EPTUS Politics Mega-thread - Page 4277
| Forum Index > Closed |
Read the rules in the OP before posting, please. In order to ensure that this thread continues to meet TL standards and follows the proper guidelines, we will be enforcing the rules in the OP more strictly. Be sure to give them a re-read to refresh your memory! The vast majority of you are contributing in a healthy way, keep it up! NOTE: When providing a source, explain why you feel it is relevant and what purpose it adds to the discussion if it's not obvious. Also take note that unsubstantiated tweets/posts meant only to rekindle old arguments can result in a mod action. | ||
|
Belisarius
Australia6233 Posts
| ||
|
mahrgell
Germany3943 Posts
In the US the people demand the police to use bodycams, so they can document the misbehaviour and unjustified actions of the police. And the police does its best to prevent it or undermine it. In Germany the police demands bodycams, because they want to document the misbehaviour by the people and to justify their own actions. And the public does its best to prevent it on claims of privacy rights, surveillance etc. | ||
|
zulu_nation8
China26351 Posts
On July 14 2016 11:10 Plansix wrote: I find it amusing that you want to post data and talk about it, but the discussion of its accuracy and quality is off limits. Do you know when data's accuracy or quality matters? 1.) When it affects the results of the study. Of which, no one's had the balls to argue. I'd love it if you did though. I'll respond so comprehensively you'll fucking love it. 2) When there's better data Oh wait there isn't. lol! Maybe you guys should respond do the results rather than talk shit about the data? Yea? | ||
|
GGTeMpLaR
United States7226 Posts
I'm fairly certain it will hurt him if he argues against the use of police body-cams for the sake of defending the police in the debates against Clinton. Granted I don't know where Trump stands on the issue, it would be one of the few cases I agreed with Clinton over Trump if he fought against it. Bodycams protect both the police and civilians they are responsible for. | ||
|
ticklishmusic
United States15977 Posts
| ||
|
thePunGun
598 Posts
| ||
|
zulu_nation8
China26351 Posts
On July 14 2016 11:16 ticklishmusic wrote: why would you bother analyzing data that is complete shit? it's like using rotten meat to make dinner. garbage in garbage out. yea bro no one should study whether cops or racist cuz they don't provide their own data. Until they do so, we should refrain from all such studies. | ||
|
Aquanim
Australia2849 Posts
1) that many are ill-trained and make many mistakes, and that this is concealed from the public 2) that they are (on average) biased against people of certain ethnicities While there are some links between the two issues, I do not see why they cannot be considered separately. In particular, I do not see why wishing to talk about one in particular would be construed as inherently dismissing the other. | ||
|
Plansix
United States60190 Posts
On July 14 2016 11:13 Belisarius wrote: I mean the study's obviously got issues, but p6 acting like a child while refusing to make an actual point in half a dozen posts is not helping the discussion. I have admitted I was incorrect when I posted information that was inaccurate, flawed, or did not support my argument. I only ask that other posters do the same. The study posted is DEEPLY flawed and Zulu acts like a child if anyone points out the data is shit. | ||
|
GreenHorizons
United States24139 Posts
On July 14 2016 11:16 ticklishmusic wrote: why would you bother analyzing data that is complete shit? it's like using rotten meat to make dinner. garbage in garbage out. Mostly because data that's better than "complete shit" is actively suppressed by the same interests that demand they see data before acting. Better shit data, than shit complaints about there not being data while ignoring why. On July 14 2016 11:19 Aquanim wrote: The way I see it, there are two problems being discussed here with respect to the police of the United States: 1) that many are ill-trained and make many mistakes, and that this is concealed from the public 2) that they are (on average) biased against people of certain ethnicities While there are some links between the two issues, I do not see why they cannot be considered separately. In particular, I do not see why wishing to talk about one in particular would be construed as inherently dismissing the other. Mostly because 1 generally only comes up to distract from 2. Notice the complaints of poor policing were ignored/actively denied previously. It's only because 2 was shown in stark relief that 1 became a more palatable topic for those who spent the previous decades denying both. | ||
|
Ghostcom
Denmark4783 Posts
On July 14 2016 11:07 zulu_nation8 wrote: you guys are welcome to collect your own data, cross-reference with other databases, or even...make an argument on how the data's flaws would impact the study's results, but lol like anyone would actually do that instead of just talking shit right? like lolol Taking the data collection into account the study you linked is mared by such serious limitations that it's publication is baffling. The only question it answers is what are the likelihood that someone enters a given incident into their dataset provided the media reported said incident, that the report can be found on google, and that the report was detailed enough to include the covariates. This is just the must superficial of scrutiny towards their datacollection - if we look at what ticklish linked, oh boy it only goes downhill from here. The amount of missing data is ridiculous (and in all likelihood not non-discriminatory). The further limitations of the study such as complete lack of important confounders (e.g. amount of encounters) (which they partly mention themselves) raises severe concerns about the seriousness of anyone wanting to use this data for anything. Unlike what GH thinks, no one is arguing about WHY the data is missing - people are pointing out that the current data is shit, including this study. Further people have argued that some of the rigorous studies conclude (such as the "lab" experiment) point towards racism playing no role in shooting. In conclusion, data quality is shit (this is, unlike what GH believes, not a reason for relying on anecdotal evidence). We should abstain from conclusions such as blacks are disproportional getting shot until we have better data as the best quality data we currently posses (e.g. lab experiment) points towards it not being so. EDIT: Actual data is missing because the US is scared of databases, people are scared of taking responsibility (including the police), databases are costly (someone is going to have to pay), and a million other reasons - all of dubious character. To pretend it is missing so a quiet genocide on a minority can be carried out is outright nefarious. | ||
