Statistics
In almost all areas of public policy, we see borne out that statistics are the most pernicious of lies. "More likely to" has become a way to enshrine in law a failure of logic.
The failure is simple to see, but, I guess, equally easy to ignore. Simply stated, it is the assignment to all members of a group characteristics applicable to only a portion of the members of the group. It is, by extension, the attribution of causality where none is statistically demonstrated. Statistical correlation is a pretty loose thing, but causality requires a 100% correlation: the factor under consideration must always lead to the outcome discussed, and it is helpful if the outcome discussed can arise only from the factor under consideration. Statistics other than 0.00% or 100.00% are meaningless when individuals or individual cases are considered.
To put it crudely, if 99.99% of all blond-haired individuals are irredeemable axe-murderers, and you meet a blond(e), absent behavior you cannot say whether that individual is from the 99.99% or the 0.01%. It is therefore prudent to be cautious around blond(e)s, but treating all of them as one would treat an axe-murderer -- that is, badly -- most likely shall result in the other members of the 0.01% becoming axe-murderers. That might make it easy to deal with blond(e)s: kill'em all. But until you've done so, you'll have more axe-murders. Add to that the fact that the non-murdering variety of blond(e) might accumulate non-blond(e) friends, who might act to avenge actions against their in-no-way-homocidal blond(e) buddy -- even chosing axes as their weapon. In short, an action based on statistics is likely to lead to an outcome contrary to that intended -- in this case, increase in axe-murders instead of security against axe-murdering blond(e)s.
The above applies to racial profiling and racism (et al.) in general, but I'm writing this because I keep hearing social conservatives spout statistics on traditional contrasted with non-traditional parenting (single-parent households, homosexual partner households, communes) as an argument to, I don't know, require individuals by law to not have children except by their heterosexual spouses. Statistics on commission of crimes, expulsion from or abandonment of school, drug-use, and violent behavioir all show greater incidence of each among children of single-parent households than those of two-(heterosexual)-parent households. But the statistics are nowhere near 0 or 100 percent. Do all children of single-parent households commit crime, drop out of school, or draw weapons? The statistics say no. Do all children who commit crime, drop out of school, or draw weapons come from single-parent households? Again, no. Our prisons, drug-rehab clinics, and anger-management classes all have more than a few individuals reared by their biological fathers and mothers. So there must be something else involved -- good parenting, perhaps? And if good children emerge from single-parent households, homosexual households, and communes, isn't it possible for unwed/divorced/widowed individuals, homosexual individuals, and good ol' group lovin' individuals to demonstrate good parenting?
There something else missing in the discussions of social conservatives. What percentage of individuals who cooked the books at Enron came from two-parent households? It is the individual, not his or her background, that makes decisions; it is those decisions that lead to actions; it is those actions that lead to effects -- and it is on effects that we must judge individuals.
So the social conservatives should just shut the hell up (yes, John Leo, I mean you) until they can incorporate logic into their sermonizing.
The failure is simple to see, but, I guess, equally easy to ignore. Simply stated, it is the assignment to all members of a group characteristics applicable to only a portion of the members of the group. It is, by extension, the attribution of causality where none is statistically demonstrated. Statistical correlation is a pretty loose thing, but causality requires a 100% correlation: the factor under consideration must always lead to the outcome discussed, and it is helpful if the outcome discussed can arise only from the factor under consideration. Statistics other than 0.00% or 100.00% are meaningless when individuals or individual cases are considered.
To put it crudely, if 99.99% of all blond-haired individuals are irredeemable axe-murderers, and you meet a blond(e), absent behavior you cannot say whether that individual is from the 99.99% or the 0.01%. It is therefore prudent to be cautious around blond(e)s, but treating all of them as one would treat an axe-murderer -- that is, badly -- most likely shall result in the other members of the 0.01% becoming axe-murderers. That might make it easy to deal with blond(e)s: kill'em all. But until you've done so, you'll have more axe-murders. Add to that the fact that the non-murdering variety of blond(e) might accumulate non-blond(e) friends, who might act to avenge actions against their in-no-way-homocidal blond(e) buddy -- even chosing axes as their weapon. In short, an action based on statistics is likely to lead to an outcome contrary to that intended -- in this case, increase in axe-murders instead of security against axe-murdering blond(e)s.
