Sniper shoots 10 Dallas Cops, kills 3

[Sua Sponte]

https://www.washingtonpost.com/news/won ... drivers-2/

This is not quite what I stated but the quickest reference I can find. The actual study was cited a fair bit during the Ferguson thing.

Thanks, Spells. Good article that. What's interesting is that one of the key take aways of the article is that despite who is searched more often than whom, searching is the exception in traffic stops. Probably something that needs more emphasis.

Something that bugs me about these analyses on either side of any issue is the people who subsequently use the results to make some point or another failing to understand that statistics tell you much in general and very little, if anything at all, in particular. I can run an extensive survivability test on a set of devices, extract good statistics about the rate of failure and, well done, the confidence I should have in the result. If the result is 10% failure in a period of time all I can say from such a study is that of 10 parts on my desk, 1 part will typically be dead in 1000hrs of normal use and, with 90% confidence interval, that if I had 10 sets of 10 parts, that would be true 9 times out of 10. Getting to a 90% confidence interval is hard. In some of those sets of 10 parts, there will be 0 failures and some will have 2 or more possibly. All I can say is that 9 time out of 10, of 10 parts, on average, 1 part will fail in 1000hrs. That information alone tells me uniquely zero about which part in the 10 will fail and, especially, why it will fail.

The latter case is the most the salient point. A study that statistically says more black persons will be subjected to search than white persons does not tell me it's because of racism. As previously mentioned, a similar approach using gender as the dependent variable, will show men being much more frequently stopped, searched, arrested and having had force used against them. I don't know one person of any political persuasion who would argue that is solely or uniquely because of sexism.

In this article, it would be worth knowing what guidelines police use in conducting searches and how often the officer requests a search but is turned down by the person stopped. It's one thing to speed, going 10 over on a highway, and another to be going 25 over in school zone. I'm curious how much type and degree of offense, prior criminal history, lapsed license or tags, rapidity in which the person pulled over after the blues went on, time of day and location impacts the rate of search. I'd guess other less difficult to quantify factors such as the car's appearance and, especially, attitude of the driver would play critical roles as well. I'd be be very surprised if race wasn't still a key factor but equally surprised if these other variables didn't play at least an equal role.

[Blaidd Drwg]

The latter case is the most the salient point. A study that statistically says more black persons will be subjected to search than white persons does not tell me it's because of racism. As previously mentioned, a similar approach using gender as the dependent variable, will show men being much more frequently stopped, searched, arrested and having had force used against them. I don't know one person of any political persuasion who would argue that is solely or uniquely because of sexism.

It would tell as a population that Race is an independent and notable factor. Whether you call this "racism" or not is not at issue. What's at issue is that race of a population is a significant factor in the treatment of an individual. Just as the difference between Male and Female rates of stops and search has a hell of a lot to do with sex, whether or not you call it sexism.

It may not be broadly acceptable to say that for instance, Blacks have a different cultural response to the police than whites as a population, or Poor Whites have a different response as population. But it would be pretty stupid to balk at the notion that men get more police scrutiny than women, not because Police are sexist, but because male sex is a more dangerous population....which means fuck all to to the guy who gets pulled over and treated poorly where if he were a woman he'd draw hardly any attention.

[Sua Sponte]

Exactly. Those points were made. The statistics don't tell you that it's entirely race-based, they tell you it plays a role but, then again, so does gender. Then so do likely other factors IF they are included in the analysis. So people who read such statistics and conclude that the entire discrepancy between white and black people is due to race, and anybody who takes issue with that conclusion is of necessity a racist, as the statistics have "proved" racism, can only do so if they also conclude that the entire discrepancy between men and women is due to sexism. If one feels the necessity to insist that part of it must be due to differences between men and women, and insist that's not sexism, just fact, then they can't conclude at the same time we need only consider race when determining the causation of other outcomes.

If you bring in the fact that a considerable contributor to the discrepancy between men and women does indeed have to do with men, as a group, being more violent and criminally inclined, a fact *not found in the statistics* but otherwise divined as a way of explaining the statistics, then you can't state that all discrepancies between black and whites are race-based *based on the statistics only* either. Yet people insist that statistical discrepancies can uniquely be explained by racism or sexism if it supports their predilection.

Again, I'd be be very very surprised if race wasn't still a key or dominant factor but equally surprised if these other variables didn't play at least a critical role. Then again, when other factors are studied, I have little doubt that you would find that some part of the discrepancy between the documented violence perpetrated by men vs women is not solely due to sex differences but other factors to include prejudices that assume men guilty and women innocent under similar circumstances play some role in the gender discrepancy.

