I feel I am better than 75+% of the student population

[iversonian]

[ QUOTE ]
-iversonian

lol at using sat as confirmation. i went to a school way less respectable that NC state, and i know a significant number of students who i would consider smarter than me in my course of study. i have also taken a professionally administered iq test and there is no way that your claim is true based on the relative intelligence of my peers.

[/ QUOTE ]

Anecdotal. Lol at the college administrators as well, then, for using SAT as confirmation. What do they know. And are you challenging the numbers in the table provided by NCSU or my interpretation of them when you say it can't be true?

[ QUOTE ]
There is another thread in here that claims a consultant has raised a verbal SAT score hundreds of points.

[/ QUOTE ]

Again, anecdotal.

[ QUOTE ]
The exams are just not an accurate measure of intelligence.

[/ QUOTE ]

What's your criteria for accurateness? Wikipedia cites a source that says 0.82 correlation. And please try not to work backwards from your conclusions.

[ QUOTE ]
A moderately intelligent person who learns the test well enough can ace that test.

[/ QUOTE ]

Not unless you want free reign to define "learn well enough" however you want, which I don't grant. Good Lord, I don't think even the shadiest test prep center would make that claim.

[ QUOTE ]
The SAT/GRE/GMAT/LSAT exams are very beatable.

[/ QUOTE ]

And Harvard students beat it more often than NCSU students -- and by a wide margin. But I don't need to beat this dead horse all over again. That's what Sklansky's for.

[iversonian]

Edit: I misread "A moderately intelligent person who learns the test..." as "Any moderately..." Not as ridiculous a statement as I had thought. Now it's just irrelevant.

[Fedfan691]

I think people are being a bit nitty about iversonian's post. He was just trying to say that the average harvard student is smarter than the vast vast majority of students at state universities. Sure schools like Berkeley and UVA throw this off, but there are many more easy to get into state universities than the elite state schools. And while there are legacies and athletes at harvard that aren't the brightest students in the country, these kids are still bright yet lie in the bottom statistical range of the harvard population. The average harvard student is insanely smart, hardworking and focused. There are hundreds of thousands of students at state universities, and the number of kids at the elite public universities isn't high enough to make iversonian's claim that unreasonable.

[RyanUSF]

Hard working and focussed does not mean intelligent in terms of pure intellectual prowess though, so don't confuse the two. While there is definitely a correlation between high SAT/ACT scores and a students' intelligence, it's definitely not a direct measure. A hard-working, focussed, semi-intelligent student can outscore a very lazy, but bright student.

[Fedfan691]

[ QUOTE ]
The average harvard student is insanely smart, hardworking and focused. There are hundreds of thousands of students at state universities, and the number of kids at the elite public universities isn't high enough to make iversonian's claim that unreasonable.

[/ QUOTE ]
[ QUOTE ]
Hard working and focussed does not mean intelligent in terms of pure intellectual prowess though, so don't confuse the two. While there is definitely a correlation between high SAT/ACT scores and a students' intelligence, it's definitely not a direct measure. A hard-working, focussed, semi-intelligent student can outscore a very lazy, but bright student.

[/ QUOTE ]

Explain to me where I said they were the same thing? All I was doing in the sentence you dispute other than pointing out that the average student was insanely smart was mentioning the student's other positive characteristics. Also explain to me where I mentioned the SAT or ACT? I never said anything about people's dispute with iversonian's use of the SAT as a measure of intelligence; I was saying there shouldn't be much of an argument about his original point.

[RyanUSF]

My post was in response to iversonian's.

[Fedfan691]

Oh, then FU quick reply button.

[dustyn]

[ QUOTE ]

What's your criteria for accurateness? Wikipedia cites a source that says 0.82 correlation. And please try not to work backwards from your conclusions.

[/ QUOTE ]

Oh come on, do you really believe this? Of course they are highly correlated. Only a very small minority of people are hiring consultants and using years of resources and test prep classes to score very high on the exam. A 5-10% minority (I'm making numbers up) are not enough to skew the overall level of accurateness that much. But just because 90-95% of people lose at poker long term does not mean the game isn't beatable. Similar logic applies here.

[ QUOTE ]
A moderately intelligent person who learns the test well enough can ace that test.

[/ QUOTE ]

I'm saying this: a "normal" above average person who can score a 1150-1250 on the SAT (without significant preparation) can get a score high enough to get into Ivy League schools (think 1400s+) if they hire consultants, take test prep classes and study their ass off for the exam. Obviously there are people (think Sklansky) that are naturally intelligent and can get a super high score on the SAT with minimal preparation. I'm saying the yuppie kids that are hiring consultants and taking SAT prep classes that cost thousands of dollars are beating the exam. What would they get without all this prep? I'm sure some would score well enough to get into Harvard. But I'm sure most would be in the 1200ish bracket without all the help.

