Stereo types and statistics

[HydrogenBond]

If one looks at stereo types, they often try to reduce a connected group to a few characteristics. I am not going to give examples. If one looks closely this is how statistical studies work, i.e, scientific stereotyping.

 

For example, stereotypes of any group contain a grain of truth. One can usually find examples of the stereo type. Yet the one size fits all does not apply to everybody, such that exceptions to rule will also occur. These people do not wish to be lumped into something they are not.

 

The same thing happens in many statistical studies. If brand X is 20% more affective, what does this mean? Does it mean that everyone taking brand X will have 20 better relief? Or does it mean that some will have better relief and some will have less relief and some will have no relief, and some will have 20% relief, with the average at 20%. The small group that actually hits the average is stereo-typical. The rest are being treated like the stereotype even though this is not true for them.

 

The question is, is this a union thing. Only science appear able to stereo type with their stereo-statistical studies. without raising the wrath of the sensitivity police. For example, blood alcohol levels stereo type all people as tea tottlers. I know many people who don't even like tea, but are forced to bear the emotional burden of being stereo typed as such.

[Ananke]

Stereotypes always have a basis in truth. People dont have the time or inclination to treat everybody as an individual, so we use shorthands, clumping people together on the basis of a shared custom or habit. The problem, if it is one, comes when the steroetype becomes outdated. A group may find itself stuck with a steroetype that simply does not fit, not because the steroetype is wrong per se, but because it is outdated.

[sebbysteiny]

Anke is right in everything he says. So I'll give him his first quality point :). But we must be careful not to ever use shorthands and judge each individual on their merits.

 

But on a further point, stereotypes may actually be accurate, but where people go wrong is by assuming they will be the only factor.

 

I am not going to give examples.

Coward.

 

For example, black people are faster runners than white. Does that mean that simply because somebody is black, that means he will be a faster runner? Of course not; it is just one factor one would need to take into account before making the assessment. Other factors would be height, build, training, determination and inner strength, trainor, personality, technique and confidence. Theretically, if one correctly balanced all the factors, one could work out who the faster runner would be. But the colour of his skin IS one factor of many required for such a calculation.

[Spiked Blood]

Humans are cognitive misers, some more so than others.:cup:

[Ananke]

You may be a cognitive miser, by which I assume you mean a retard, however I think deeply on any topic to which I turn my thoughts.

[HydrogenBond]

Although most people can see some grain of truth in many stereo-types, almost everyone agrees that stereo-types are not true enough to be accurate for all people in a group

 

The original topic was stereo-types and statistics. Where I was heading with this is that some social statisitcal studies are very much analogous to stereo-typing. They attempt to find a generalization and apply it to everyone. Ironically, most people are OK withscientific stereo-typing even though it often creates one size fits all solutions.

 

For example, the legal blood alcohol level is based on an average of alcohol related affects. Everyone is under this law, as though this stereo-type is true for everyone. The reality is that there are those who do not fall within this stereo type, yet culture will pigeon hole them as being covered by the scientific stereo type. This is no different than saying because most terrorist are Muslim, we should force social compliance on all Muslims, to be on the safe side. The ASLU would not stand for that, but seems to ignor scientific stereotyping. The only difference is that the first is done in the head, and second is done in a more formal way.

 

For example, a less stereo-typing solution to blood alcohol would be to create a test, which is voluntary, where a person would actually get a hard number for them, with respect to alcohol blood level and corrdination/reaction. This is placed on the licence. This would supersede the scientific stereo type. Currently, all Muslims are not placed under detension. Evidence needs to be gathered to prove the stereo type of Muslim equals terrorists before action can be taken.

[InfiniteNow]

Where I was heading with this is that some social statisitcal studies are very much analogous to stereo-typing. They attempt to find a generalization and apply it to everyone.

Not really, no. They study populations, hopefully obtaining an accurate cross section first, then plot those numbers and observe trends, applying various forumlae in attempts to predict future outcomes, continually adjusting based on what they learn with each iteration.

 

Stereotyping is saying that all blacks steal, or all mexicans cannot speak english, or some other ridiculous thing like that.

[HydrogenBond]

Sometimes the interpretation of scientific studies associated with populations comes down to politics. This is often where the money behind such studies stem from. If the result came out differently than desired, the research money would dry up.

