Regression toward the mean

[Bungee Jumper]

Its pointless. He still cant even understand what he's reading, no does he know the proper statistical definition. You've explained the above statement to him hundreds of times, and he still cant comprehend it. He isnt going to any time soon.

 

Until he can successfully learn the definitions of error and variance, theres no hope.

873290[/snapback]

 

It's not pointless. Like I said, his delusions entertain me. You've got to love how he posts ten sources that prove my point, along with statements of "this proves you were wrong." I need to invent a new phrase to describe his consistent inability to see things as they are, but rather as the complete opposite of what they are. Let's call it "conceptual dyslexia".

 

I believe his response to this will be along the lines of: "I won! I won! I won!" Accompanied by much diaper-pissing.

[Chilly]

So, if removal of error causes regression toward the mean, does this mean the only way to be perfect is to actually be imperfect?

 

Or is this a hippie world where everyone is perfect, we just can't measure it yet?

[Orton's Arm]

1) It's coming from a Stanford web page, which makes it a little easier to ridicule.

2) It contains no math, which makes it a LOT easier to ridicule.

3) I don't need to ridicule it.  You simply don't understand what that's saying and what it means: in a test that has a certain measure of error, people with extreme amounts of error will have less extreme amounts of error upon retesting as the error regresses to the mean.  This is not the same as regression to the mean of the population.  That's what everything you've linked to today has said.  That's what I've said.  That's ENTIRELY different from what you've been saying.

The Stanford quote said exactly what I've been saying all along: someone who scores a 140 on an I.Q. test will tend to get a somewhat lower score upon being retested. I've made my position on this so abundantly clear over the past 50 pages that you can't now pretend that the Stanford website is communicating a different message than I've been communicating. But just in case the Stanford quote didn't quite make it through your thick skull, the Berkeley quote should do the trick:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test.

I may not have been able to persuade very many people over the last 50 pages, but at least I gave them a very clear idea as to where I stand. This means your efforts to redefine what I've been saying are futile. Now it's become clear that Berkeley, Stanford, the University of Chicago, and other credible sources support me. Your refusal to acknowledge defeat even at this point damages only your own credibility.

[Wraith]

Once again, you guys are arguing for the sake of arguing. Hell, you're practically arguing the same side.

 

HA: "The sky is blue!"

 

BJ: "No! The cloud's are white!"

[Orton's Arm]

Once again, you guys are arguing for the sake of arguing. Hell, you're practically arguing the same side.

 

HA: "The sky is blue!"

 

BJ: "No! The cloud's are white!"

873346[/snapback]

You're giving Bungee Jumper entirely too much credit. Consider the I.Q. test score example, which I've repeatedly given, and which he's repeatedly ridiculed. Take someone who scored a 140 on an I.Q. test. This person could be a true 140, a lucky 130, or an unlucky 150. Of the latter two possibilities, the lucky 130 one is more likely than the unlucky 150 one, because there are more 130s available for getting lucky, than there are 150s available for getting unlucky. If you were to take a group of people who scored a 140 on an I.Q. test, and ask them to retake the test, the average for the group would be less than 140.

 

Bungee Jumper has ridiculed my intelligence and my knowledge of statistics for saying the things in the above paragraph. Happily, I was able to find a Stanford article which said the same thing I've been saying. Had Bungee Jumper acted like a gentleman in our debate, I'd allow him to leave with dignity. His conduct toward me has been so utterly despicable that, having won this debate, I'm not going to let him off the hook by turning a blind eye toward the errors of his position.

[Ramius]

the Berkeley quote should do the trick:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test.

I may not have been able to persuade very many people over the last 50 pages, but at least I gave them a very clear idea as to where I stand. This means your efforts to redefine what I've been saying are futile. Now it's become clear that Berkeley, Stanford, the University of Chicago, and other credible sources support me. Your refusal to acknowledge defeat even at this point damages only your own credibility.

873335[/snapback]

 

How in the hell does a positive correlation between test scores in any way tell that the person's second score is going to automatically be lower than their first? That may be the most assinine thing you've posted yet, which is definitely saying something. The 2 things have absolutely NOTHING to do with one another.

[Orton's Arm]

How in the hell does a positive correlation between test scores in any way tell that the person's second score is going to automatically be lower than their first? That may be the most assinine thing you've posted yet, which is definitely saying something. The 2 things have absolutely NOTHING to do with one another.

873366[/snapback]

You are aware that you're ridiculing a quote from Berkeley, right? You've been acting like an ignorant loudmouth throughout the regression toward the mean debate, but this takes the cake.

[Wraith]

How in the hell does a positive correlation between test scores in any way tell that the person's second score is going to automatically be lower than their first? That may be the most assinine thing you've posted yet, which is definitely saying something. The 2 things have absolutely NOTHING to do with one another.

873366[/snapback]

 

I think your confusion is caused by not getting the full context from where this quote was taken.

