Discrimatory statistics, is it really discrimination?

For the serious discussion of weighty matters and worldly issues. No off-topic posts allowed.

Moderators: Azrael, Moderators General, Prelates

User avatar
PandasOnProzak
Posts: 19
Joined: Sun Dec 21, 2008 11:25 pm UTC
Location: The Dragon Dreams Drearily of Drapes

Discrimatory statistics, is it really discrimination?

Postby PandasOnProzak » Sat May 16, 2009 2:00 am UTC

I have been thinking about this for a while and was wondering what SB's take on it was. Are statistics based on race/gender/etc that show a certain group to be "worse" truly discriminatory or does it give scientific backing to stereotypes/discrimination. A good example of this is crime rates in the united states, a plethora of surveys show African American's to contribute disproportionately to crime. Whenever i hear this mentioned it's always from the perspective that the Justice system is racist and not that more African American's are brought up in the ghetto/crime culture. I believe that statistics don't lie, while there might be some racist cops/judges our system as a whole seems to be unbiased. Shouldn't these results be used to identify a cultural problem and help solve it?

Yet whenever someone mentions this they get slammed by Civil rights groups and the media for racism. A good example of this is Obama's comment on black males not raising their children, you would think this would be legitimate coming from an African American who worked to benefit the community, but no :? .

so this may have seemed like a small rant, and it is, but i feel that this is an interesting topic and hope i can spark some discussion.

EDIT: found a better way to explain my thoughts.
1. these statistics expose discrimination
2. these statistics validate discrimination
3. they are worthless without case specific data
4. they are cannot be used to come to a conclusion.

i agree with number three but it seems that in most cases this data supports discrimination and that bothers me.
Last edited by PandasOnProzak on Sat May 16, 2009 5:55 pm UTC, edited 2 times in total.
҉

quartercirclefish
Posts: 60
Joined: Sun Nov 16, 2008 8:11 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby quartercirclefish » Sat May 16, 2009 7:11 am UTC

There is probably a negative feeling towards them because of the concept of controlling for socio-economic demographics. Ie compare the IQ of caucasians and african-americans while controlling for socio-economic and other factors (schooling, upbringing etc) and you might find that the caucasians score better than the african-americans.
This doesn't tell you anything. How can you distinguish between it being an inherent racial characteristic or that you just missed out some controlling factor or your adjustment for controlling factors wasn't 100%.
It just gives fuel for racists

sje46
Posts: 4730
Joined: Wed May 14, 2008 4:41 am UTC
Location: New Hampshire

Re: Discrimatory statistics, is it really discrimination?

Postby sje46 » Sat May 16, 2009 9:07 am UTC

I wonder about this too, a lot. But one thing to keep in mind is that it is basically impossible to control for everything. If a black person is raised in the same neighborhood and his family had the same income, etc etc, as a white kid, and everything is controlled for, you still can't control his view of the world, his experiences as a black person, etc. I think there can be a little learned helplessness in there. Or perhaps some feel that they don't have to try as hard, for some reason. Or perhaps the tests favor certain races over others.
And sometimes I wonder if perhaps there is an inherent difference in the brains of different races. Have we proved that it isn't true? No, but just talking to different races than your own shows that their minds work the same as ours, and if there is a difference, it would be negligible. And even if it were true, would that justify treating them inequally? I don't think so.
I know that there is more variation within race than between race, but that doesn't make their facial features look the same as other races, so why should the brain be different? For that matter (and this is a little off-topic), do internal organs differ between races as well?
I would like to hear from a neurologist or anthropologist or whoever would study this to hear their thoughts on this. I highly doubt it's biological, however. I'm 99% sure that the reason blacks don't do as well on IQ tests is entirely sociological.
But to answer the OP, statistics aren't discriminatory. People can choose to use statistics to justify discrimination, if they are right or wrong ("Well, statistics say that muslims are most likely to hijack planes, so we'll search you") or prejudice, which is almost always wrong, since pretty much not all of anything will have a certain quality.
General_Norris: Taking pride in your nation is taking pride in the division of humanity.
Pirate.Bondage: Let's get married. Right now.

General_Norris
Posts: 1399
Joined: Fri Apr 17, 2009 12:10 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby General_Norris » Sat May 16, 2009 9:48 am UTC

People don't understand simple logic most of the time. They NEVER understand nor probability nor statistics.

Example => Lottery

In my wonderful country, Spain, we are chockedwith "positive discrimination" as in "Men must be X height to be part of the firemen, Women must be X-Y because there are not as many women and that's bad and sexist"

On the same topic. People always say that women are worse paid than men. Well considering that men have better jobs than women on average it's logical. That doesn't mean there's some ilegal payments around.

All those "arguments" fail because they rely on probability. That's is, women have worse jobs than men as whole but if you pick a woman she may have a really good job. Spaniards as a whole love Paella. I don't. No true Spaniard? I don't think so.

Seraph
Posts: 343
Joined: Mon Jul 16, 2007 4:51 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Seraph » Sat May 16, 2009 10:31 am UTC

Statistics lie all the time. I remember when I was in high school running into a similar issue with a "racial bias" in AP tests, but when you looked at the results for individual states the bias disappeared. The real issue was that a number of southern states (with large black populations) would pay for any student to take the tests. A lot of these students were poor performers, and so dragged down the black statistics more then the other racial groups.

Another example, the book "Innumeracy" by John Allen Paulos discusses a discrimination case brought against a large university. When you looked at the college as a whole, a much greater percentage of male applicants were being accepted when compared to female applicants. However, when the school administrators looked at the individual departments to figure out which ones were the problem, they were all accepting a higher percentage of female applicants. It turns out the issue was that the departments with high admission rates tended to have lots of male applicants, and the departments with low admission rates tended to have lots of female applicants.

