Regression toward the mean

[Orton's Arm]

I'm tired of the regression toward the mean discussion, so here are a number of links with which to put this issue to bed.

 

http://www.socialresearchmethods.net/kb/regrmean.htm

 

"Assume that the sample for your study was selected exclusively from the low pretest scorers. . . . the sample's mean appears to regress toward the mean of the population from pretest to posttest."

 

http://www.ruf.rice.edu/~lane/stat_sim/reg...mean/index.html

 

Simulation of regression toward the mean.

 

http://www.drjimtaylor.com/homtemplate/sophomoreslump.html

 

"An alternative explanation that has been offered for the phenomenon of the sophomore slump is that the decline in performance is a function of a regression toward the mean (Gilovich, 1984; Nisbett, Krantz, Jepson, & Kunda, 1983), that is, a statistical tendency of extreme scores to move toward the group mean (Campbell & Stanley, 1963). From this perspective, outstanding rookie performances are likely to regress toward their actual level of ability (Gilovich, 1984)."

 

http://www.ipmaac.org/acn/aug96/stat.html

 

"A group of children took a reading test. Those who scored more than one standard deviation below the mean were put into a special reading program. Students in the reading program took the reading test again at the end of the semester and, on the average, the scores were higher. A matched t-test showed a significant difference between the means.

 

The reading teacher, George Bernard Phonics, claimed that this showed that the program did some good. Is that true? You know, I have to say: 'No, the procedure described does not show the program did any good whatsoever.' . . .

 

George 'learned' about regression toward the mean in his first statistics course, but forgot about it right after the final exam."

 

These four links are in addition to the two I've already supplied. I invite anyone who'd like to dispute this issue to provide even a single link from a credible source to support his or her view.

[Bungee Jumper]

I'm tired of the regression toward the mean discussion, so here are a number of links with which to put this issue to bed.

 

http://www.socialresearchmethods.net/kb/regrmean.htm

 

"Assume that the sample for your study was selected exclusively from the low pretest scorers. . . . the sample's mean appears to regress toward the mean of the population from pretest to posttest."

 

http://www.ruf.rice.edu/~lane/stat_sim/reg...mean/index.html

 

Simulation of regression toward the mean.

 

http://www.drjimtaylor.com/homtemplate/sophomoreslump.html

 

"An alternative explanation that has been offered for the phenomenon of the sophomore slump is that the decline in performance is a function of a regression toward the mean (Gilovich, 1984; Nisbett, Krantz, Jepson, & Kunda, 1983), that is, a statistical tendency of extreme scores to move toward the group mean (Campbell & Stanley, 1963). From this perspective, outstanding rookie performances are likely to regress toward their actual level of ability (Gilovich, 1984)."

 

http://www.ipmaac.org/acn/aug96/stat.html

 

"A group of children took a reading test. Those who scored more than one standard deviation below the mean were put into a special reading program. Students in the reading program took the reading test again at the end of the semester and, on the average, the scores were higher. A matched t-test showed a significant difference between the means.

 

The reading teacher, George Bernard Phonics, claimed that this showed that the program did some good. Is that true? You know, I have to say: 'No, the procedure described does not show the program did any good whatsoever.' . . .

 

George 'learned' about regression toward the mean in his first statistics course, but forgot about it right after the final exam."

 

These four links are in addition to the two I've already supplied. I invite anyone who'd like to dispute this issue to provide even a single link from a credible source to support his or her view.

858325[/snapback]

 

The argument isn't whether or not it exists. The argument is whether or not it's caused by "error", as you mistakenly and ignorantly claim, or by the statistical dispersion of data according to a given probability distribution, as is claimed by everyone else in the known universe except yourself.

[ieatcrayonz]

Cat.house

[RuntheDamnBall]

Statistics deathmatch begins in... 3....2...1...

[Taro T]

So, does this officially put the number of threads regarding HA's math skills into double digits, or were we already there?

