Average IQ at Harvard?

Colleges and Universities A-Z Harvard University
121

<p>
[quote]
The difficulty of getting into Harvard as compared to MIT comes from its greater fame and popularity, and from the large number of applicants who also can generate the noncognitive, or less-cognitive, credentials that MIT students may not have. The non-well-rounded, insufficiently athletic or communitarian or minority, applicants who went to MIT, Caltech, Berkeley and UCLA make a comeback to Harvard for the PhD program, where it now helps to be Top Geek

[/quote]
</p>

<p>That's not really what I'm talking about. My point is this. Take a brilliant high school senior. He gets into both Harvard and MIT. Which one will he choose? Probably Harvard. Why? Maybe it's due to pressure from his parents. Maybe because he just wants to impress his friends. Maybe it's because he rationally determines that he doesn't really know what he wants to do later in his life, so to be risk-averse, he chooses the school with the strongest brand name (an entirely rational decision, IMO). </p>

<p>But the point is, Harvard wins the cross-admit battle with MIT for undergrads, as it does with every other schools. And as I have shown with regards to SAT scores, there is no evidence that MIT ends up with a more brilliant undergrad student body. </p>

<p>PhD is different. If you are getting a Phd, you generally know EXACTLY what you want to do with your career, so you are usually not swayed by general school brand names. Like I said, there is very little reason to choose Harvard for an engineering PhD if you can get into MIT. In fact, I am convinced that a major reason why Harvard engineering is ranked so relatively low is because the presence of MIT sucks all of the oxygen out of the room. As far as the other sciences are concerned, MIT's PhD programs are ranked at least as high, and usually higher. The same is true of economics. Generally, as a PhD student, you want to attend the strongest program in your particular discipline. Somebody admitted to the PhD programs at Harvard and MIT in engineering is almost certainly going to pick MIT. Somebody admitted to the PhD program in physics or economics at both Harvard and MIT is going to have a tough decision, but marginally prefer MIT, however marginally. </p>

<p>But the point, again, is not to romanticize the strength of Harvard's PhD programs, particularly its technical ones. Harvard has extremely strong competition for the top technical PhD students.</p>

122

<p>This discussion fails to address the varying types of intelligence as well as what I perceive to be the respective goals of Harvard and MIT. From the people I have known who went to MIT, it seems like MIT is looking for more raw, intellectual power. MIT wants undergrads who are going to make remarkable gains in the hard sciences and mathematics (i.e. developing a new postulate). On the other hand, I think Harvard is in the interest of finding more well-rounded students who are of course intelligent, but whose intelligence relates more to linguistics, introspection and social skills. To put it in the simplest terms, Harvard is looking for future leaders, revolutionaries, politicians, humanitarians, etc. whereas MIT is looking for the soon-to-be brilliant mathematicians and physicists.</p>

<p>To tie this back to the OP, I would guess MIT students have higher IQs based on our current IQ tests, which seem to relate more to mathematical and numerical logic. But I'd say IQ is naturally less of a requisite for Harvards goals.</p>

<p>On a last note, I'll entertain my bias a bit more and cater to Harvard's defense by pointing out that if it wanted Harvard could just take every perfect SAT/ACT applicant and easily have the highest mean IQ of any school.</p>

123

<p><a href="sakky%20wrote:">quote</a>there is essentially no difference in the SAT scores of the undergraduate students pools at Harvard and MIT, particularly at the 'lower pool'. The 25th/75th percentile of the SAT scores at MIT is 1410/1560. At Harvard, it's 1400/1580.