User avatar
Indon
Posts: 4433
Joined: Thu Oct 18, 2007 5:21 pm UTC
Location: Alabama :(
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby Indon » Sat May 16, 2009 3:09 pm UTC

Seraph wrote:It turns out the issue was that the departments with high admission rates tended to have lots of male applicants, and the departments with low admission rates tended to have lots of female applicants.


And to give further idea of how unreliable statistics are to make judgments, this could still be a result of systemic sexism, if for instance it could be demonstrated that a field's admission rates dropped when the number of female applicants for the field increased.

Statistics, particularly demographic statistics in a social system, tend to be clusterfuckingly complex.
So, I like talking. So if you want to talk about something with me, feel free to send me a PM.

My blog, now rarely updated.

Image

User avatar
PandasOnProzak
Posts: 19
Joined: Sun Dec 21, 2008 11:25 pm UTC
Location: The Dragon Dreams Drearily of Drapes

Re: Discrimatory statistics, is it really discrimination?

Postby PandasOnProzak » Sat May 16, 2009 4:05 pm UTC

what always irks me is discrimination cases against private schools/universities based on acceptance statistics, it's gotten to the point where diversity boards exist to prevent such suits.
҉

Captain_Thunder
Posts: 83
Joined: Tue Mar 27, 2007 9:18 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Captain_Thunder » Sat May 16, 2009 8:57 pm UTC

Seraph wrote:Statistics lie all the time. I remember when I was in high school running into a similar issue with a "racial bias" in AP tests, but when you looked at the results for individual states the bias disappeared. The real issue was that a number of southern states (with large black populations) would pay for any student to take the tests. A lot of these students were poor performers, and so dragged down the black statistics more then the other racial groups.

Another example, the book "Innumeracy" by John Allen Paulos discusses a discrimination case brought against a large university. When you looked at the college as a whole, a much greater percentage of male applicants were being accepted when compared to female applicants. However, when the school administrators looked at the individual departments to figure out which ones were the problem, they were all accepting a higher percentage of female applicants. It turns out the issue was that the departments with high admission rates tended to have lots of male applicants, and the departments with low admission rates tended to have lots of female applicants.


But were these statistics really lying? These explanations only become interesting and relevant when you try to look at the cause of the statistics, and your preconceived notions with regards to the causes turn out to be false. The correlations stated were still true, regardless of the underlying reasons.

It's important to realize that just because someone points out statistics that say blacks are involved in more crime doesn't mean they are saying that blacks are more violent/less intelligent/whatever; they are only saying, and only can be saying, that a correlation exists. It's their own fault if they want to try and establish causality.

User avatar
Indon
Posts: 4433
Joined: Thu Oct 18, 2007 5:21 pm UTC
Location: Alabama :(
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby Indon » Sat May 16, 2009 11:41 pm UTC

Captain_Thunder wrote:It's important to realize that just because someone points out statistics that say blacks are involved in more crime doesn't mean they are saying that blacks are more violent/less intelligent/whatever; they are only saying, and only can be saying, that a correlation exists. It's their own fault if they want to try and establish causality.


For rational beings, correlation does not imply causation.

Humans are not rational.
So, I like talking. So if you want to talk about something with me, feel free to send me a PM.

My blog, now rarely updated.

Image

btilly
Posts: 1877
Joined: Tue Nov 06, 2007 7:08 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby btilly » Sun May 17, 2009 9:33 pm UTC

Statistics are just a tool. On the one hand without statistics, you can't know how you're doing, or where you need to improve. On the other hand statistics can be used to lead people to bad conclusions. Personally I'm of the opinion that if we are so touchy about a subject that we're not even willing to look at the facts, that is a problem. So I'd prefer to have the statistics, even though they can be dangerous.

One of my favorite examples of an interesting way to use statistics involves the correlation between SAT scores, grades, and gender. Statistically men outperform women on timed tests. Statistically women outperform men in the classroom. There is a straightforward explanation - women are more likely than men to spend time double-checking their work. This helps you in the classroom, but hurts you in a timed test. (This is not the full cause, but it is part of it.)

In the 90s MIT looked at this, and made a controversial (and soon reversed) decision. They reasoned that they were using the SAT as a predictor of classroom performance, and they wanted to pick people who . But the SAT was under-predicting how well women would do. So before showing applications to the admission committee, they adjusted SAT scores of women up until they had an unbiased predictor of classroom performance. Was this discrimination?

quartercirclefish wrote:There is probably a negative feeling towards them because of the concept of controlling for socio-economic demographics. Ie compare the IQ of caucasians and african-americans while controlling for socio-economic and other factors (schooling, upbringing etc) and you might find that the caucasians score better than the african-americans.
This doesn't tell you anything. How can you distinguish between it being an inherent racial characteristic or that you just missed out some controlling factor or your adjustment for controlling factors wasn't 100%.
It just gives fuel for racists

One obvious factor is culture. A lifetime of encounters with strangers who have expectations about what to expect from blacks and whites leaves differing impacts on blacks and whites. There is no way to eliminate this.

However there is some evidence suggesting that a more important factor is differing cultural attitudes towards child rearing between blacks and whites. You can see this from looking at cross-racial families. If one parent is black and the other is white, then genetically it shouldn't matter much which parent is black. Nor does it affect how others interact with you. But children with a black mother score lower on IQ tests than ones with a black father. Given that mothers are more involved in child rearing than fathers, this suggest that child rearing is the difference. However studies of the Flynn effect suggest that the gap is closing.
Some of us exist to find out what can and can't be done.