[Blaidd Drwg]

If you bring in the fact that a considerable contributor to the discrepancy between men and women does indeed have to do with men, as a group, being more violent and criminally inclined, a fact *not found in the statistics* but otherwise divined as a way of explaining the statistics, then you can't state that all discrepancies between black and whites are race-based *based on the statistics only* either. Yet people insist that statistical discrepancies can uniquely be explained by racism or sexism if it supports their predilection.

Again, I'd be be very very surprised if race wasn't still a key or dominant factor but equally surprised if these other variables didn't play at least a critical role. Then again, when other factors are studied, I have little doubt that you would find that some part of the discrepancy between the documented violence perpetrated by men vs women is not solely due to sex differences but other factors to include prejudices that assume men guilty and women innocent under similar circumstances play some role in the gender discrepancy.

Are men and women equally violently and criminally inclined based on any information? Hell yeah.

discrepancy between men and women does indeed have to do with men, as a group, being more violent and criminally inclined, a fact *not found in the statistics*

Do you have evidence to the contrary? or just an absence of one type of evidence.

Yet people insist that statistical discrepancies can uniquely be explained by racism or sexism if it supports their predilection.

For the man on the street, you might well call it a predilection...just as your predilection of considering factors such as the physical state of the automobile or time of day as being salient factors. What's in question is whether that's a logical conclusion rather than a one completely supported by statistics which have not been gathered. If you rephrase that to....

Yet people insist that statistical discrepancies can uniquely be explained by racism or sexism if it supports their Hypotheses.

Well now you have something. Because certain blank spots in the data are more or less telling about a situation than others. For instance, it may be more rational to look at the time of day of a stop as important as opposed to income level, less important because it's a fact not in evidence during the time of the stop. Both of these are factors to consider but some are more readily dropped to second tier priority than others.

[Sua Sponte]

The primary idea is discerning what statistics can tell us as a fundamental mathematical venture, regardless of specific application, what statistics can and cannot tell us based upon the construct of the study and analysis, and what statistics can or cannot tell us about causal mechanisms. What worries me about these articles, and I do like the one from The Post, is the ensuing arguments about what can actually be concluded. One cannot, from the study as conducted or presented, conclude that the discrepancies between blacks and whites in traffic stop-based searches are due to racism, if reasonable alternative explanations exist, such as type of offense, time of day, priors, maybe even unpaid parking tickets. Those might be procedural triggers for the police more often true of some groups than others. Condition or age of the car as a reason for search wasn't stated as a legitimate reason for search but rather as a possible form of prejudice but not a race related one. All similar to your argument elsewhere of race as a proxy for class and poverty.

A common parable in statistics courses is the study showing longer armed children are statistically better readers. The numbers really do fully support that assertion. All sorts of reasons are posited by the students. In the end, it's true because older kids are typically better readers than younger and are also more physically developed. The physical attribute is a marker of age, not a causal mechanism for reading ability. Simple minded but nowhere do the statistics even hint at longer arms being the driver of reading ability. There are people who would hear this story and carry to their grave the belief that if they could stretch their kid's arms, they'd have the next Shakespeare.

[Blaidd Drwg]

The primary idea is discerning what statistics can tell us as a fundamental mathematical venture, regardless of specific application, what statistics can and cannot tell us based upon the construct of the study and analysis, and what statistics can or cannot tell us about causal mechanisms. What worries me about these articles, and I do like the one from The Post, is the ensuing arguments about what can actually be concluded. One cannot, from the study as conducted or presented, conclude that the discrepancies between blacks and whites in traffic stop-based searches are due to racism, if reasonable alternative explanations exist, such as type of offense, time of day, priors, maybe even unpaid parking tickets. Those might be procedural triggers for the police more often true of some groups than others. Condition or age of the car as a reason for search wasn't stated as a legitimate reason for search but rather as a possible form of prejudice but not a race related one. All similar to your argument elsewhere of race as a proxy for class and poverty.

I understand where you're coming from but I disagree that the "primary idea" is discerning what statistic can tell us. i don;t think that's the primary idea at all.

What's at question...in the larger sense, is what do we know about the discrepancies between the ways blacks and whites are treated and what do we DO about that. Statistical analysis is a deep cutting, but narrowly applicable tool. We should use it as well as every other tool of rational thought to figure out why those differences exist. I have my theory (race is less relevant than class) but I wouldn't limit the scope of inquiry to statistical analysis for the very reasons you're describing above...at the end of the day, simple observation in those communities erase any doubt that blacks and whites DO receive different treatment be it based on their cultural postures, income levels, level of education etc. This is the important part...now understanding the breakdown of the individual influences is critical but not as critical and beginning with an understanding that the difference exists. I think all sides of the "debate" are too quick to dismiss one the most easily agreed upon findings of fact and jump right to attempting to debunk the stories people tell each other about why.

[TerryB]

mediatakeout pretty much debunks all of BD's arguments.

[Blaidd Drwg]

mediatakeout pretty much debunks all of BD's arguments.

That's racist.