Baltimore riots

[Danzig]

Originally Posted by Rupert Pupkin

I will repeat what I just said. "The article says that the FBI data on police shooting by race is incomplete. But you don't need 100% complete data to do an accurate analysis. For example, when polls are done, the pollsters don't poll 100% of the population. They may only poll 1-2% of the population and this will yield very good data, as long as there was a true random sample."

If you can get some pretty accurate information from a poll that only polls 1% of a population, then I think the numbers from the FBI (which are probably 80-90% complete) are probably pretty accurate. You don't need exact numbers to analyze data. If you have ballpark numbers, you are going to draw the same conclusions

you think? based on what exactly?






Those internal figures show at least 1,800 police killings in those 105 departments between 2007 and 2012, about 45% more than the FBI’s tally for justifiable homicides in those departments’ jurisdictions, which was 1,242, according to the Journal’s analysis.
The full national scope of the underreporting can’t be quantified. In the period analyzed by the Journal, 753 police entities reported about 2,400 killings by police. The large majority of the nation’s roughly 18,000 law-enforcement agencies didn’t report any.

http://www.wsj.com/articles/hundreds...ics-1417577504

and then the article you posted has this in it:

His results, posted last week on his blog Cop in the Hood, arrived with several caveats, notably that 25 percent of the website’s data, which is drawn largely from news reports, failed to show the race of the person killed.

you're wanting this bs study with god knows what numbers from a site, where one quarter of the stats included don't even know the race of the person shot, to prove something??

if you wish to believe what the person in the article you cited tries to conclude, go for it.
but good luck getting anyone to read it and glean anything useful from it.