Others exist to hold the beer.

User avatar
athelas
A Sophisticated Plagiarism Engine
Posts: 584
Joined: Sat Oct 21, 2006 2:37 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby athelas » Sat Jun 13, 2009 4:26 pm UTC

Before we tackle the question we'd better figure out what we are trying to do. Are we trying to eliminate discrimination? Or are we trying to figure out the truth behind social trends? These two goals sometimes overlap but are sometimes in opposition, and if we try to do both at once it will lead to muddled thinking and cognitive dissonance.

User avatar
Durandal
Posts: 659
Joined: Thu Nov 08, 2007 12:12 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Durandal » Sat Jun 13, 2009 5:00 pm UTC

This topic reminds me of a talking point put forth by someone - forget who. Say you are on board a train with two railcars. One car is filled with caucasians, while the other car is filled with african-americans. Statistically speaking you are more likely to be assaulted in the african-american car, so is it racist of you to move to the caucasian car?

User avatar
Lucrece
Posts: 3558
Joined: Mon Feb 25, 2008 12:01 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Lucrece » Sat Jun 13, 2009 6:36 pm UTC

Durandal wrote:This topic reminds me of a talking point put forth by someone - forget who. Say you are on board a train with two railcars. One car is filled with caucasians, while the other car is filled with african-americans. Statistically speaking you are more likely to be assaulted in the african-american car, so is it racist of you to move to the caucasian car?



Except that most literature in criminology points to crimes being intraracial, not interracial. As a white person, you will most likely be victim to a criminal of your own race than a criminal of another race/ethnicity.*

What you could ask in such a scenario is whether a sense of cultural adhesion might lead the majority individuals to have less problem prioritizing the minority race for victimization.

*http://www.allacademic.com/meta/p_mla_apa_research_citation/1/2/6/1/0/p126101_index.html
Belial wrote:That's charming, Nancy, but all I hear when you talk is a bunch of yippy dog sounds.

sje46
Posts: 4730
Joined: Wed May 14, 2008 4:41 am UTC
Location: New Hampshire

Re: Discrimatory statistics, is it really discrimination?

Postby sje46 » Sat Jun 13, 2009 7:39 pm UTC

Lucrece wrote:
Durandal wrote:This topic reminds me of a talking point put forth by someone - forget who. Say you are on board a train with two railcars. One car is filled with caucasians, while the other car is filled with african-americans. Statistically speaking you are more likely to be assaulted in the african-american car, so is it racist of you to move to the caucasian car?

Except that most literature in criminology points to crimes being intraracial, not interracial. As a white person, you will most likely be victim to a criminal of your own race than a criminal of another race/ethnicity.*

What you could ask in such a scenario is whether a sense of cultural adhesion might lead the majority individuals to have less problem prioritizing the minority race for victimization.

*http://www.allacademic.com/meta/p_mla_apa_research_citation/1/2/6/1/0/p126101_index.html

SB doesn't care what you find boring. -Az

Anyway, even if that research is true (I doubt it is, really, and I can only see the abstract), that would just be plain fortuitous. What if it isn't true? What if this moral dilemma wasn't avoided by luck?
But again, I really doubt that it is safer to be some average white guy living in an impoverished neighborhood than an average black guy. Perhaps it has to do with white people living in fear of constantly being jumped, and the black people more likely to join gangs, to sell drugs, etc, if they live in the same neighborhood. There are a lot of factors in this. I do think, persoanlly, as a strange white person in Compton or Harlem or whatever is supposed to be bad, that I would be more likely targeted than the average black stranger. Why? Because I am more likely to be rich, less street-savvy, and because of plain black-pantherist racism. Of course, I'm sure demeaner and muscle size and dress etc would be important factors over whether you'd get jumped or not. But this apples to blacks as well as whites or any race.
This makes it sound like I think that anyone who walks in a bad neighborhood will definitely get stabbed. To be honest, I think people probably exaggerrate how dangerous neighborhoods are, but I wouldn't know.
General_Norris: Taking pride in your nation is taking pride in the division of humanity.
Pirate.Bondage: Let's get married. Right now.

User avatar
Clumpy
Posts: 1883
Joined: Thu Jul 19, 2007 4:48 am UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby Clumpy » Sun Jun 14, 2009 1:05 am UTC

It's interesting to note how many of our racial responses are the result of cultural conditioning. I live in a community and state that isn't very diverse. I've often wondered how I would be able to analyze others' responses to me were I black. For example, I, as a caucasian, will often say high to people on the street and get no response or just a vague nod and a look away. That's fine - many people are just nervous speaking to strangers and aren't as outgoing.

If I were black, how would I know firstly that many people act that way anyway or if it happened more often because of nervousness related to my race? (Either because of fears regarding stereotypes of African-Americans or, more likely, a fear of appearing either nervous or patronizing.)

Likewise it's my firm belief that attempting to paint pictures of a race's "superiority" in one field or another is likely to be tainted by several reasons (Relevant: the book The Bell Curve and firestorm surrounding it):

1) Consciousness and unconsciousness racism and discrimination including its associated effects (generational poverty, access to education and medicine and the like), as well as cultural biases taint things like standardized tests.

2) Vast numbers of blacks perform better than vast numbers of non-blacks on standardized tests, so even a general trend is not indicative of an individual's chances of success.

3) Given our incomplete data gathering and analyzing ability, making racial assumptions or pursuing the data from this direction will have a damaging effect on our chances of fixing the problems. (Treating the symptoms rather than the sickness.)

