Orton's Arm

The fact that you weren't able to understand my example the first time suggests that your natural aptitude for statistics isn't particuarly high, and that none of the training you've received can make up for this.

Because there are very few conservative Democrats left. Because many soldiers don't want to accept the possibility that they're risking their lives for nothing. They've been told (rightly or wrongly) that there's a very good reason for what they're being asked to do. Because many people are more easily led by social tactics such as namecalling than they are by information and logic. See answer to #3. Plus, there are so many things competing for people's time. A simple message more easily penetrates this noise than a more nuanced view.

Yes. Had they intended the definition you have in mind, they would have used an upper case H.

I'll respond here, in order to (belatedly) avoid hijacking yet another thread. You are incorrect in saying that my example illustrates regression away from the mean. Here's why. Consider a world in which the following was true: you had a perfect I.Q. test, which gave you the same score every time. Also in this example, the heritability term in the I.Q. equation is now 1. So far, so good--there is no regression toward the mean. But now, keep everything the same, except that you allow your I.Q. test to be flawed. The underlying truth is that if two parents have an average I.Q. of 140, their children will be expected to have I.Q.s that are the same. But as I illustrated in my earlier post, measurement error creates a distorted view of things. So the people who got 140s on the I.Q. test really have an average I.Q. that's somewhat lower. This is because there are more people with I.Q.s of 130 or 135 (available for getting lucky) than there are people with I.Q.s of 150 or 145 available for getting unlucky. So if you have a group of people who all got a 140 on an I.Q. test, the true average I.Q. of that group will be less than 140. If you give that group a second I.Q. test, its true average will make itself clear. But in this case, the second I.Q. test could take the form of those people having kids. The average child from that group will reflect that group's true (130s-style) average, and not its inflated average of 140. "Look!" people might say, "The kids are regressing toward the mean. Why is this?" The reality is that the kids' I.Q.s are, on average, the same as their parents'. It's just that you overestimated how smart those parents were.

At least we're in agreement that you've been using the first definition. Clearly, the American Psychological Association had the second definition in mind when discussing heritability, because they used a lower case h. As for that second definition that the American Psychological Association was using, the Answers.com article had this to say: In other words, the Wikipedia article's equation was using the term in exactly the right way.

You're still unable to understand how measurement error can cause regression toward the mean? Let me spell things out in simpler terms. Example 1: consider a population where everyone's I.Q. is exactly 100. But the I.Q. test is error-prone, with some people getting lucky (a score that's 10 points higher than it should be) or unlucky (a score that's 10 points too low). Suppose you were to look just at the people who scored 110 on that I.Q. test. Suppose you were to ask them to take a retest. Their scores would completely regress to the mean. That is, the average retest score would be 100, because in the retest the number of lucky and unlucky people would balance each other out. Example 2: Consider a population with a bell-curve (Gaussian) distribution of I.Q.s and the same error-prone I.Q. test described above. Any individual person is equally likely to get lucky or unlucky on this I.Q. test. Suppose for example you're looking at those who scored 140 on the I.Q. test. You know that a certain percentage (say 20%) of the people with I.Q.s of 150 will get unlucky and appear to have an I.Q. of 140. You know that a certain percentage (20% for example) of the people with an I.Q. of 130 will get lucky and appear to have an I.Q. of 140. Because the overall population is normally distributed with an average I.Q. of 100, the number of people with an I.Q. of 130 will be a lot larger than the number of people with I.Q.s of 150. 20% of a large number (the population size for I.Q.s of 130) is a lot bigger than 20% of a small number (the population size for people with I.Q.s of 150). The lucky 130s will outnumber the unlucky 150s. Therefore, the people who scored 140 on that test are, on average, less intelligent than their 140 score will indicate. Should that group ever be asked to take a retest, the presence of all those unlucky 130s will make itself felt. The second time around, lucky and unlucky people will balance each other out, and the measured score will more truly indicate that group's potential. Example 3: Consider a population with a bell-curve distribution of I.Q.s, and an error-prone I.Q. test that produces normally distributed results centered around the true mean. That is, if you were to take the I.Q. test 1000 times, your scores would vary. Your average score would be your true I.Q. 64% of your scores would be within one standard deviation of the mean, 95% within about two standard deviations, etc. The logic here is the same as in example 2. Consider the population of those who scored a 140 on the test. That group will contain more lucky people with true I.Q.s of 130, 135, etc. than unlucky people with I.Q.s of 145 or 150. This is because there are more people with I.Q.s of 130, available for getting lucky, than there are of people with I.Q.s of 150 who are available to get unlucky. Once again, the people whose I.Q.s were measured to be 140 will, on average, have I.Q.s that are somewhat lower. Measure them a second time, and now you'll find out the true average I.Q. of that particular group of people. That population will appear to regress toward the mean.