 

A NASCAR driver would still be a much better driver than most of population, even with triple the legal blood alcohol limit. Yet the statistical stereo-type will say even such as these are incapable of driving an auto beyond the current set point. It creates a social stereo-type that attempts to turn everyone into a lower common demoninator.

 

The political factor may be similar to the tabboo of racial profiling. One is not able to profile the lightweight drinkers, who are indeed a hazzard at the set limit, so we need to assume everyone is a light weight. It would make more sense, as a scientist, for each driver to have a hard number that is appropriate to them. This is called reality science.

[InfiniteNow]

So is your thread about statistics and stereotypes, or is it about the legal definition of drunk as pertains to the driving laws? :cup:

[Freddy]

The laws against drunk driving are now uniform across the US with the legal limit set at .08% alcohol in the blood. They are not about if some drivers drive better at or above that limit than others do. It is about that this society does not want impaired people behind the wheel and it does not want drivers to drink and drive. Driving is a priviledge and not a right and if drivers want to keep the priviledge then follow laws concerning motor vehicles. Almost 44k died in automobile accidents in 2005. The statistics did give the breakdown for alcohol related deaths at 39% in the link below. Driving while under the influence is a crime.

 

http://www.cbsnews.com/stories/2006/08/16/national/main1900989.shtm

[sebbysteiny]

I think the argument posed by Hydrogen Bond is slightly tougher than people think. It goes like this.

 

Steriotyping is assuming traits about all individuals based on a higher probability of having that trait from their race. This is morally wrong, unreliable and irrational.

 

Statistics often assumes outcomes about individual data points based on a higher probability of having that outcome from the type of data point.

 

Therefore, so the argument goes, statistics is morally wrong too, unreliable and irrational.

 

Most people here seem to be tackling this from the point of view that statistics are accurate. Whilst this is true, it does not directly break Hydrogen's argument. Instead, we must find the distinction between steriotyping (wrong) and statistics (right).

 

My first attempt is to say that steriotyping is fundamentally bad statistics born out of deliberately ignoring more important factors and also possibly drawing false conclusions as a consequence of cultural ignorance.

[sebbysteiny]

x

[InfiniteNow]

This is morally wrong, unreliable and irrational.

Okay, but define those this for me in a way that is not relative to your own viewpoint and applies equally to all observers. Morally wrong? What the ___ is that?

[sebbysteiny]

Steriotyping is assuming traits about all individuals based on a higher probability of having that trait from their race. This is morally wrong, unreliable and irrational.

 

Okay, but define those this for me in a way that is not relative to your own viewpoint and applies equally to all observers. Morally wrong? What the ___ is that?

 

Well since steriotyping is essentially identical to racism, I didn't expect too many people to question its lack of moral validity. I agree that somebody who has shwazstiker posters on every wall of their house might feel steriotyping is a morally correct tool to judge a fellow human being. Sorry about the dodgy assumption.

[InfiniteNow]

Well since steriotyping is essentially identical to racism

 

I know what you are attempting to say, but your point is not accurate. I could stereotype based on gender. That could be analogous to sexism, but does not have to be. I could stereotype based on age, which might, but does not have to be analogous to agism. I could stereotype based on hair color. That could be analogous to chromafolliclism... :confused:

 

The distinction is important though... to stereotype is to assume traits by generalizing on an incomplete set of attributes, where making something into an "ism" tends to imply a negative generalization. The stereotype is not in and of itself negative. I think the portion of the word that's most relavant is the "typing" portion.

 

Sorry about the dodgy assumption.

No worries. I will, however, occastionally ask you to clarify. Here again, you still haven't answered how we define "morally wrong" in a way that is not subjective to the observer. Then again, I'm not sure anybody can.

 

 

Cheers. :confused:

[HydrogenBond]

If we did a statistical study between men and women connected to physical strength, the study would say that men are stronger than woman. This would be considered a sexist stereo type. There are some very strong women and some very weak men. This statistical stereo-type sort of creates a false sense of reality. Reality is case by case and not a stereo-type.

[sebbysteiny]

If we did a statistical study between men and women connected to physical strength, the study would say that men are stronger than woman.

 

This is where you are wrong. The statistics would say that men are stronger than women IN GENERAL. That implies that there are exceptions. The added words are essential in understanding the statistics. I think once you add those necessary words, the point that you were making falls into the abyss.