[/quote]
</p>

<p>Your data are bogus, as covered in CC many times. Those are not the 25/75 percentiles of combined SAT but the sum of the intervals for math and verbal separately; that's math and verbal scores of different students added together. Correct data are discussed below, and reveal the statistical ranking to be the expected one, Caltech > MIT > Harvard.</p>

<p>The closest thing to a head-to-head comparison of combined math and verbal percentiles between Harvard and MIT is ACT composite scores, where the reported "25th percentile" at Harvard is the 13th percentile at MIT (but we don't know the actual percentage of 30+ ACT's at Harvard to compare with 88 at MIT). That's consistent with the idea that the bottom of the pool is larger at Harvard. Harvard also admits more people who rank below the top 10 and top 25 percent of their class (MIT admits none from the latter category); the class rank data are not conclusive but also point to a weaker bottom of the pool at Harvard.</p>

<p>The SAT data also have Caltech > MIT > Harvard, with test-score cutoff of the pool setting the ranking. The bottom of the pool is killed off completely at Caltech, partly at MIT, and significantly less so at Harvard. The correct 25-75 percentile ranges, from 2007 US News rankings (2005-6 CDS), are: </p>

<p>700-780 V 770-800 M ??-?? ACT CalTech
690-770 V 740-800 M 31-34 ACT MIT
700-790 V 700-790 M 31-34 ACT Yale
700-790 V 700-790 M 30-34 ACT Harvard
690-770 V 690-790 M ??-?? ACT Princeton
690-770 V 690-780 M 28-33 ACT Stanford</p>

<p>The Common Data Sets of all these schools but Harvard are online. Yale is essentially the same as or slightly stronger than Harvard according to the US News (CDS derived) data, so one can use its CDS as a reasonable approximation to Harvard, and check against the data for Princeton, Stanford and other Ivies. The result:</p>

<p>Caltech is as strong in the verbal SAT as all other schools, and by some measures is the strongest (percent at 700+, no scores under 600). In the math SAT it exceeds MIT, and thus all other schools as well.</p>

<p>MIT dominates all schools other than Caltech in the math SAT; on verbal SAT it is better than Stanford and Brown, and somewhat below Princeton. MIT scores below Harvard and Yale on the verbal SAT by about 10-20 points while about 40 points higher on the math SAT. The extent of MIT and Caltech math dominance is understated because 800 SAT is below their 75th percentile (for Caltech, 700 is their 4th percentile and 770 is their 25th); for all other schools the 25/75 range of SAT is 90-100 points across but for the tech schools this hits the 800 maximum and we don't see the true spread.</p>

<p>Ignoring Caltech which has no apparent weakness from data at this resolution, here is the SAT 100-point tranche comparison for MIT with Yale (equal to or stronger than Harvard for humanities these days). They have the same proportions of students with very low verbal scores, below 600 and 500, which probably can be attributed to international students or extreme cases of legacies, development etc. Setting those aside, the chance that a given 600+ verbal score is weak (below 700) is about 50 percent higher at MIT than Yale, whereas the chance that a given math score is weak is 3 times higher at Yale than at MIT.</p>

124

<p>Not to mention that the math score is a significantly better indicator of "IQ", in its traditional form, than the verbal score.....which tilts it even greater in the direction of MIT/Cal Tech at the undergrad level.</p>

125

<p>
[quote]
On a last note, I'll entertain my bias a bit more and cater to Harvard's defense by pointing out that if it wanted Harvard could just take every perfect SAT/ACT applicant and easily have the highest mean IQ of any school.

[/quote]
Interesting hypothesis, I doubt it could have a higher IQ than Cal Tech though. Harvard would have to stop admittance into lower IQ majors...Tech schools have lower proportions in low IQ majors to begin with so it's quite easy for them to rachet up their average IQs because they can flex their competitive advantage in Engineering over Harvard (of which the Harvard name has no cache).</p>

126

<p><a href="sakky%20wrote:">quote</a>Harvard wins the cross-admit battle with MIT for undergrads, as it does with every other schools.

[/quote]
</p>

<p>Harvard also wins against Caltech which has a statistically stronger set of undergrads, so it's not clear what your point is.</p>

<p>The revealed preference study (cross-admit tournament) that you mention is at</p>

<p><a href="http://www.economics.harvard.edu/faculty/hoxby/papers/revealedprefranking.pdf%5B/url%5D"&gt;http://www.economics.harvard.edu/faculty/hoxby/papers/revealedprefranking.pdf&lt;/a&gt;&lt;/p>

