Orton's Arm

Maybe you should a) spend a little less time throwing insults my way and b) spend a little more time actually explaining what you mean. Instead, you've been hiding behind statements like "you don't understand X." If you feel communication errors have taken place, and you want someone to blame, maybe you should start by looking in the mirror.

Joe Six Pack's response to this was a slam dunk. But I'll add to what he wrote by saying that I know a woman who lived in Sweden. She had a low income while she was there. Paid high taxes. In fact, her taxes were so high that by the time she'd finished paying them, her rent, and her utilities, it was really tough to buy food. In a bad month, the last week or so would have no money at all for food, or for anything else. But if going hungry ever caused her medical problems, at least the hospital stay would be free.

I'm curious as to exactly what aspect of the Washington political process you feel results in money going to make a better society. Because what I see is money going to those with the most political leverage. (Consider agricultural subsidies for example.) The larger the government's role in the economy becomes, the more the messed-up political process in Washington will determine winners and losers.

I put a greater level of faith in that Wiki article than I do in either Ramius or Bungee Jumper. I feel the latter two are entirely too willing to allow their feelings about me to distort their posts about statistics. For example, Ramius came to the defense of syhuang; despite the fact that syhuang's attack against me was an embarrassment to syhuang. I don't see any flaw with the logic of the Wikipedia article. Bungee Jumper and Ramius can be as arrogant and pompous as they choose, but their emotions--no matter how strongly held--are not a substitute for logic. Their unsupported statements are not a substitute for statements about logic. They claim to speak for the scientific community; but are actually treating this discussion in an entirely unscientific fashion. They're assuming that genes are innocent of affecting intelligence until proven guilty. The spirit of science is not to make assumptions one way or the other. Then, when I found a document showing that genes do affect intelligence, they proceeded to question whether a parent's intelligence really is passed along to the child. These questions, by the way, were driven by Bungee Jumper's own misunderstanding of heritability. Its true definition is mathematical, and is one of the terms of the equation that predicts a child's intelligence. There is no doubt that intelligent parents are more likely to have intelligent children than are less intelligent parents. The fact that both Bungee Jumper and Ramius have appeared to ridicule this concept shows their partisanship rises above their respect for objective truth. To return to the Wikipedia article. None of those attacking its explanation of regression toward the mean were able to identify a flaw with its logic. I'm confident there is no flaw. I'm also confident that if Ramius and Bungee Jumper took a little time to stop celebrating themselves for being so well-credentialed; and actually thought about what the article was trying to say, they'd realize it was correct. Whoever wrote that article may well know more about statistics than either of them.

I'm not so sure about that. Mike Williams played his best football for McNally (2nd half of 2004). You don't see former McNally players going on to have more success elsewhere than they did in Buffalo, do you? The Bucs are already displeased with big Mike Williams, Trey Teague is hurt for the Jets, and I don't think Bennie Anderson is setting the world on fire for the Dolphins. At some point, you have to start holding players responsible for their own mistakes--including mental errors.

The point I'm making is that TD messed up either by a) not allocating enough rounds 1 & 2 picks to the OL, or b) not doing well enough in rounds 3 - 7. One or the other. Maybe both. I agree it's too early to call Romo anything more than a one game wonder. But I just get the sense this guy might amount to something.

http://sports.espn.go.com/nfl/columns/stor...hael&id=2645741

Is it that time of month? Because you've managed to do the following: - Miss the whole point about regression toward the mean article and my follow-up example. Weiss obviously understands that point. Why don't you? - Use the simplified example I provided to imply (falsely) that I was unable to conceptualize more statistically realistic examples. - Once again attempt to create confusion about the word "heritability" - Falsely claim that my only source is Weiss, when I've cited other sources to butress my case. In fact, I don't need Weiss at all--the formula in the article, together with the American Psychological Association's determination that heritability is about 0.75, are enough. - Falsely claimed that I don't know the difference between error and regression. Then you failed to provide any support for this baseless claim. Come to think of it, you failed to provide support for any of the accusations you made against me. The heritability accusation. The genetics accusation. The mean/discrete value accusation. - Represented yourself as speaking for the entire scientific community, when in fact you speak only for yourself. - Oh, and one last thing: Weiss's paper wasn't about the career choices of East German children. In fact, I couldn't find a single true statement in your entire post. Not one. Are you trying to create this much confusion about what's actually a fairly simple issue?

Thanks for bringing a little fresh air into what's become a stuffy discussion. I innocently created a system, which while not perfect seemed like it could shed some light on how well a QB is doing to help his team score points. I didn't realize at the time that this system would inspire syhuang to launch a personal crusade against me.