 

The fact that there is a positive correlation between the test and the retest says nothing about how a person is likely to score closer to the mean on a retest, as you pointed out.

 

HOWEVER, the positive correlation does mean that "...individuals who score much higher than average on one test tend to score above average...on the other test."

 

Make sense?

[Orton's Arm]

I think your confusion is caused by not getting the full context from where this quote was taken.

 

The fact that there is a positive correlation between the test and the retest says nothing about how a person is likely to score closer to the mean on a retest, as you pointed out.

 

HOWEVER, the positive correlation does mean that "...individuals who score much higher than average on one test tend to score above average...on the other test."

 

Make sense?

873423[/snapback]

You're wasting your time. If he doesn't understand this phenomenon after having it explained to him for 50 pages, he's never going to get it.

[Wraith]

You're wasting your time. If he doesn't understand this phenomenon after having it explained to him for 50 pages, he's never going to get it.

873452[/snapback]

 

It has nothing to do with your phenomenon. Your quote from Berkely seemed to imply that a positive correlation between the test score and the retest score leads to regression towards the mean. That implication is bizarre and wrong.

 

I am simply pointing out that the snippet from Berkely was saying that despite your phenomenon, people who scored very high still tend to score above average on a retest, due to the positive correlation between test and retest.

[Bungee Jumper]

The Stanford quote said exactly what I've been saying all along: someone who scores a 140 on an I.Q. test will tend to get a somewhat lower score upon being retested. I've made my position on this so abundantly clear over the past 50 pages that you can't now pretend that the Stanford website is communicating a different message than I've been communicating.

 

No, you've been calling it "regression toward the mean", specifically toward the population mean. And you're wrong. It's not. Stop changing your story and thinking no one will notice; generally, most of the people here have better reading comprehension skills than you, and won't be caught out by your "That's not what I said!" act.

 

But just in case the Stanford quote didn't quite make it through your thick skull, the Berkeley quote should do the trick:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test.

 

Can you even define "correlation"? You don't even know what that quote from Berekley means, do you? If you did, you'd see it's bull sh--. Care to guess why, specifically? (Hint: describe the meaning of a correlation coefficient of "1". Then define "tautology".)

 

I may not have been able to persuade very many people over the last 50 pages, but at least I gave them a very clear idea as to where I stand. This means your efforts to redefine what I've been saying are futile. Now it's become clear that Berkeley, Stanford, the University of Chicago, and other credible sources support me. Your refusal to acknowledge defeat even at this point damages only your own credibility.

873335[/snapback]

 

You've been very clear. You just haven't been consistent. Or you have, in that you've consistently changed your argument when cornered by facts in a vain attempt to appear correct. Now you've taken it to the point of posting stuff you don't even understand and shouting "But it's from Stanford! I won!"

 

Well...sorry Sparky. First off, this isn't about winning to me. It's not a competition, it's about watching you flop around in mathematical ignorance like a landed fish. Second off...it's easy to say "I won", but it's tough to win any competition when you're flopping around in mathematical ignorance like a gutted fish.

[Bungee Jumper]

It has nothing to do with your phenomenon. Your quote from Berkely seemed to imply that a positive correlation between the test score and the retest score leads to regression towards the mean. That implication is bizarre and wrong.

 

I am simply pointing out that the snippet from Berkely was saying that despite your phenomenon, people who scored very high still tend to score above average on a retest, due to the positive correlation between test and retest.

873477[/snapback]

 

You mean...that Berkely snippet doesn't mean what HA thinks it means?

 

I thought it was a brilliant piece of double-talk, myself..."Because of positive correllations between test and retest scores, people who score high the first time tend to score almost as high the second time." In other words, "Because of positive correlation, correlation is positive." Yeah, no sh--...

 

Berkely's also known for inventing LSD. Doesn't seem all that coincidental, does it?

[Orton's Arm]

It has nothing to do with your phenomenon. Your quote from Berkely seemed to imply that a positive correlation between the test score and the retest score leads to regression towards the mean. That implication is bizarre and wrong.

 

I am simply pointing out that the snippet from Berkely was saying that despite your phenomenon, people who scored very high still tend to score above average on a retest, due to the positive correlation between test and retest.

873477[/snapback]

Here's the quote again:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test.

The quote directly addresses the phenomenon I've been describing. A correlation of 0 from one test to the next would imply the test results were obtained due entirely to random chance. A correlation of 1 would imply that test results were obtained due strictly to successful measurement. The Berkeley quote is implying the correlation between scores on a test and a retest is positive (it's measuring something innate) but less than one (there is measurement error involved).

 

In situations like this, the phenomenon I've been describing predicts that those who obtained high scores on the first test will still be above-average upon being retested, but not by quite as much as their initial scores would indicate. That's also what the Berkeley quote said, and that's the phenomenon I've been trying to communicate to Ramius and Bungee Jumper these last 50 pages.

[Bungee Jumper]

Here's the quote again:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test.