<p>Note the graphs (p.7) of admissions rate by SAT percentile at Harvard, MIT, Yale and Princeton. At MIT, admissions rate increases steadily with SAT score, with the top of the range being sampled at a rate much higher than HYP. At Harvard, there is no difference between SATs in the 91 to 98th percentiles, though at the very top there is a sharp climb in selection rates. This matches what is reported informally about Harvard admissions: at the top 10-15 percent of the enrolled pool, selection is purely academic (so we see the same increase in admission rates as at MIT). For the solid but not amazing candidates in the rest of the pool, selection is on other and apparently not SAT-correlated factors such as athletics and hobbies. Yale and Princeton play other games but have a dramatic rise in admission rate at the very highest SAT scores as they try to get enough superstars to maintain statistical parity with Harvard.</p>

<p>MIT and Caltech, which use SAT (and thus IQ) loaded selection throughout their applicant pool, will not only oversample the top range compared to Harvard, Yale, and Princeton, but will tend to cut off more of the bottom of the pool. MIT doesn't have 10 percent recruited athletes as Harvard does.</p>

127

<p>Siserune, even if everything you are saying is true, it only indicates that Harvard undergrads are * slightly * worse off than MIT undergrads are (note, neither I nor microsoft never made any comparisons to Caltech, so that's irrelevant). </p>

<p>But how does that carry in a comparative analysis. Like I've said throughout this thread, Harvard has large graduate departments in the humanities and social sciences, relative to MIT. Like I said, you generally don't have to be a genius to get into and survive one of these graduate programs. There are PLENTY of people getting PhD's in English at Harvard, but none at MIT. You don't have to be a genius to get a PhD in English. You have to be well-read, know how to do archival research, and have stamina and creativity. But 'brilliance'? Not unless you're using some definition of 'brilliance' that is far from the mainstream.</p>

<p>Furthermore, like I said, take the entire suite of technical disciplines that microsoft was talking about. MIT has clearly higher ranked graduate programs in engineering, and slightly better programs in all of the sciences. The upshot is that Harvard graduate programs in these technical disciplines are almost certainly not as good, on average, as their counterparts at MIT. This is particularly noteworthy in engineering - I think it's safe to say that the majority of Harvard engineering graduate students would rather be going to MIT, but didn't get in. I certainly know of a quite a few like that. </p>

<p>So even if we relax the analysis to accomodate what you are saying, it still means that there is little evidence that Harvard graduate students are really far more 'brilliant' than Harvard undergrads, which is what microsoft and others have asserted. This is particularly true when you're talking about the technical disciplines. LEt's face it. If you're really in the bottom X% of your class as a Harvard undergrad, you're probably not majoring/concentrating in physics, or engineering, or mathematics. You're almost certainly going to be majoring in something easier - yet the fact is, Harvard has plenty of graduate students who are also in those easy subjects. </p>

<p>But the bottom line is this. I see no evidence to indicate that Harvard graduate students are more 'brilliant' than Harvard undergrads.</p>

128

<p>IQ doesn't really matter.</p>

129

<p>^ agreed, its only what you do with it that does</p>

130

<p>OK...let's get something straight here. Standardized tests, like the SAT or ACT, do not measure IQ at all. The only thing they measure is how well you test. Some test better than others. </p>

<p>I've read many articles and discussions on this controversial topic about SAT/IQ correlation...and the evidence is stacked against standardized tests. They don't measure anything. Go out and research, and you will know what I'm talking about.</p>

<p>I just wanted to get that out in the air. And please, don't turn this thread into a debate about this topic.</p>

131

<p><a href="sakky:">quote</a> even if everything you are saying is true, it only indicates that Harvard undergrads are slightly worse off than MIT undergrads are (note, neither I nor microsoft never made any comparisons to Caltech, so that's irrelevant).