The above applies to racial profiling and racism (et al.) in general, but I'm writing this because I keep hearing social conservatives spout statistics on traditional contrasted with non-traditional parenting (single-parent households, homosexual partner households, communes) as an argument to, I don't know, require individuals by law to not have children except by their heterosexual spouses. Statistics on commission of crimes, expulsion from or abandonment of school, drug-use, and violent behavioir all show greater incidence of each among children of single-parent households than those of two-(heterosexual)-parent households. But the statistics are nowhere near 0 or 100 percent. Do all children of single-parent households commit crime, drop out of school, or draw weapons? The statistics say no. Do all children who commit crime, drop out of school, or draw weapons come from single-parent households? Again, no. Our prisons, drug-rehab clinics, and anger-management classes all have more than a few individuals reared by their biological fathers and mothers. So there must be something else involved -- good parenting, perhaps? And if good children emerge from single-parent households, homosexual households, and communes, isn't it possible for unwed/divorced/widowed individuals, homosexual individuals, and good ol' group lovin' individuals to demonstrate good parenting?
There something else missing in the discussions of social conservatives. What percentage of individuals who cooked the books at Enron came from two-parent households? It is the individual, not his or her background, that makes decisions; it is those decisions that lead to actions; it is those actions that lead to effects -- and it is on effects that we must judge individuals.
So the social conservatives should just shut the hell up (yes, John Leo, I mean you) until they can incorporate logic into their sermonizing.
posted by heavynettle at 09:45
5 Comments:
One of the reasons that public education is provided is the belief in the postulate that a free society must first be an educated society. I personally agree with this, but I can see that it's not an absolute dichotomy: educated = free or free = educated.
Anyway, some of your recent posts have shown that ignorance of current events, the Constitution, or even of a fully articulated goal, i.e. the American dream, seem to be common elements in a society and nation rushing headlong into the burning barn with its blinders on.
So, this should really make your day. Not only do many Americans not realize the San Where-ever-the-hell in your new favorite TV show is not a real place; apparently only 4 in 5 are aware that the earth revolves around the sun. Ahttp://www.goatview.com/august3scopesgallup.htm poll by the elder statsmen at Gallup revealed this fact. (But don't feel bad, even fewer Germans or Brits knew the true nature of the solar system.) Galileo and Copernicus sacrificed for THIS??? So that 20% of the members of a democratic and secular society could grow up without knowing such a basic fact I would expect most 5-year-olds to know??? Well, yes.
Other fun facts in the poll reveal that, far from being able to name how many amendments were in the Bill of Rights or other arcane facts, only 3 in 4 American adults know that the country won its independence from Great Britain. And the vote of each one of these people counts for as much as yours.
Everyday I believe more and more that the rights of citizenship should have to be earned by native born Americans, just as it is by naturalized citizens, all of whom, to be sure, must learn about who it was we fought at Concord and Lexington.
Okay, 1) Commander-in-Chief is not my favorite show!
2) Copernicus is overrated. One point I like to make when I'm talking to such morons who consider Intelligent Design a scientific hypothesis is that science exists not to determine the true nature of the world as to provide us with a means of understanding (and, presumably, predicting) the world. The Ptolemaic system, with Earth orbited by the other nine known heavenly bodies in epicyclical orbits, had one big advantage over the Copernican system: it worked. That is, if one wanted to determine the position of Mars on a given day in the future or past, one had merely to plug in the hideously complicated math to get the information -- Copernicus' model, with perfectly circular orbits, could do no such thing. It wasn't until Keppler came up with elliptical orbits that a working heliocentric model could be devised. The difference is that, while the Church was dogmaticallly asserting that the Earth was the center of the universe and Copernicus was, with insufficient understanding, parroting something he translated from Greek, the scientists were keeping their heads down and using observation to make the former dogma work for everyday persons. The fact is that, mathematically, we can say any point in the universe is the center of the universe -- it's just a bear to make all the math work. The heliocentric system works because it simplifies our understanding of how astronomical objects interact, not because it's "true".