The quote directly addresses the phenomenon I've been describing. A correlation of 0 from one test to the next would imply the test results were obtained due entirely to random chance. A correlation of 1 would imply that test results were obtained due strictly to successful measurement. The Berkeley quote is implying the correlation between scores on a test and a retest is positive (it's measuring something innate) but less than one (there is measurement error involved).

 

In situations like this, the phenomenon I've been describing predicts that those who obtained high scores on the first test will still be above-average upon being retested, but not by quite as much as their initial scores would indicate.

873519[/snapback]

 

Which is completely and utterly different from regression toward the mean in the population as a whole, you nitwit.

 

Like I keep saying - and you continually keep misunderstanding - the regression of one person's test scores is due to the regression of ERROR to the mean. Population variance is a totally different concept.

[Orton's Arm]

No, you've been calling it "regression toward the mean", specifically toward the population mean.

I'm not alone: the EPA has been calling it the same thing.

And you're wrong.  It's not.  Stop changing your story and thinking no one will notice; generally, most of the people here have better reading comprehension skills than you, and won't be caught out by your "That's not what I said!" act.

I'm not sure why on earth you'd expect anyone to believe that bit about me changing my story. Earlier this page, I gave that I.Q. test example with which the PPP board is now no doubt familiar--someone who scored a 140 on an I.Q. test is more likely to be a lucky 130 than an unlucky 150. Hence, if a group of people who scored a 140 the first time around chose to retake the test, that group's average on the retest would be less than 140. I've been saying this from the beginning of the debate. Why would I feel a need to change this message at the very moment when I've found numerous websites that back me up?

First off, this isn't about winning to me.

Do you honestly think anyone on these boards is stupid enough to believe a line like that?

[Ramius]

It has nothing to do with your phenomenon. Your quote from Berkely seemed to imply that a positive correlation between the test score and the retest score leads to regression towards the mean. That implication is bizarre and wrong.

 

I am simply pointing out that the snippet from Berkely was saying that despite your phenomenon, people who scored very high still tend to score above average on a retest, due to the positive correlation between test and retest.

873477[/snapback]

 

Thanks, i know what positive correlation is. Berkeley was implying that the positive correlation had some bearing on the directionality of the scores when the test was re-taken.

 

That was what i was questioning, because as you stated, that implication is bull sh--. Holcombs arm is just too stupid to realize this.

[Ramius]

You are aware that you're ridiculing a quote from Berkeley, right? You've been acting like an ignorant loudmouth throughout the regression toward the mean debate, but this takes the cake.

873387[/snapback]

 

Once again, this response demonstrates your utter lack of reading comprehension on this entire topic.

 

I point out the ridiculous nature of an implication from berkeley, and, instead of understanding a concept, you simply state "its from berkeley, its true!" The is typical of someone who doesnt understand what they are trying to explain, aka you in these threads.

[Orton's Arm]

Once again, this response demonstrates your utter lack of reading comprehension on this entire topic.

 

I point out the ridiculous nature of an implication from berkeley, and, instead of understanding a concept, you simply state "its from berkeley, its true!" The is typical of someone who doesnt understand what they are trying to explain, aka you in these threads.

873542[/snapback]

Here's the Berkeley quote again, this time with more explanation:

In most test-retest situations, the correlation between scores on the test and scores on the re-test is positive, so individuals who score much higher than average on one test tend to score above average, but closer to average, on the other test. . . . Similarly, individuals who are much lower than average in one variable tend to be closer to average in the other (but still below average). Those who perform best usually do so with a combination of skill (which will be present in the retest) and exceptional luck (which will likely not be so good in a retest). Those who perform worst usually do so as the result of a combination of lack of skill (which still won't be present in a retest) and bad luck (which is likely to be better in a retest). . . . A particularly high score could have come from someone with an even higher true ability, but who had bad luck, or someone with a lower true ability who had good luck. Because more individuals are near average, the second case is more likely; when the second case occurs on a retest, the individual's luck is just as likely to be bad as good, so the individual's second score will tend to be lower. The same argument applies, mutatis mutandis, to the case of a particularly low score on the first test.

The message in that quote should sound familiar to you. I've been saying it for 50 pages.

[Bungee Jumper]

I'm not sure why on earth you'd expect anyone to believe that bit about me changing my story.

 

Uhhh...because they can read?

 

Do you honestly think anyone on these boards is stupid enough to believe a line like that?

873529[/snapback]

 

Why on earth not? I don't need to "win". I'm right, and you're an idiot. There's nothing for me to "win"...no matter what happens, I'm still right and you're still an idiot.

[Bungee Jumper]

Once again, this response demonstrates your utter lack of reading comprehension on this entire topic.

 

I point out the ridiculous nature of an implication from berkeley, and, instead of understanding a concept, you simply state "its from berkeley, its true!" The is typical of someone who doesnt understand what they are trying to explain, aka you in these threads.

873542[/snapback]

 

But it's so much !@#$ing fun to read, i'n't it?