It's the same deal with women and tests: women are often stereotypically seen to be "better" than men at verbal tasks and men better at spatial and analytical tasks, but when viewing a distribution you'll see from the outliers that the difference is marginal and outlying women and men will often outperform the opposite sex in the areas society sees them as weak. So saying "women are worse at [x]" is counterproductive because it literally gives no information about a particular individual's gifts, talents and knowledge.

User avatar
Lucrece
Posts: 3558
Joined: Mon Feb 25, 2008 12:01 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Lucrece » Sun Jun 14, 2009 2:01 am UTC

Being a dick is not the proper SB response to a poorly supported position. -Az
Belial wrote:That's charming, Nancy, but all I hear when you talk is a bunch of yippy dog sounds.

sje46
Posts: 4730
Joined: Wed May 14, 2008 4:41 am UTC
Location: New Hampshire

Re: Discrimatory statistics, is it really discrimination?

Postby sje46 » Sun Jun 14, 2009 4:39 am UTC

I doubt that we can trust your study in this situation, because there are a lot of factors that simply are not discussed in the abstract. We do not know if the white people are extra careful and don't go out in night, etc. We do not know if the black people tend to engage in high risk activities more, like selling drugs, joining gangs, etc. I think that all this makes more intuitive sense, you know? Because white people tend to have more money, tend to be "out of place" if they are in a black neighborhood, less street savvy, etc. I'm not saying that the black man are out for white blood. I wouldn't be surprised at all if they are only five percent more likely to mug some average white dude than some average black dude given equal chance.
But regardless. Ignore the study anyway. Lets say that it is true that black people are more dangerous, statistically, to be around. We have this moral dilemma. How to we solve it?

Edited. I suggest we refrain from continuing the removed personality conflict. -Az
General_Norris: Taking pride in your nation is taking pride in the division of humanity.
Pirate.Bondage: Let's get married. Right now.

User avatar
athelas
A Sophisticated Plagiarism Engine
Posts: 584
Joined: Sat Oct 21, 2006 2:37 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby athelas » Sun Jun 14, 2009 6:55 pm UTC

Even more importantly, your study says that blacks do not specifically pick on whites. It does not say that black-on-white violence is less likely than white-on-black violence.

Just like the OP says, we need to think carefully about what statistics actually say, and not just use them to confirm our previous biases. This is including the bias towards blindly imagining no differences between the races.

User avatar
Lucrece
Posts: 3558
Joined: Mon Feb 25, 2008 12:01 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Lucrece » Sun Jun 14, 2009 9:12 pm UTC

athelas wrote:Even more importantly, your study says that blacks do not specifically pick on whites. It does not say that black-on-white violence is less likely than white-on-black violence.

Just like the OP says, we need to think carefully about what statistics actually say, and not just use them to confirm our previous biases. This is including the bias towards blindly imagining no differences between the races.



But who's saying you may not find cultural differences? Is there a difference of backgrounds that most likely will present some difference in behavior? Yes. Does that mean that we know what the actual differences are, and to what degree? No, and speculation informed by socially imposed stereotypes is most certainly not the way to go about it.

Furthermore, the study I linked was used to deny the notion that a white man would be safer among whites than among blacks, not which race placed under a minority position would be less likely to be victimized by the majority population.
Belial wrote:That's charming, Nancy, but all I hear when you talk is a bunch of yippy dog sounds.

User avatar
setzer777
Good questions sometimes get stupid answers
Posts: 2762
Joined: Sun Nov 23, 2008 9:24 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby setzer777 » Thu Jun 18, 2009 5:06 pm UTC

quartercirclefish wrote:There is probably a negative feeling towards them because of the concept of controlling for socio-economic demographics. Ie compare the IQ of caucasians and african-americans while controlling for socio-economic and other factors (schooling, upbringing etc) and you might find that the caucasians score better than the african-americans.
This doesn't tell you anything. How can you distinguish between it being an inherent racial characteristic or that you just missed out some controlling factor or your adjustment for controlling factors wasn't 100%.
It just gives fuel for racists



Agreed. Though one issue seems to be that non-racist people are more willing to make definitive claims in the other direction (it's cultural) even though there are no completely controlled studies proving *that* either. Though this is definitely the safer position to take morally (better to assume everyone is biologically equivalent and be wrong than to imagine certain differences and be wrong), empirically it's just as incorrect as assuming that biological difference is proven.
Meaux_Pas wrote:We're here to go above and beyond.

Too infinity
of being an arsehole

tantalum
Posts: 221
Joined: Sun Jan 20, 2008 9:28 pm UTC
Location: cambridge, MA

Re: Discrimatory statistics, is it really discrimination?

Postby tantalum » Mon Jun 29, 2009 3:21 pm UTC

I would guess that the reason discriminatory statistics is so abominable (at least in the U.S.) is because individualism is so strong here. Here, everyone is a free agent and should be treated on a case-by-case basis, or so we all like to think.

Imagine China. If you were a chinese citizen and the government announces that it is cracking down on some ethnic minority because it is widely known* that they are more likely to commit crimes. I would imagine most Chinese citizens**, who appreciate their government (no matter how oppressive it may be to Western sensibilities), would agree and call it a wise move.

*known in the sense that all americans know that black people are associated with higher crime rates, but may or may not be racially biased.
**to western audiences, it would seem that I'm taking a cheap shot at the Chinese people. The chinese, on the other hand, would probably nod and think that I'm on their side. I guess this only shows how different the mentality (individualism v. socialism) really is.