[Bungee Jumper]

So, does this officially put the number of threads regarding HA's math skills into double digits, or were we already there?

858504[/snapback]

 

Trick question. HA has no math skills.

[Taro T]

Trick question.  HA has no math skills.

858509[/snapback]

I thought about adding the phrase "or lack thereof" but figured it was understood (by nearly everyone).

[Ramius]

Kelly Holcomb sucks. He's a decent backup, but thats it. He's got no arm, and cant make throws in RWS december winds.

[Kelly the Dog]

http://www.ipmaac.org/acn/aug96/stat.html

 

"A group of children took a reading test. Those who scored more than one standard deviation below the mean were put into a special reading program. Students in the reading program took the reading test again at the end of the semester and, on the average, the scores were higher. A matched t-test showed a significant difference between the means.

 

The reading teacher, George Bernard Phonics, claimed that this showed that the program did some good. Is that true? You know, I have to say: 'No, the procedure described does not show the program did any good whatsoever.' . . .

 

George 'learned' about regression toward the mean in his first statistics course, but forgot about it right after the final exam."

 

These four links are in addition to the two I've already supplied. I invite anyone who'd like to dispute this issue to provide even a single link from a credible source to support his or her view.

858325[/snapback]

I think it is safe to say that any serious person, anywhere, with any point to make, should not use as his prime example a study based on the findings of a friggin' reading teacher named George Bernard Phonics.

 

I'm just sayin'.

 

Not to mention that the article was written by a dead guy who used to write Peanuts.

[Orton's Arm]

The argument isn't whether or not it exists.  The argument is whether or not it's caused by "error", as you mistakenly and ignorantly claim, or by the statistical dispersion of data according to a given probability distribution, as is claimed by everyone else in the known universe except yourself.

858339[/snapback]

Suppose you were to take a very large random sample of people, and have them all take I.Q. tests. Randomly choose 100 people who got a 140 on the test, and have them retake it. I contend that the expected average score for those 100 people on the retake will be less than 140. Do you agree or disagree?

[X. Benedict]

I think it is safe to say that any serious person, anywhere, with any point to make, should not use as his prime example a study based on the findings of a friggin' reading teacher named George Bernard Phonics.

 

I'm just sayin'.

 

Not to mention that the article was written by a dead guy who used to write Peanuts.

858603[/snapback]

Good Grief.

 

Sometimes I lie awake at night, and I ask, "Where have I gone wrong?" Then a voice says to me, "This is going to take more than one night."

 

Charlie Brown

[Bungee Jumper]

Suppose you were to take a very large random sample of people, and have them all take I.Q. tests. Randomly choose 100 people who got a 140 on the test, and have them retake it. I contend that the expected average score for those 100 people on the retake will be less than 140. Do you agree or disagree?

858621[/snapback]

 

I disagree with the methodology. Randomly choose a non-random set of data? You don't even understand the meaning of "random".

[Orton's Arm]

I disagree with the methodology.  Randomly choose a non-random set of data?    You don't even understand the meaning of "random".     

858647[/snapback]

Nice way to weasel out of the question. Suppose 1000 people had scored a 140 on the test. Of those 1000, you'd randomly choose 100 to retake the test.

 

I can see why you're afraid to answer this question. If I'd committed myself to an erroneous view as strongly as you did, I'd be afraid of being too specific myself.

[Bungee Jumper]

Nice way to weasel out of the question. Suppose 1000 people had scored a 140 on the test. Of those 1000, you'd randomly choose 100 to retake the test.

 

I can see why you're afraid to answer this question. If I'd committed myself to an erroneous view as strongly as you did, I'd be afraid of being too specific myself.

858665[/snapback]

 

I'd answer the question if it made sense. It doesn't. You can't randomly choose a non-random sample, no matter if you "randomly" choose 100 from a general population, or 100 from a non-random subset.