I did some more research, and it turns out the word "heritability" has two definitions. There's "broad sense heritability" which is the definition you seem to have in mind, as well as "narrow sense heritability" which is the way the word was being used in the Wikipedia article and equation. "Broad sense heritability" is represented by a capital H^2, while "narrow sense heritability" is represented by a lower-case h^2. The American Psychological Association used the term "heritability" in the narrow, lower-case h sense in its response to the Bell Curve controversy In other words, the American Psychological Association is saying that parental intelligence predicts 75% of a child's I.Q., with the remaining 25% predicted by the intelligence of the child's relevant population group.

Someone with your, um, conversational style should know how to spell "vitriol."

You need to go back and reread the I.Q. equation, and surrounding text. Here it is

Why should you listen to me? Because just about every time we disagree on something, I'm right and you're wrong. That's why.

It's statements like this which make me feel your contribution to the eugenics discussion has been worth little. The Weiss study determined the intelligence of the proband children through their performance on a standardized math test; not by asking them what they want to be when they grow up. Misguided comments like the one you just made added significantly to the length of the discussion, while at the same time decreasing its overall usefulness.

Other than the fact that you have zero ability to evaluate intelligence, your post didn't have much of a point either.

I see my thinking has made a strong impression on you. I can see why.

Unless you're an anarchist, you too feel the government has a job to protect us. To protect us from criminals via a functional police force and criminal justice system. To protect us from foreign invasion through a strong military presence. Maybe even to protect us from traffic fatailities by maintaining a safe road system. In trying to make a fool of Bill, you only made yourself appear foolish. There are legitimate ways of disagreeing with Bill's position. Calling him an idiot for the "government has a job to protect us" statement isn't one of them.

Before the Green Bay game, the question was asked about how we should respond if Losman played well but the Bills lost anyway. I had the following response It works the other way too. If you win but your FG kicker goes 0-6, you know you have to do something at kicker. If you win but your QB barely breaks the 100 yard mark, you know you need better play from your QB. The Bills won because the Packers shot themselves in the foot. The Bills won despite the lack of production at QB. If that lack of production continues, the Bills should start thinking about what they need to do to correct the situation. At some point their thoughts may turn to Nall. Obviously they thought highly enough of Nall to promise him a legitimate shot at the starting position. But he got off on the wrong foot because it took him too much time to learn the new offense. Then came the injury, and before he knew it he was third string. As an outsider looking in, it seems like Nall was written off before the Bills' coaching staff had truly figured out what he could do. Some fans feel that Losman should be judged with a Drew Brees ruler, and should be given years of opportunity to start looking like a real QB. A few such people are willing to write Nall off based on what he did when he was still learning the new offense. Was Nall written off too quickly? Is it time to start remembering the reasons why the Bills brought him here in the first place? Is it getting time to put him in a game or two, just to see if he's the next Tony Romo? The Bills will have to ask themselves these questions going into next week's game.

From 2001 - 2005, these are the QBs who were taken in the first four picks of the draft: 2001: Michael Vick 2002: David Carr, Joey Harrington 2003: Carson Palmer 2004: Eli Manning, Philip Rivers 2005: Alex Smith From 2001 - 2005, these are the offensive linemen taken in the first four picks of the draft: 2001: Leonard Davis 2002: Mike Williams 2003: none 2004: Robert Gallery 2005: none No, I'm not saying that first round offensive linemen in general work out as poorly as the guys on the above list. But there may be more risk to a high OL pick than you might realize; without the same high reward a star QB would bring. Unless Losman starts playing better than he did today, the Bills might be better served taking a QB early in the first, and going OL in the 2nd and 3rd.