User avatar
Oculus Vespertilionis
Posts: 434
Joined: Thu Jun 04, 2009 7:42 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Oculus Vespertilionis » Tue Jun 30, 2009 4:56 am UTC

setzer777 wrote:(better to assume everyone is biologically equivalent and be wrong than to imagine certain differences and be wrong)

Actually, I'm not so sure. This is all off the top of my head. Think of it from the point of view of the minority in question, and your consequences in both directions.
And, to get a little bit away from the touchiness, let's talk about the Welsh and soccer.
Now, let's say that for whatever reason there's a possibility that people of Welsh descent may be worse soccer players than your average American of some other descent.
You assume (incorrectly) that the difference is cultural, and can be corrected for. You therefore set expectations for the Welsh players equal to the other players as the cultural factors are accounted for and removed, and are consistently let down when they don't make it. Individual Welsh under this scenario will come to believe it's their fault that they fail at soccer more than their peers -- after all, if they're inherently the same, clearly they're just not putting in the same effort into it. And, because you'll probably never actually be certain that you've "fully corrected" for the cultural factors, this deficiency will never be identified. Instead of working to deal with the inherent deficiency (which doesn't make them any less valuable as people, just less able to play soccer), we keep propping them up. The Welsh themselves, in this case, seem to suffer most of all.
Other case. You assume (incorrectly) that the difference is biological, and can't be helped. You therefore set lower expectations for the Welsh players, and provide them extra help and other advantages (which you assume will always be necessary). They will soon meet and exceed your lower expectations with the help, designed to deal with their inherent deficiency, as there is no actual deficiency. In this scenario your assumptions are rather quickly disproved by a rapid rise in ability; as they meet and overtake their peers, the misconception naturally disappears and the unnecessary compensation diminishes accordingly.
The fact is, in terms of consequences of assumption, it seems that a false positive belief in racial inequality may be more easily self-correcting, and less harmful in the long-term, than a false negative belief. In practice, however, belief in racial inequality is by definition racist. Racism is bad, disrespectful, and contrary to individualist thought (no sarcasm at all here; I really believe those three things), so consequences aside, racialist assumptions may not be a reasonable starting point.
Just my thoughts; probably silly. Look forward to the response though.
-OcV
You do what you can to make relationships and to respect yourself and others. Everything else is bookkeeping.

User avatar
setzer777
Good questions sometimes get stupid answers
Posts: 2762
Joined: Sun Nov 23, 2008 9:24 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby setzer777 » Tue Jun 30, 2009 1:53 pm UTC

I'd say the main issue I have with your scenario is the assumption that Welsh players, having (what is assumed to be) a racial disadvantage at soccer, would be given extra help and other advantages. More likely they just wouldn't be put on soccer teams at all. What's the team owner's incentive for putting inferior players on their team, who on average take extra investment just to reach the same level of competence most players start with?
Meaux_Pas wrote:We're here to go above and beyond.

Too infinity
of being an arsehole

User avatar
Oculus Vespertilionis
Posts: 434
Joined: Thu Jun 04, 2009 7:42 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Oculus Vespertilionis » Tue Jun 30, 2009 2:48 pm UTC

Ah. So you're saying that if we assume racial equality and cultural disadvantage, we'll be more motivated to fix the problem than if we assume racial inequality instead.
I just figured if we assumed racial inequality, we'd treat the inequality the same that we treat other disabilities -- by providing accomodations. This is where the analogy may break down.
OcV
You do what you can to make relationships and to respect yourself and others. Everything else is bookkeeping.

User avatar
andrewclunn
Posts: 135
Joined: Mon Jun 29, 2009 7:33 pm UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby andrewclunn » Tue Jun 30, 2009 4:57 pm UTC

In order to have this discussion properly I think we need a representative who isn't a white male. Would someone who is female or of another racial background please state that they are such, so that we might engage in using statistics in an attempt to prove or disprove differences between your group and ours.
I program in languages that would make your motherboard blush.

I also shave with a +2 Occam's razor.

User avatar
setzer777
Good questions sometimes get stupid answers
Posts: 2762
Joined: Sun Nov 23, 2008 9:24 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby setzer777 » Tue Jun 30, 2009 5:07 pm UTC

andrewclunn wrote:In order to have this discussion properly I think we need a representative who isn't a white male. Would someone who is female or of another racial background please state that they are such, so that we might engage in using statistics in an attempt to prove or disprove differences between your group and ours.


Eh...which part of the thread is this actually a response to? It seems like the main discussion is about the possible assumptions to make given incomplete statistical data and the empirical/ethical ramifications of those assumptions. As far as I can tell nobody is actually trying to "prove or disprove differences between [two groups]", because everyone here recognizes that we do not even remotely have enough evidence to do either.
Meaux_Pas wrote:We're here to go above and beyond.

Too infinity
of being an arsehole

User avatar
andrewclunn
Posts: 135
Joined: Mon Jun 29, 2009 7:33 pm UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby andrewclunn » Tue Jun 30, 2009 5:19 pm UTC

setzer777 wrote:As far as I can tell nobody is actually trying to "prove or disprove differences between [two groups]", because everyone here recognizes that we do not even remotely have enough evidence to do either.


I'm not so sure on this point, but I recognize that to discuss such a thing among people of similar backgrounds might lead to bias influences the outcome of such a discussion. And if it is true that we cannot currently prove or disprove such things, then to say that one believes that such statistics are or are not discrimination is simply sharing one's own biases rather than an informed position.

EDIT
I'm disagreeing with point 4: "they cannot be used to come to a conclusion."
I program in languages that would make your motherboard blush.

I also shave with a +2 Occam's razor.

User avatar
Oculus Vespertilionis
Posts: 434
Joined: Thu Jun 04, 2009 7:42 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Oculus Vespertilionis » Tue Jun 30, 2009 5:35 pm UTC

andrewclunn wrote:In order to have this discussion properly I think we need a representative who isn't a white male. Would someone who is female or of another racial background please state that they are such, so that we might engage in using statistics in an attempt to prove or disprove differences between your group and ours.