My point is simply that the raw intelligence of the median student at Harvard is greater than a state univ, but it's not 95% of the students there. It's way lower. What is way higher is the access to top flight resources to prepare for the exam and beat it - to propel students from the 1200 bracket to the 1450 bracket. I'm not saying consultants can get morons into Harvard. I'm saying they can get people with 75th percentile IQs 98th percentile SAT scores and get them into Harvard.

[T50_Omaha8]

[ QUOTE ]
What's your criteria for accurateness? Wikipedia cites a source that says 0.82 correlation.

[/ QUOTE ]When the median score for the (old) SAT was just over 1,000, I'm not sure the correlation coefficient means too incredibly much. Just because two variables are correlated over their entire domains doesn't mean they are well correlated over all subsets of these domains.

And in this instance it would make very good sense that the range over which the SAT fails to serve as an intelligence test is at the higher end of the spectrum.

[ QUOTE ]
Lol at the college administrators as well, then, for using SAT as confirmation. What do they know.

[/ QUOTE ] They know that it doesn't only test intelligence, but also other things they are interested in, namely a student's ability to grind out memorization of a few hundered big words.

And you should really do a little reading on network externalities if you're going to continue touting the SAT as a good guideline for college admissions. That's why they all use this test.

[iversonian]

dustyn,

The comparison here is between the median Harvard student and the top 95th percentile State U student. What you're saying is that the former group spends far more resources than the latter group in preparing for the SATs. In fact, they outspend the latter group to such a degree that it renders any comparison of scores between the two invalid. Especially considering (however much benefit you think test prep confers) that the marginal benefit gets smaller and smaller as you reach those stratospheric scores, it must be a ridiculously large gap.

Do you have any reason to believe that that's true, besides that you want it to be true? It's certainly not obvious to me, at least, so you'll have to (sigh, I know, it's a pain) make a case for it. Point to a news story or anything that took a quantitative measure of time/money spent by elite school students vs. not so elite school students. And remember, we're talking about the most motivated of those NOESS's, not the median bum slacker student.

[ QUOTE ]
I'm saying this: a "normal" above average person who can score a 1150-1250 on the SAT (without significant preparation) can get a score high enough to get into Ivy League schools (think 1400s+) if they hire consultants, take test prep classes and study their ass off for the exam.

[/ QUOTE ]

I've heard of cases like that happening, too. It's still a long way from -- the median Harvard student (1490-ish) is generally something like a 1200 who took ridiculous amounts of classes financed by his rich parents.

BTW, what does everyone else who's on the other side of the fence think that figure should be? Median Harvard = X% State U? Ok, less than 95. How much less?

[ QUOTE ]
Oh come on, do you really believe this? Of course they are highly correlated. Only a very small minority of people are hiring consultants and using years of resources and test prep classes to score very high on the exam. A 5-10% minority (I'm making numbers up) are not enough to skew the overall level of accurateness that much. But just because 90-95% of people lose at poker long term does not mean the game isn't beatable. Similar logic applies here.

[/ QUOTE ]

It's a good argument, and I agree that on (overall) average, the SAT scores of Harvard students overstates their intelligence compared with the average school, but given that we're talking about a selected sample of the brightest students, I doubt that there is much skew in that overstatement (of SAT as a proxy for IQ) between Harvard students and State U students.

[ QUOTE ]
Just because two variables are correlated over their entire domains doesn't mean they are well correlated over all subsets of these domains.

And in this instance it would make very good sense that the range over which the SAT fails to serve as an intelligence test is at the higher end of the spectrum.

[/ QUOTE ]

True, the SAT wasn't designed to get very granular scores at the extremes. E.g., something like 2% get perfect math scores. It may be true that the correlation is weaker at the upper extreme. However, you're bringing this up to rebut my argument which was 'yes they are a valid measure, see here' to rebut dustyn's 'SATs are not a valid measure'. For your argument to be meaningful, it has to show that the relationship b/t the two breaks down to such a degree at the extremes that it is no longer a valid measure. It does not do that. It merely posits that, even if .82, it may be less than that when considering only the range of interest.

Though it may be less reliable, it is probably reliable enough (I would think so, given that Harvard et al relies upon it), and I would also say that may even be less biased by the test-prep effect at the upper extremes. The SAT is most coachable at around the 1000 level, which is where the main focus is on increasing vocab base and studying up on remedial math. At the 1400 level, you have to teach students how to tackle math problems of a kind they've never seen before and to analyze writing samples for its true underlying meaning. It's less about "tricks" and more about... being smarter.

[ QUOTE ]
They know that it doesn't only test intelligence, but also other things they are interested in, namely a student's ability to grind out memorization of a few hundered big words.

[/ QUOTE ]

They really actively select for students who are good at memorizing a lot of stuff? News to me.