Given that some people appear to have had difficulty understanding the point I was making, I'll express my thoughts on regression toward the mean in simpler, easier to understand terms. Consider an I.Q. test. 60% of the people who take the test get the correct score. 20% get lucky, and score 10 points higher than they should. 20% get unlucky, and score 10 points lower than they should. Now imagine a room filled with people who scored a 140 on the test. Most of those people will indeed have I.Q.s of 140. But some of the people will have I.Q.s of 150, but who due to bad luck scored 10 points lower than they should have. There will be people in this room with I.Q.s of 130, but who through good luck scored a 140 on the test. Because there are a lot more people with I.Q.s of 130 than there are with I.Q.s of 150, the lucky 130s in the room will significantly outnumber the unlucky 150s. Therefore, the average intelligence of the people in the room will be lower than the 140 you measured. Suppose you were to ask everyone in the room to take a second I.Q. test. You'd expect the number of lucky people to balance out the number of unlucky people. Once you'd finished averaging the results of the second test, you'd know the true average I.Q. for the people in the room. That true average will be lower than 140. This is what the Wikipedia article meant when it described regression toward the mean. What did I mean when I wrote that regression toward the mean could be caused by error in measurement? Suppose that the expected value of a child's intelligence is the average of the two parents' I.Q.s. Now suppose that the people in the I.Q. 140 room had been kept there so long they started having kids with each other. Based on the I.Q. tests those people took, you'd expect their kids to have I.Q.s of 140. That was the measured I.Q. But the true average I.Q. in that room is lower--perhaps 135. Once those people's children start taking I.Q. tests, those children will get an average score of 135. This is what Weiss meant when he pointed out that measurement error could explain regression toward the mean.

I wrote: Later on, you wrote: You followed it up with I'm glad you've done the math to prove what I was already saying: the equation I mentioned is a mathematical definition of regression toward the mean. If you think I didn't know that already, go back and reread my above posts. The question is what's causing the observed regression toward the mean that the equation describes? As the article I found points out, measurement error produces the appearance of regression toward the mean.

In practical terms, the bulk limit theorem states the following: suppose the true population is normally distributed. You begin sampling. At first, your sample will be a t-distribution, but as you add to your sample size the t-distribution will get taller and skinnier to look more and more like the normal distribution of the true population. In other words, the bigger your sample, the more closely it will resemble the true population. You appear to be confusing this with another statistical phenomenon: regression toward the mean:

Thanks for once again seeming to say something without having said anything. Beyond the fact that you disagree, of course. But we knew that already, because you've repeated it a few hundred times.

"Analyze the equation at the bulk limit where any measurement error averages out?" Let me put this to you in simple terms: you have an equation that deals with children one by one. It deals with the individual parents of these children. There are several ways in which measurement error can occur: 1. The child's intelligence can be underestimated 2. The parents' intelligence can be underestimated 3. The child's intelligence can be overestimated 4. The parents' intelligence can be overestimated Let's do a thought experiment, and pretend these error terms are very strong. In fact, let's pretend they're so strong that no individual parent or child measurement has any validity whatsoever. In a world where measurement was as bad as this, whatever correlation might exist between parental and child I.Q. would be completely masked. It's masked on the individual level, and it's masked in the aggregate. But now measurement gets better; so there's now some relationship between a person's measured I.Q. and his or her real I.Q. This is the point where we can at least glimpse whatever correlation exists between parental and child I.Q. But only partially, because measurement error is still masking most of the relationship. If your measurement error was zero, you could truly see whether regression toward the mean was taking place. But due to the imperfect measurement systems we have, what appears to be regression toward the mean may actually be the result of measurement error partially occluding the true relationship between parental and child I.Q.s. So this is the argument against regression toward the mean. But there is also a chance regression toward the mean is real. Does this possibility mean a eugenics program wouldn't work? Not at all. Your earlier statements to the contrary notwithstanding, a heritability of 0.75 does in fact mean that, on average, 75% of a smart person's intelligence will be passed onto his or her children. So a couple with an average I.Q. of 200 will produce children with average I.Q.s of 175. Yes, there's a feeling of disappointment that the children probably won't be as smart as their parents, but at least they'll most likely be smarter than Ryan Leaf's kids. By encouraging the geniuses of the world to have more kids, the population will be smarter than it otherwise would have been.