You may be right in saying that TD's performance in rounds 3 - 7 was par for the course. I don't know one way or the other. Partly, my sig is a response to people who make the claim that TD didn't neglect the offensive line. These people will point to some pick in round 4 or 6 or whatever, and to them that's at least a good faith effort. If that's how they want to look at it, fine. But the implication there is that those picks in rounds 3 - 7 have a fair amount of value. If that's what these people are implying, then in order to be consistent they have to say that TD messed up by only finding one starter in rounds 3 - 7 over the course of five years.

While your post doesn't contain a single word of truth, it does contain varying degrees of falsehood. First, the definition of heritability is mathematical, and has to do with the extent to which a trait is passed from parent to child. It's one of the terms in the equation about which we're arguing. You haven't demonstrated consistent commitment to this definition, as shown by the confusion you tried to create about whether intelligence was passed from parent to child. Secondly, I provided an example of how measurement error can cause the appearence of regression toward the mean. If you don't understand the example, I'll be happy to explain it again. If you feel you understand it but don't agree with it, I'd be happy to listen to whichever legitimate reasons you feel you may have for having disagreed. You were too busy insulting me to mention any such reasons. Thirdly, you accused me of lying, cheating, and manipulating to get my way. After lying about my knowledge of statistics, I'm amazed you have the nerve to accuse me of dishonesty. Then again, Lenin said to always accuse your enemies of that which you yourself are guilty. Maybe that's what you're doing here. Fourth, you claim that you know much more about this subject than I do. I have yet to see evidence of this claim. You repeatedly demanded that I provide evidence of a link between smart parents and smart children. You would look equally silly if you'd a) refused to see a link between smoking and cancer until I'd shown you evidence, b) ignored the evidence once it was presented, and c) represented yourself as a health expert. Fifth, I think that a lot of people probably are confused by exactly what meaning (if any) you're trying to hide between insults. But some people may not want to ask questions because they assume (falsely) that you know what on earth you're talking about, and that the issue is settled. Others may be more curious, but might be too afraid of seeming foolish to speak up. Then there are those who wisely stopped reading this thread once you made it into a flame war. Because in general, it's good for questions about science or statistics to get decided by flame wars.

Walk away with my tail between my legs? You've got to be absolutely joking, right? Are the fact-free insults coming from you and Bungee Jumper supposed to accomplish that? Is it the fact that I have the stronger case? Or is it the fact that neither of you is letting himself get pinned down? Or is it the fact that neither of you supposed statistics experts can understand a very simple concept such as how measurement error can cause the appearance of regression toward the mean? You've made a number of posts in this thread. None of them have had any information content whatsoever. Zero. Nill. Nada. Is behavior like that supposed to make me run away with my tail between my legs? If this was a bar fight, you wouldn't need to use tricky things like facts, or logic. You could just push people around if they disagreed with you, and if you were bigger. Maybe that would suit your personality better. But on a discussion board, facts and logic actually matter, at least up to a point. The facts and logic happen to favor the pro-eugenics side of this debate.

The next time you think it's a good idea to post . . . don't.

Is this why you've been so consistent in ignoring facts?

Your posts have been 1% explanation, 99% insult. Other than the fact I'm being insulted, there is very little in your posts for me to understand. Bungee Jumper is a little better, but he typically likes to avoid being pinned down. "You don't understand heritability," he'll say, without committing himself to either a) a definition of the word, b) where he thinks I've misunderstood the term, or c) an explanation of how this affects the discussion. How is someone reading this thread possibly supposed to learn anything from a post like that? The only message either you or Bungee Jumper are trying to send is that people should disagree with me even if they don't exactly know why. That's fine for people who like to be led, but it might be frustrating for those who like to know the reasons for what they believe.

It usually is.

You don't think it's possible that a Wikipedia contributor just might know more about statistics than either Ramius or Bungee Jumper? If you think that possibility is zero, then so is your own knowledge of statistics.