[/quote]
</p>

<p>Caltech, whose superiority you don't contest, is a valuable data point for these comparisons, which is why both microsoft (post #119) and I have discussed Caltech > MIT > Harvard. The first inequality clarifies the second as well as the H grad/undergrad question.</p>

<p>MIT is roughly a statistical midpoint between Harvard and Caltech. This is visible in the 25/75 percentiles above, in the amount of known admissions dilution (Caltech- none, MIT- AA, Harvard- AA and athletics), and other measures. As another example, a good indicator of the level of meritocracy versus games in admission is percentage of Asians enrolled. From the Common Data Sets ("unknown" are the students with no identified race, "guess" is my subjective estimate of the Asian numbers taking this into account):</p>

<p>31.1 Caltech (guess 31.4?, 1.5 unknown)
26.6 MIT (guess 28.5?, 11.1 unknown)
24.3 Stanford (guess 25.7, 5.3 unknown)
18.0 Harvard (guess 20.0?, 7.9 unknown)
13.6 Yale (guess 16?, 11.0 unknown)
12.1 Princeton (guess 12.1, 0.0 unknown)</p>

<p>Excluding Stanford (which is atypical for this metric), the ranking is identical to the statistical ranking by SAT's. Note that at the Ivies, Asian admits average about 50 SAT points higher than ordinary admits, so larger numbers are directly correlated with population SAT as well as being a measure of merit-based admission. Even taking into account the greater popularity of engineering among Asians, it is fair to say that MIT is at least as close to Caltech's admission model as it is to Harvard's.</p>

<p>It is clear that the degree of meritocratic, which is essentially by-the-numbers, admission is what drives the strength of the undergrad enrolled pool, and that the reason Caltech is different is not that its top students are so much better than Harvard or MIT (they aren't), but that Caltech has deleted the bottom of the pool by using far fewer non-academic selectors. </p>

<p>Compared to undergraduate admission, Harvard PhD admission is almost entirely academically based. The 30 percent of internationals at GSAS come from a far stronger population than anything seen at undergraduate admissions; the bottom 10 percent of athletes are gone; the 20-30 percent of minorities are divided by a factor of 2 or 4 or 20 depending on the department. Recommendations are more meaningful and likelier to represent reality. Professors, not the admissions office, determine the admissions. Caltech is the control experiment that tells us what the results of such a procedure are.</p>

132

<p>This last post looks rather impressive.</p>

133

<p>
[quote]
Take a brilliant high school senior. He gets into both Harvard and MIT. Which one will he choose? Probably Harvard.Why? Maybe it's due to pressure from his parents. Maybe because he just wants to impress his friends.

[/quote]
</p>

<p>Sorry I don't buy that. MIT has as good a brand name as Harvard or even better nowadays. Suggesting that Harvard is a top choice for everybody would have meant a 100% yield figure for Harvard. It is not so.</p>

<p>
[quote]
If you are getting a Phd, you generally know EXACTLY what you want to do with your career, so you are usually not swayed by general school brand names.

[/quote]
</p>

<p>No. Many people decide to do PhD because they want to do some research and add to the existing knowledge before moving on to money making business. The school brand name matters immensely for all PhDs. There are many physics PhDs (for example) who move into quantitative finance. Many PhDs start their own business in the silicon valley (esp in stanford). Many people go into consulting (which pay very large salaries to PhDs). That is why school name matters VERY much for these people. And these are the people who will prefer a Harvard PhD degree than a MIT degree as they are not going to be doing research for the rest of their lives. In fact it matters even more those whi decide to become professors. Saying that one who has decided to do a PhD knows exactly what he wants to do with his life is wrong.</p>

<p>
[quote]
Like I said, there is very little reason to choose Harvard for an engineering PhD if you can get into MIT.

[/quote]
</p>

<p>Maybe true if you are an american but not if you are an international student because the "Harvard" name helps more than MIT tag internationally. And it is a fact that in Harvard SEAS most students are of international origin. I personally know people in Harvard who turned down Stanford for Harvard just because the name is bigger internationally.</p>

<p>
[quote]
In fact, I am convinced that a major reason why Harvard engineering is ranked so relatively low is because the presence of MIT sucks all of the oxygen out of the room.

[/quote]
</p>

<p>No. (If that were true why doesnt MIT suck out oxygen from mathematics or economics programs at harvard?) The correct reason is Harvard never gave engineering much attention. Engineering was somehow never intellectually appealing to Harvard. But Harvard has realized its mistake and is concentrating big time on SEAS and I feel Harvard will be a top-notch engineering school within a decade. </p>

<p>
[quote]
Siserune: Harvard also wins against Caltech which has a statistically stronger set of undergrads, so it's not clear what your point is.