*eyebrow raise* How would this work, exactly? You get a single member of a minority group... how will that allow you to make a generalization about the group as a whole, or do any sort of statistics?
Especially since I doubt that the people of this forum represent the "normal" of any demographic group they fall into.
But maybe I'm just misunderstanding you.
OcV
You do what you can to make relationships and to respect yourself and others. Everything else is bookkeeping.

User avatar
andrewclunn
Posts: 135
Joined: Mon Jun 29, 2009 7:33 pm UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby andrewclunn » Tue Jun 30, 2009 5:56 pm UTC

Oculus Vespertilionis wrote:*eyebrow raise* How would this work, exactly? You get a single member of a minority group... how will that allow you to make a generalization about the group as a whole, or do any sort of statistics?
Especially since I doubt that the people of this forum represent the "normal" of any demographic group they fall into.
But maybe I'm just misunderstanding you.
OcV


If we cannot prove or disprove any differences because of memetics (through comparisons of religions or culture) or genetics (through comparisons of gender or race) then this whole thread is moot. I think we can all agree that memetics DOES in fact correlate to various other activities (example, likelihood to be a suicide bomber between atheists and people or certain faiths.) It's still culturally taboo to say that genetics can influence behavior, but that does not mean that it doesn't. It is my goal to have someone of another memetic or genetic group than myself identify themselves and to then attempt to prove that we can first correlatively show a difference and then that can lead to valid hypothesis about the causal justification for that difference.
I program in languages that would make your motherboard blush.

I also shave with a +2 Occam's razor.

User avatar
Oculus Vespertilionis
Posts: 434
Joined: Thu Jun 04, 2009 7:42 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Oculus Vespertilionis » Tue Jun 30, 2009 8:40 pm UTC

Isn't this called "misuse of anecdotal evidence through unsupported generalization"?
OcV
You do what you can to make relationships and to respect yourself and others. Everything else is bookkeeping.

User avatar
setzer777
Good questions sometimes get stupid answers
Posts: 2762
Joined: Sun Nov 23, 2008 9:24 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby setzer777 » Tue Jun 30, 2009 9:15 pm UTC

Oculus Vespertilionis wrote:Isn't this called "misuse of anecdotal evidence through unsupported generalization"?
OcV


Not necessarily, he could want someone of another group to identify themselves and then try to establish a difference in the groups using statistically valid studies.
Meaux_Pas wrote:We're here to go above and beyond.

Too infinity
of being an arsehole

User avatar
andrewclunn
Posts: 135
Joined: Mon Jun 29, 2009 7:33 pm UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby andrewclunn » Wed Jul 01, 2009 3:08 pm UTC

Instead it appears I'm accidentally providing evidence that the xkcd forums are populated almost entirely by white men.
I program in languages that would make your motherboard blush.

I also shave with a +2 Occam's razor.

User avatar
athelas
A Sophisticated Plagiarism Engine
Posts: 584
Joined: Sat Oct 21, 2006 2:37 am UTC

Re: Discrimatory statistics, is it really discrimination?

Postby athelas » Wed Jul 01, 2009 3:49 pm UTC

Related, and underexposed, is the theory of Rational Discrimination. (Also here)


If a concessionaire at a baseball stadium sells hot dogs for a dollar each and makes ten cents per sale, at a wage of five dollars per hour he will hire another hot-dog seller only if the new seller can sell at least fifty hot dogs per hour. If the concessionaire keeps all revenues in excess of his costs, he will naturally prefer employees who can sell more than fifty hot dogs an hour. Suppose the concessionaire notices that his highest producers are young males with little body fat who wear track-meet T-shirts. Suppose he also notices that the majority of those who failed to meet the fifty-hot-dogs-per-hour standard were overweight women. Faced with a choice between an overweight woman and a male runner, a rational employer would hire the man.

An exceptionally motivated overweight woman might outperform the average male runner, of course. Unfortunately, without requiring extensive physical examinations, the cost of which could wipe out much of the profits from hot-dog sales and make any hiring moot, the manager cannot separate exceptional overweight women from ordinary ones.

Lacking information about individuals, the concessionaire bases his decision on the average characteristics of the groups with which he has had experience. Most of his employees will be young, lean, fit, and male. Economists call this “statistical discrimination.” The employer’s workforce will look the same whether the manager discriminated fairly on the basis of real differences in productivity (no overweight woman is likely to cover as much area as a young male runner), fairly on the basis of incomplete information (the overweight woman was exceptionally fit but the manager did not know it), or unfairly on the basis of managerial taste (the employer dislikes female employees)...

Studies of woman’s wages that conclude women are discriminated against in U.S. labor markets are also misleading when they suffer from the bias caused by omitted information. The gap between male and female wages narrowed from about 40 percent in 1970 to about 24 percent in 2003. Claims that the 24 percent difference results from bigotry typically ignore the fact that as a group women are more likely to work part time, choose careers in lower-paying fields, work for government or a nonprofit, and have fewer years of labor market experience than men of the same age. These differences could all create a wage gap and may reflect choices made to accommodate family responsibilities. Researchers who adjust for standard factors such as education, experience, and line of work find no significant difference between the earnings of men and women who never married and never had a child....

Rather than reducing discrimination, many of the U.S. government policies adopted since segregation have simply changed the groups discriminated against. Along with real or implied hiring quotas, the restrictions created under affirmative action blunt the market mechanisms that make discrimination expensive. Barriers to hiring and firing make employers less likely to try out types of people with whom they have little experience. Minimum-wage laws and union wage scales keep wages higher than market wages, reducing the number of people employers wish to hire while simultaneously attracting more applicants. When this happens, bigots pay less for turning away applicants who match their productivity requirements but not their tastes.