Still, the level of scientific ignorance in our society always makes me want to go buy a gun. For perspective, I guess, we should remember that a reasoned, analytical approach to life is something that the overwhelming majority of humans have always been unable to handle. The difference today is, as you point out, that each of the subhumans has the same political authority as I do. We should minimize their dangerous influence by eliminating the popular vote for President and (after the appropriate amendment) Senators.
A test for citizens would be useful, too, but what would be on it? I would think the Constitution would be enough, but look at the variety of interpretations of that rather brief document offered by the Supreme Court for all these years.
Sigh.
I agree that Senators should not be directly elected; they should be elected by our local representatives of the third world--state legislatures. Why? Because state legislatures have a lot more direct impact on our day to day lives than is generally appreciated but they are ignored by most of the population (quick - name your state senator! OK, YOU probably can) as a cute little "baby congress" in which to cut your political teeth.
Giving state legislatures back the power to elect Senators would make the relevance of those local bodies clear to everyone, and perhaps spur interest in local processes and policies.
Interestingly enough, the official U.S. Senate website has a blurb explaining the status quo as the best possible system. Go figure.
OK, let us be free here to talk about racial profiling.
It sucks. It's self-defeating, or at best a self-fulfilling prophecy. You should look at some of Leonard Pitts recent columns about his son's experience with the police to see how horrible it is for young person to be busted for D.W.B. or other such nonsense.
But how to you stop people from learning from experience? You mention the 99.9% of blonds being psychos as not being good enough to pre-judge. Well, I agree with you intellectually; but in my gut I still know that the vast majority of serial killers are white, and that, as a rule, gang-bangers of black and Latino background at least usually have to have some personal connection to you before they'll off you for no reason. But then again, I don't know very many black folks who've been mugged by white guys. Do you? We all live in fear of "the other," not knowing how to hold their hand and sing kumbayaa.
You've talked a lot on this blog about the need for individuals to build communities around themselves--to be the social creatures we are--in order to thrive. I would argue that the intent behind the federal government was in fact to create a mega-community, and that we pay taxes voluntarily to support the same. When we opt out of the megacommunity, we criminalize ourselves in the eyes of the other three hundred million members of the group.
If the federal entity were really a community, I would embrace it. Unfortunately, it isn't. It's an enforcement mechanism for a farm where some animals are more equal than others, where African decendents of slaves lobby for "the government" to pay them reparations because, well, they don't identify enough with that government to realize it would only be a case of them paying themselves.
No real comment here, just looking for answers on how to avoid the traps our instincts have programmed us to embrace; mainly, that the best course of action is to sneak over the hill and kill the strangers or neighboring family of chimps in their sleep before they do it to you.
Experientially my blond analogy bears out, if you adopt a good neutral standard of behavior. A blond carrying an axe would be more suspicious than a one carrying a puppy, and of two blonds with axes, the one that is advancing threateningly on me would be more suspicious than the one with axe over shoulder whistling while walking in the opposite direction.
Racial profiling is useless, but profiling can be constructive. The black guy in a three-piece suit arouses no more suspicion in me than a white guy in one, but put gang colors, a scowl, and knuckle-dragging swagger on either a black or a white guy and I'll be much more attentive.
In short, then, learning from experience is not to be prevented, but expanded: mingle with persons who arouse irrational fear in you, and you'll likely be able to dismiss any such fear in the future -- basing a reaction on the actions of individuals, rather than incidental characteristics.
As to another of your points, the federal government is most definitely not a super-community: it exists to protect the rights of individuals, not to bind us together. That's what fascist governments do. Community is something the individual must form, with like-minded individuals, not something that the guys with guns cause us to do.
Post a Comment
<< Home