率上升还是下降。在就是典型的拿数据讲故事，适合搞传销。

ca1903

Pascal，原來的文章沒有數字，只有論證。

同理可證，擁槍多的犯罪率少。

turbopascal

ca1903 发表于 2017-1-11 05:46
Pascal，原來的文章沒有數字，只有論證。

同理可證，擁槍多的犯罪率少。

原文引用了很多篇peer reviewed文章，那些文章都有数据的。

HRQYH

这种统计方法不科学。结果自然不正确。结论必然是错的。

moses80h

纽约、芝加哥犯罪率如何？

foxea

任何一个对争议性话题（controversial topic ）的学术评论 （review），如果只列了一方的观点，那作者肯定是有偏见的。很简单的逻辑：如果数据都是一边倒，这个话题怎么可能有争议性（controversial）？

Magix

foxea 发表于 2017-1-11 07:54
任何一个对争议性话题（controversial topic ）的学术评论 （review），如果只列了一方的观点，那作者肯定 ...

没错, 所谓的数据都是为了观点服务的.