[/quote]
</p>

<p>That is true. I dont find that any reason for assuming harvard undergrads are any better than caltech's or MIT's. In fact I think the avergae caltech and MIT undergrad is much sharper.</p>

<p>
[quote]
Like I said, you generally don't have to be a genius to get into and survive one of these graduate programs.

[/quote]
</p>

<p>You don't have to be genius no doubt but you have to be extremely bright academically to get in. We never said Harvard grads are all geniuses. They are just more brilliant than the undergrads because the number of graduate seats are extremely limited and there is a huge competition for those spots on an international level. That makes the average quality of the grads much much higher than undergrads.</p>

<p>
[quote]
If you're really in the bottom X% of your class as a Harvard undergrad, you're probably not majoring/concentrating in physics, or engineering, or mathematics.

[/quote]
</p>

<p>No, really why will someone make the assumption that anyone concentrating in physics, maths will always do well and not be in the bottom rungs. I dont understand.</p>

<p>
[quote]
Harvard has large graduate departments in the humanities and social sciences, relative to MIT. Like I said, you generally don't have to be a genius to get into and survive one of these graduate programs. There are PLENTY of people getting PhD's in English at Harvard, but none at MIT. You don't have to be a genius to get a PhD in English.

[/quote]
</p>

<p>I dont agree with your association of low IQ with the humanities and social sciences. There are many brilliant people who chose to study humanities just for the love of the subject (among other reasons) and in all probability the strongest of them are in Harvard grad programs. </p>

<p>As siserune said, "The 30 percent of internationals at GSAS come from a far stronger population than anything seen at undergraduate admissions" and that is very true.</p>

134

<p>
[quote]
Sorry I don't buy that. MIT has as good a brand name as Harvard or even better nowadays. Suggesting that Harvard is a top choice for everybody would have meant a 100% yield figure for Harvard. It is not so.

[/quote]
</p>

<p>You deliberately misquoted me. I never said that Harvard had a 100% yield figure. I said that Harvard wins the majority of undergrad cross-admits with MIT, something that even MIT does not dispute. </p>

<p>
[quote]
No. Many people decide to do PhD because they want to do some research and add to the existing knowledge before moving on to money making business. The school brand name matters immensely for all PhDs. There are many physics PhDs (for example) who move into quantitative finance. Many PhDs start their own business in the silicon valley (esp in stanford). Many people go into consulting (which pay very large salaries to PhDs). That is why school name matters VERY much for these people. And these are the people who will prefer a Harvard PhD degree than a MIT degree as they are not going to be doing research for the rest of their lives. In fact it matters even more those whi decide to become professors. Saying that one who has decided to do a PhD knows exactly what he wants to do with his life is wrong.

[/quote]
</p>

<p>And the same is true EVEN MORE SO for undergrad. After all, as an undergrad, you REALLY don't know what you want to do with your whole life. But as a PhD student, you have a far far better idea. Hence, general brand name matters less. What matters much more to you is the specific strength of the specific department you will be in. </p>

<p>That's simply because at least you actually know what you want to study. As an undergrad, you can switch majors freely. Not so as a PhD student. You can't just enter a PhD program in physics, found out that you don't like and simply "decide" to switch over to a PhD program in English. </p>

<p>
[quote]
Maybe true if you are an american but not if you are an international student because the "Harvard" name helps more than MIT tag internationally. And it is a fact that in Harvard SEAS most students are of international origin. I personally know people in Harvard who turned down Stanford for Harvard just because the name is bigger internationally.