What we should realize is that business owners have a strong incentive to disprove false stereotypes, and make use of true ones. (Think about it. If everyone else hates hiring Poles, but they work just as hard, then I can snap them up at cheap wages (because where else would they go?) and turn a huge profit.) They only way discrimination can persist in a competitive atmosphere is if the customers don't like being served by Poles.

Now what happens if Poles are lazy? Well, an owner will try to get information about each applicant's performance. If you find a hard worker, then you can hire them at a discount, just as before. If they're not, then you might hire them at lower wages or not hire them at all. Poles will draw lower wages, and this is as it should be - each one is judged on the basis of his abilities.

Now what happens if you don't let employers test out their hires, because you don't like the results that come back? (This happened recently in the Ricci case). Then suddenly the cost of hiring Poles rises, because you can't tell if you have a hard worker or (more likely) a bum on your hands. Thus instead of making matters better, this makes discrimination more unfair, because you penalize the good along with the bad.

And the solution?
4. Tim seems pretty fatalistic about statistical discrimination. He shouldn't be. Are you disturbed by the fact that the market fails to judge people as individuals? Then allow employers to write contracts to solve the problem.

Example: Some young women are 100% focused on their careers, and don't want kids. Most young women, however, do want kids, and intend to strike a balance between work and family. That balance often involves receiving expensive job training from a firm, then quiting before the firm can recoup its expenses.

Under current law, an employer isn't even allowed to ask about a female applicant's child-bearing plans. If you wanted to blow up the glass ceiling, though, you should not only allow employers to ask; you should allow them to offer deals like "We'll hire you, but your health insurance doesn't cover pregnancy." The career woman would be happy to sign, reassuring the employer.

How will that help women? It won't! On average, it's a wash: It will help career-minded women, and hurt the rest. And if you want to judge female workers on the basis of individual productivity, that is exactly what should happen.

Cheshire
Posts: 19
Joined: Thu Jul 02, 2009 11:58 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Cheshire » Sat Jul 04, 2009 8:02 pm UTC

andrewclunn wrote:In order to have this discussion properly I think we need a representative who isn't a white male. Would someone who is female or of another racial background please state that they are such, so that we might engage in using statistics in an attempt to prove or disprove differences between your group and ours.

I'm black.

Even starting to have an opinion on this whole thing is just really depressing. I know that I can, you know, not be a criminal so I'm not. I'm fine with being a statistic just not a negative one.

User avatar
Enuja
Posts: 1576
Joined: Tue Apr 14, 2009 9:40 pm UTC
Location: Chicago, IL
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby Enuja » Sun Jul 05, 2009 7:31 pm UTC

To address the OP: the more information we have, the better. The example of France, which does not collect statistics on race and ethnicity in employment and the like, is very instructive. Officially ignoring perceived differences does not make people stop perceiving and acting on these differences. There are a lot of relevant examples in this thread about data that can be misleading, but the solution is to collect more data to figure out what is going on. On the issue of racial profiling and crime, when a lower percentage of black people pulled over or asked for ID are eventually arrested, this means that a larger percentage of them actually breaking the law will be caught (because more white people will simply never be stopped in the first place). The more information we have the better, but it's also really important to known what information we don't have, to know when we just have correlation instead of causation.

athelas, let me get this straight ...
athelas wrote:Now what happens if Poles are lazy? Well, an owner will try to get information about each applicant's performance. If you find a hard worker, then you can hire them at a discount, just as before. If they're not, then you might hire them at lower wages or not hire them at all. Poles will draw lower wages, and this is as it should be - each one is judged on the basis of his abilities.
I read this to say that, if Poles are, on average, lazy, than the hard working Poles should draw lower wages and therefore be an economic boon to the employers willing to hire them. This is not as it should be, from my perspective of should. Hard-working Poles should get the same wages as hard working anybody else!

And on Bryan Caplan's suggestion about gender and hiring and child-raising: personally, my goal is to prevent employers from punishing anyone, male or female, from taking time off to raise children. My goal is not efficiency or rewarding nose-to-the-grindstone, my goal is to improve society, and I think that society is improved by an overwhelming expectation that family should be more important than work. I think the solution is for more people to take time off for children (my stay-at-home-Dad was much better than my stay-at-home-Mom, who should stay home after lactation depends on the individuals, not the gender). I can come up with all sorts of specific perverse incentive problems with asking applicants to sign something that says they won't leave to take care of/have a family, but quite beyond that, that requirement would absolutely need to be enforced on all prospective employees, male or female. From the specific example, the health insurance couldn't cover the employees or the employee's spouse's pregnancy or children. If the requirement only applied to females, that would radically increase the chance that individual women, instead of individual men, chose to stay home and take care of children.

General_Norris
Posts: 1399
Joined: Fri Apr 17, 2009 12:10 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby General_Norris » Tue Jul 07, 2009 9:44 am UTC

I think both parents should be able to stop working when they have a child, there are some countries that do that and it works fine.

User avatar
andrewclunn
Posts: 135
Joined: Mon Jun 29, 2009 7:33 pm UTC
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby andrewclunn » Tue Jul 07, 2009 4:52 pm UTC

Cheshire wrote:
andrewclunn wrote:In order to have this discussion properly I think we need a representative who isn't a white male. Would someone who is female or of another racial background please state that they are such, so that we might engage in using statistics in an attempt to prove or disprove differences between your group and ours.

