News for hackers. Stuff that matters.

A comment on Bayesian Analysis of Racial Bias in Police Shootings in the United States. Parent comment

Re:Bayesian Analysis of Racial Bias in Police Shootings in t... by bduerst on Friday July 08, 2016 @04:36PM (#12058465)
> African Americans are significantly more likely to be involved in all crimes - according to arrest and conviction data.

And according to studies in criminal behavior, African Americans are not any more likely to be committing crimes than other races. They're just more likely to be arrested by police and convicted for behavior than other races. If you don't believe me, you can look into marijuana use and possession arrests between the different races.

[ Parent | on HN ]
Re:Bayesian Analysis of Racial Bias in Police Shootings in t... by michaelbuddy on Friday July 08, 2016 @08:42PM (#12059531)
You're Wrong. And murder rates tell you that. MJ posession and murder are two different things. Per capita, blacks murder more than whites. AND in other countries, blacks also commit more than their "fair share" of crime including murders. Fortunately the United States still keeps the data on this. Canada is so bad now they purposely scrubbed racial information.
[ Parent | on HN ]
Re:Bayesian Analysis of Racial Bias in Police Shootings in t... by bduerst on Saturday July 09, 2016 @01:34PM (#12062386)
You're wrong, and it's because you're cherry picking yet again. Murder is highly race-on-race, and because blacks have a smaller percentage of the population, they only appear to be higher per capita [1].

Of course, since your last comment seems to hint at confirmation bias, I'm sure you'll dismiss any real statistical analysis as "scrubbing".

[1] http://qz.com/556988/here-are-four-charts-on-race-and-murder...

[ Parent | on HN ]