[/quote]
</p>

<p>Wait, I thought you said that the MIT brand name is just as strong as Harvard's. So if that's true, then why would you choose Harvard? </p>

<p>And again, it's all relative. Just like I agree that you have some grad students going to Harvard for engineering just because of the Harvard brand name, you have undergrads choosing Harvard just because of the Harvard brand name. The point is, it's far more prevalent to do so as an undergrad, when you often times don't even know what you want to major in or what you want to do with your life. </p>

<p>
[quote]
No. (If that were true why doesnt MIT suck out oxygen from mathematics or economics programs at harvard?)
The correct reason is Harvard never gave engineering much attention. Engineering was somehow never intellectually appealing to Harvard. But Harvard has realized its mistake and is concentrating big time on SEAS and I feel Harvard will be a top-notch engineering school within a decade

[/quote]
</p>

<p>Yeah, well, first off, how does that help you, as a prospective engineering student, now? You are going to get your engineering job prospects based on how good your school is NOW (or when you graduate in a few years), not how it might be in a decade. </p>

<p>Secondly, even in a decade, I highly doubt that even people at Harvard expect to be able to match MIT in engineering. That's a goal that will take many decades to attain, if it ever will be attained. </p>

<p>
[quote]
You don't have to be genius no doubt but you have to be extremely bright academically to get in. We never said Harvard grads are all geniuses. They are just more brilliant than the undergrads because the number of graduate seats are extremely limited and there is a huge competition for those spots on an international level. That makes the average quality of the grads much much higher than undergrads.

[/quote]
</p>

<p>Uh, no, again, not really. I don't buy the notion that the competition for these seats requires 'brilliance'. Again, what it requires is * research ability * which is not the same thing as brilliance. To give you an example, one of the most well-known ways to get into a Harvard PhD program is to first work for a Harvard professor as a research associate. What that means is that you graduate from undergrad, and get a job at Harvard doing research, and then get the prof to write a rec for you to let you in. I would say that of all the possible paths to getting into Harvard for your PhD, this is probably the most well-worn one. That path has little to do with your 'brilliance'. It has to do with your work ethic and your research ability, but not 'brilliance'. </p>

<p>You have to keep in mind the primary reason why people are admitted to Harvard PhD programs (or any PhD programs at any school). Let's be perfectly honest about what's really going on here. You're being admitted primarily to help the professors advance their own research agendas. That's the main reason. Getting into a PhD program is not a 'reward' for being brilliant. In fact, plenty of brilliant people don't get in if the faculty doesn't think that they can help any faculty advance their research (i.e. if the research interests don't align). Now, granted, if 2 people apply who both have strong research abilities that align well with the faculty, and there's only 1 spot available, they will take the more brilliant one. But the faculty will take the not-so-brilliant guy who is doing research that aligns well with the faculty over the brilliant guy who does not align with the faculty. </p>

<p>
[quote]
o, really why will someone make the assumption that anyone concentrating in physics, maths will always do well and not be in the bottom rungs. I dont understand.

[/quote]
</p>

<p>When did I make that assumption? I said that if you're one of the worst students, you're probably going to choose an easier subject. And that's not physics, math, or engineering. </p>

<p>
[quote]
I dont agree with your association of low IQ with the humanities and social sciences. There are many brilliant people who chose to study humanities just for the love of the subject (among other reasons) and in all probability the strongest of them are in Harvard grad programs.

[/quote]
</p>

<p>Oh really? Again, Harvard holds no monopoly when it comes to strength of humanities/social sciences PhD programs. Far from it, in fact. </p>

<p>
[quote]
As siserune said, "The 30 percent of internationals at GSAS come from a far stronger population than anything seen at undergraduate admissions" and that is very true.

[/quote]
</p>

<p>But again, that's irrelevant. Like I said, you're not being admitted to a PhD program for your 'brilliance'. You're being admitted to help the faculty advance its research. Nothing more, nothing less. I know some people who got admitted to every PhD program they applied to except one - Harvard. And that includes programs that are ranked far lower than Harvard. The reason is simple. Their research interests aligned very well with one particular prof at Harvard, but not with the profs at the other schools. </p>

<p>Furthermore, just being admitted is only the beginning of the game. You next have to choose to go. Plenty of people get admitted to Harvard phD programs, and then choose to go elsewhere. In fact, I know several particular departments that would consider themselves very lucky if they got a 50% yield. The Harvard physics and math department, for example, are almost certainly not going to hit 50% yield this year for their PhD admittees - heck, may not even hit 33% (for some reason, a lot of Harvard admittees are apparently choosing to go to MIT this year). Contrast that with undergrad, where about 80% of those who are admitted will choose to come. </p>