I'm black.

Even starting to have an opinion on this whole thing is just really depressing. I know that I can, you know, not be a criminal so I'm not. I'm fine with being a statistic just not a negative one.


Okay good. I'm from the United States. What country are you from? I'm asking to see if I can use national census data in attempting to do a comparison.
I program in languages that would make your motherboard blush.

I also shave with a +2 Occam's razor.

User avatar
Spacemilk
Posts: 936
Joined: Tue Dec 02, 2008 9:03 pm UTC
Location: Hugh ston
Contact:

Re: Discrimatory statistics, is it really discrimination?

Postby Spacemilk » Tue Jul 07, 2009 5:59 pm UTC

I am a white female. I'll try to give my perspective on the issue of discriminatory (and reverse discriminatory) statistics. Sorry if I'm regressing back to earlier in the topic instead of continuing recent posts.

I do not like reverse discrimination, or affirmative action - whatever you like to call it. As I see it, in its most basic form the idea is to give a greater advantage to someone who has possibly had fewer advantages in their lives. However, I think it is extremely detrimental on the collegiate level and when considering applications for jobs in workplaces. I believe that it actually causes more discrimination. People tend to assume, when they see a minority get a job or get into college or what have you, that they've achieved this not on the basis of their intelligence or their skills, but because they are necessary to fill a statistical quota. The sad thing is, with affirmative action this does occasionally happen, just not THAT often; and so everyone else in that minority group gets lumped together with those one or two "quota-filling" people.

Here is an example from my life: I'm a female engineer. When I was applying to colleges, I wasn't entirely sure I wanted to be an engineer - I thought maybe I'd do chemistry, or what about psychology, it looked quite interesting... basically I had the same dilemma a lot of high school seniors have when faced with that "intended major" line. But I was pushed to put down engineering on all my applications, regardless of what I actually wanted to do, because it was believed that I'd have a better shot at getting in to tougher colleges because I was a woman looking at engineering. When I got into a respectable college and started my engineering classes, I found out most of the girls in engineering had done exactly the same thing. But I wanted to stay in engineering; a lot of those girls treated the first semester as a "blow-off" semester since they knew they'd be changing majors. As a result, the rather lazy, non-engineering-inclined (they weren't stupid, they just weren't good at engineering) fostered a mentality that ALL women were bad at engineering. It wasn't so bad in my major since I was a chemical engineer (~50% or more women in my classes) but I can't imagine what it must've been like to be, say, a woman electrical engineer (<5% women in classes) where you are constantly going against prejudice and trying to prove yourself thanks to affirmative action.

And now, as a result, I feel a little uncertain about myself - yeah, I made it into a good college and did decently well in my major, but would I even have made it in if I weren't a woman? And when I applied for a job and got it, would I even have gotten that job if I weren't a woman? It becomes an issue where not only do the people around you doubt your abilities because they assume you've gotten unfair advantages, you begin to doubt yourself as well.

The problem that reverse discrimation tries to solve is a perceived lack of advantage for people when they were young - at least that's my opinion. (some of the reverse discrimination stuff is done to counteract still-held prejudices, but as I talked about above, you're shooting yourself in the foot with that one - you're only making the problem worse) The way to do that is NOT with affirmative action. You need to start at the source - give those advantages to kids and people in their homes and early in their lives, so they can then go out and achieve things on their own.

My point is: In our personal lives, in our homes, and in the early years of children's lives, everyone should have equal ameneties and advantages. So provide for that, first. Then, what do you with those things, what you achieve, is your own doing and is a product of your own will, not because someone had to prop you up.

edit: To tie back to the OP: I guess I'd say those statistics cause more discrimination, or at least don't do any net good. However, rereading the OP, it seems more like you were trying to focus on cultural stereotypes and people's reactions to that. I guess I focused on the extremely specific case of affirmative action. I'll try to think more about the original topic and do a broader post on that.
milk from space is good for you!



Patch
Posts: 42
Joined: Sun Apr 13, 2008 6:29 pm UTC

Re: Discrimatory statistics, is it really discrimination?

Postby Patch » Tue Jul 07, 2009 11:16 pm UTC

There has been a real world experiment done that might shed light on the OP. It goes like this:

1) A high proportion of death row inmates (presumably the worst sort of criminals) are a) Black, and b) borderline mentally retarded.

2) This might be because Black people with a low IQ are more likely to commit murders or other heinous crimes. Or it might be because Black people with a low IQ are more likely to be falsely accused of murder, due to biases in our justice system.

3) Recent advances in technology have allowed us to re-evaluate some of those cases using DNA evidence.

4) If Black people with low IQ are more likely to commit crimes, you would expect that DNA evidence typically would not reverse the verdict. If there is bias in the system, you would expect to see the opposite.

I know that several former death row inmates have been exonerated due to DNA evidence. I don't have hard statistics on what percentage of cases that are re-examined in the light of DNA come up with the supposed criminal being innocent, though. Anybody know where one might track that info down?

Peace Out,
- Patch

Chen
Posts: 5580
Joined: Fri Jul 25, 2008 6:53 pm UTC
Location: Montreal

Re: Discrimatory statistics, is it really discrimination?

Postby Chen » Thu Jul 09, 2009 4:14 pm UTC

Patch wrote:4) If Black people with low IQ are more likely to commit crimes, you would expect that DNA evidence typically would not reverse the verdict. If there is bias in the system, you would expect to see the opposite.


Note here you'd need to compare the amount of overturned verdicts of non-blacks as well. A lot of reversals could just imply the system in general is poor at convicting the right people.


Return to “Serious Business”

Who is online

Users browsing this forum: No registered users and 9 guests