 

What you really mean to ask is: Suppose you take a very large randomly chosen sample of people and administer an IQ test. How will the people that score 140 score on the second administration of the test. And I'll tell you EXACTLY how they'll score, as soon as I get a chance to do the math. Shall we agree on a mean of 100, and a standard deviation of 15, with a measurement error of mean 0 and standard deviation of...let's say 3?

[Orton's Arm]

I'd answer the question if it made sense.  It doesn't.  You can't randomly choose a non-random sample, no matter if you "randomly" choose 100 from a general population, or 100 from a non-random subset.

 

What you really mean to ask is: Suppose you take a very large randomly chosen sample of people and administer an IQ test.  How will the people that score 140 score on the second administration of the test.  And I'll tell you EXACTLY how they'll score, as soon as I get a chance to do the math.  Shall we agree on a mean of 100, and a standard deviation of 15, with a measurement error of mean 0 and standard deviation of...let's say 3?

858693[/snapback]

Sounds fair to me.

[justnzane]

Suppose you were to take a very large random sample of people, and have them all take I.Q. tests. Randomly choose 100 people who got a 140 on the test, and have them retake it. I contend that the expected average score for those 100 people on the retake will be less than 140. Do you agree or disagree?

858621[/snapback]

 

from my psychology professor last year, "IQ does not change much throughout time, so a person who scores a 125 as a teen should expect to score around 125 as an adult"

[Orton's Arm]

from my psychology professor last year, "IQ does not change much throughout time, so a person who scores a 125 as a teen should expect to score around 125 as an adult"

859154[/snapback]

This isn't a question of I.Q. changing over time. Suppose someone was to take an I.Q. test, and score a 140. One month later, suppose this person were to retake the I.Q. test. Clearly this person's underlying I.Q. is still the same, unless it was decreased by excessive exposure to these boards. But regression toward the mean indicates that this person's measured I.Q. will probably be lower the second time he or she takes the test. In other words, that initial 140 score was probably based mostly on intelligence, but also on a little luck.

[justnzane]

This isn't a question of I.Q. changing over time. Suppose someone was to take an I.Q. test, and score a 140. One month later, suppose this person were to retake the I.Q. test. Clearly this person's underlying I.Q. is still the same, unless it was decreased by excessive exposure to these boards. But regression toward the mean indicates that this person's measured I.Q. will probably be lower the second time he or she takes the test. In other words, that initial 140 score was probably based mostly on intelligence, but also on a little luck.

859166[/snapback]

 

The point of what i was saying is that if you take it once and take it a second time, you're score will not change by much, if at all Again, you amaze me with your lack of intellect.

[ieatcrayonz]

This isn't a question of I.Q. changing over time. Suppose someone was to take an I.Q. test, and score a 140. One month later, suppose this person were to retake the I.Q. test. Clearly this person's underlying I.Q. is still the same, unless it was decreased by excessive exposure to these boards. But regression toward the mean indicates that this person's measured I.Q. will probably be lower the second time he or she takes the test. In other words, that initial 140 score was probably based mostly on intelligence, but also on a little luck.

859166[/snapback]

So that 85 you scored would probably turn into a 79 if you took it again?

 

Do I have it now?

[Orton's Arm]

The point of what i was saying is that if you take it once and take it a second time, you're score will not change by much, if at all  Again, you amaze me with your lack of intellect.

859226[/snapback]

Since you don't know what we're discussing, you can read up on it here

A person who scored 750 out of a possible 800 on the quantitative portion of the SAT takes the SAT again (a different form of the test is used). Assuming the second test is the same difficulty as the first and that there was no learning or practice effect, what score would you expect the person to get on the second test? The surprising answer is that the person is more likely to score below 750 than above 750; the best guess is that the person would score about 725. If this surprises you, you are not alone. This phenomenon, called regression to the mean, is counter intuitive and confusing to many professionals as well as students.