I'm not sure how much is due to Hollywood changing. Probably most Hollywood types had airheaded views 30 years ago, just as they do today. But there are always exceptions to the airheaded trend. Harrison Ford is such an exception today, just as Ronald Reagan was an exception in the past. The thing about Reagan is that his views were formed by getting out of the Hollywood ivory tower, and actually listening to real people talk about their problems. When he said, "government isn't the solution to the problem, government is the problem," his words weren't spoken at random. They were the natural conclusion of a man who'd listened to countless people tell him horror stories of out-of-control government bureaucracies. Limousine Liberals may have disagreed with him, but their views were formed strictly by their experience inside the Hollywood ivory tower. Reagan earned the publicity he was given. But others are given publicity for their political views because they're famous, and not because they have anything worthwhile to say. This is the problem a lot of people have with Limousine Liberals.

There you go again. The issue we were arguing about was whether intelligent adults are more likely than average to have intelligent children. You, apparently, weren't convinced that intelligent parents were any more likely to produce smart children than were any other type of parents. The formula demonstrates, with mathematical precision, that the confusion you've been trying to create about the word "heritability" is counterproductive. Intelligence is passed from one generation to the next. The formula is a mathematical description of regression toward the mean. Is regression toward the mean taking place? Assuming it is, successfully convincing smart people to have more children will result in a population that's smarter than it otherwise would have been. But suppose Weiss is right, and the appearance of regression toward the mean is due entirely to measurement error. In that case, the benefits of a eugenics program would be even greater.

1. You're wrong. 2. The article used the word "heritability" in the equation, as well as in the following sentence: We're talking apples to apples. 3. The article fits in nicely with what I've said--that intelligence is largely genetic, that parental intelligence drives the intelligence of children, etc. The formula which I cited is basically a mathematical definition of regression toward the mean. I've addressed the issue elsewhere. And at least for Weiss, regression toward the mean is simply an artifact of measurement error, rather than a genuinue phenomenon. Certainly it's not something that should be allowed to stand in the way of an otherwise welcome eugenics program.

Now that I know reverse psychology is this easy to use on you, let me offer you a few suggestions. First, please keep the statistics thread open, because it isn't a flame war at all . . .

You don't think he's angry?!?!?!? What on earth planet are you living on? Have you been abducted by space aliens or something--aliens that are using your fingers to transmit their own weird thoughts? This guy has "anger management issues" written all over him. In big neon letters. If you're too blind to see that . . . well, I guess it wouldn't be the first time you were too blind to see something obvious.

Mocking me for what? For believing that genetics play a major role in determining intelligence? That's mainstream science, buddy. Or are you mocking me for saying that the intelligence of biological parents is a strong predictor of the intelligence of children? Because that, too, is something which is supported, and supported very strongly. The predicted I.Q. of a child is determined by the following formula: Expected I.Q. = (average I.Q. of population) + (heritability)^2 * ((average I.Q. of parents) - (average I.Q. of population)) In 1995, the American Psychological Association's task force on "Intelligence: Knowns and Unknowns" concluded that the heritability of I.Q. was "around .75."

The article I read classified feathers according to their complexity. Type 1 feathers were the simplest, and supposedly the first to appear. IIRC, type 1 feathers were down. Over time, animals evolved more complex types of feathers. Type 5 feathers are asymmetric, and are associated strictly with flight. Dinosaurs had type 4 feathers; which is basically a type 5 feather without the asymmetry.

There isn't a single true word in your post. I just see more bull-in-the-china-shop type hostility and anger.

Oh really? Because the study never implied that every last janitor is stupid, nor did it imply that every last nuclear physicist is smart. Weiss clearly believed that there would be a lot more smart people in a group of 100 nuclear physicists than in a group of 100 janitors. Yes, this belief is integral to his study. Yes, he used career choice as part of his method for estimating parental intelligence. But you know what? He also estimated that, for men, this method would be in error 10% of the time. The humility implied by the error term shows that Weiss would have been too wise to dismiss the possibility that two people of humble career origins could produce an intelligent son or daughter.

I see you failed to comprehend the study.

It's comments like this which led to the bull in a china shop reference. How is it possible to have an intelligent discussion about a subtle and complex issue with someone who comes across as enraged and hostile as you? If you go back and reread my post, you'll see that you refuted a straw man. I acknowledged that tens millions of years ago, carbon levels in the Earth's atmosphere were higher than they are today. I pointed out that this gradual decrease in the atmosphere's carbon level is one possible reason for why the world was generally warmer during the days of the dinosaurs than it is today. As for the feathers on dinosaurs, I'm well aware of the fact that at least some predatory bipedal dinosaurs had type-4 feathers. (Type-5 feathers are the most complex; and are found on modern flying birds. Type 4 feathers are basically symmetrical type 5 feathers, and are also found on birds.)