|
Aquanim
Australia2849 Posts
On July 14 2016 11:20 GreenHorizons wrote: Mostly because data that's better than "complete shit" is actively suppressed by the same interests that demand they see data before acting. Better shit data, than shit complaints about there not being data while ignoring why. The problem is that *regardless of the accessibility of better information*, drawing and presenting conclusions from data which is inherently biased in a "desirable" direction to your argument is at best unhelpful and at worst actively harmful to your cause. | ||
|
zulu_nation8
China26351 Posts
On July 14 2016 11:01 ticklishmusic wrote: i mean he's not the one who posted the harvard study after it had been discussed for 10+ pages i did a quick scrub on the dataset. it seems like an ambitious project, and i would even say its a worthy one. however the data quality is pretty shitty because the collection method has some problems even though there's a rudimentary protocol. it's also very incomplete. out of about 2100 entries, 800+ list ethnicity as unknown. heck, 250+ of them dont have a gender OR ethnicity listed. that means close to half the dataset is junk. ![[image loading]](http://i.imgur.com/h7GQpCX.png) Do you think the researchers are rolling dice to guess the ethnicity of the 800+ then adding it to the dataset? | ||
|
Ghostcom
Denmark4783 Posts
On July 14 2016 11:23 zulu_nation8 wrote: this is the only thing I've seen: Do you think the researchers are rolling dice to guess the ethnicity of the 800+ then adding it to the dataset? Do you think those 800+ who are missing crucial variables are non-discriminatory (in the statistical sense) cases? | ||
|
ticklishmusic
United States15977 Posts
| ||
|
zulu_nation8
China26351 Posts
On July 14 2016 11:23 Ghostcom wrote: Taking the data collection into account the study you linked is mared by such serious limitations that it's publication is baffling. The only question it answers is what are the likelihood that someone enters a given incident into their dataset provided the media reported said incident, that the report can be found on google, and that the report was detailed enough to include the covariates. This is just the must superficial of scrutiny towards their datacollection - if we look at what ticklish linked, oh boy it only goes downhill from here. The amount of missing data is ridiculous (and in all likelihood not non-discriminatory). The further limitations of the study such as complete lack of important confounders (e.g. amount of encounters) (which they partly mention themselves) raises severe concerns about the seriousness of anyone wanting to use this data for anything. Unlike what GH thinks, no one is arguing about WHY the data is missing - people are pointing out that the current data is shit, including this study. Further people have argued that some of the rigorous studies conclude (such as the "lab" experiment) point towards racism playing no role in shooting. In conclusion, data quality is shit (this is, unlike what GH believes, not a reason for relying on anecdotal evidence). We should abstain from conclusions such as blacks are disproportional getting shot until we have better data as the best quality data we currently posses (e.g. lab experiment) points towards it not being so. Best quality data are The Guardian and Washington Post's own databases. They seem to say that blacks are not being disproportionately shot, but it depends on what you compare the killing data to. Unfortunately, not everyone is as cautious with their opinion as you. People want answers and are protesting on impressions that may or may not be right. | ||
|
GreenHorizons
United States24139 Posts
On July 14 2016 11:23 Aquanim wrote: The problem is that *regardless of the accessibility of better information*, drawing and presenting conclusions from data which is inherently biased in a "desirable" direction to your argument is at best unhelpful and at worst actively harmful to your cause. Again so we end up where black people are screaming at the tops of their lungs that it's real and it's happening, then someone saying "well we can't believe them without data", then someone says "here's the data", then they respond "well it's incomplete, we can't draw any conclusions", then they are presented with why we don't have the data, then they say, "well we can't draw conclusions without the data", as if they don't know why the data is incomplete. No one should be able to comment on police shooting data unless they are willing to admit why the data we have is incomplete in the first place, otherwise it's a totally disingenuous discussion from step 1. | ||
|
Aquanim
Australia2849 Posts
On July 14 2016 11:20 GreenHorizons wrote:... Mostly because 1 generally only comes up to distract from 2. Notice the complaints of poor policing were ignored/actively denied previously. It's only because 2 was shown in stark relief that 1 became a more palatable topic for those who spent the previous decades denying both. Well, I'm certainly not familiar with the history of this issue going back decades. Nevertheless, I think that *in the present conversation and context* BOTH of these problems deserve to be discussed without anybody trying to obfuscate one problem with the other. "But they did it first" is not a recipe for solving problems. | ||
|
Ghostcom
Denmark4783 Posts
On July 14 2016 11:27 ticklishmusic wrote: if half the data points found like this are worthless then maybe the collection methodology needs to be revised. who knows what gremlins are lurking in shit-my-data-sucks land? May I use this for my presentation on data-validation and why it is important? | ||
|
zulu_nation8
China26351 Posts
On July 14 2016 11:26 Ghostcom wrote: Do you think those 800+ who are missing crucial variables are non-discriminatory (in the statistical sense) cases? Don't know, I'm trying to read the tables that factor in crime rates by race, then will decide. | ||
| ||