<p>But again, I would emphasize, we should not 'romanticize' what the PhD process is all about. They're not really there for their 'brilliance'. That's only a secondary consideration. The primary consideration is how helpful you will be as far as having the faculty advance their own research interests. . It's really a 'tit-for-tat' arrangement. You're cheap labor for the faculty, and in return you get a degree. So the question is, are they going to get that labor out of you? Plenty of brilliant people can't do research, or if they can, can't do the kind of research that any of the faculty is interested in.</p>

135

<p>
[quote]
Compared to undergraduate admission, Harvard PhD admission is almost entirely academically based.

[/quote]
</p>

<p>You're actually making the same mistake that microsoft is. See above. PhD admissions are NOT academically based. They are RESEARCH based. Just like all PhD admissions are. As I explained above, the primary criteria for being admitted to a PhD program is whether the faculty thinks you will be a valuable researcher to advance their own research agendas. Nothing more, nothing less. </p>

<p>Now, I agree that you are cutting out the athletes, cutting out the legacies, cutting out what you call the 'games' that go into undergraduate admissions. But you are replacing it with an EVEN BIGGER GAME - that being whether you are a good research fit for the department. A lot of 'brilliant' people can't do research. Or if they can, they aren't doing research that fits what any of the faculty want to do. </p>

<p>Secondly, Harvard has very heavy competition in every single PhD category. Harvard is the clear #1 in only a few categories. Hence, even if you can get into Harvard for your PhD, there is a strong chance that you will choose to go elsewhere (if you can get into that other place). Contrast that with undergrad, where 80% of the admittees will choose to come. I don't know of a single Harvard PhD program that can boast of an 80% yield rate, and certainly, Harvard can't boast of such a PhD yield rate across the board. Hence, even if Harvard really is admitting plenty of 'brilliant' people through an ungamed process (which is not true at all, but let's say it is), that doesn't mean that Harvard ends up with a brilliant PhD student population. </p>

<p>But the bottom line is this. Let's not romanticize what PhD admissions are all about, and what PhD programs are all about. You are cheap research labor for the faculty. Nothing more, nothing less. The faculty doesn't admit you because of your 'brilliance', they admit you for your ability to help them on their research.</p>

136

<p>According to Simonton (1988), the average IQ of the undergraduate (this includes state schools) was 115. I think that the survey was conducted years before 1988, so it's likely that the IQ has decreased since then (since students from more disadvantaged backgrounds are enrolling in university in larger numbers - though those at the top of the SES ladder are also enrolling in university in higher numbers, and there is a fairly strong correlation between socioeconomic status and IQ).</p>

<p>Bear in mind, of course, that only ~25% of Americans currently have a bachelor's degree.</p>

<p>==</p>

<p>Also, IQ tests were originally based on assumptions - assumptions that may be flawed. It's entirely possible that modern neuroscientific research can find more cognitive factors that have stronger correlations with whatever g is than current IQ tests. Of course, there's Howard Gardner's "theory of multiple intelligences," but even Herrnstein and Murray acknowledge that g only measures a specific portion of those multiple intelligences. Nonetheless, that g is vital for success in the information age. </p>

<p>The power of IQ tests lies in correlation.</p>

<p>==</p>

<p>Heh, it seems like the long thoughtful posts above are all made by adults (it depends on crystallized intelligence, which doesn't decline with age as fluid intelligence does - although I'm not that sure as to how accurate the distinction is). Experience matters for something.</p>

137

<p>At my high school, people with IQ's of 140+ are not exceedingly rare, and those with an IQ of 150+ come up every once in a while. 160+, I doubt there have been more than a handful in the school's history.</p>

<p>These "top" students often go to Harvard, MIT, Caltech, and such. It's definite that the students, at least, would qualify for Mensa, and likely by a substantial margin. A mean IQ of 130 wouldn't be too hard to believe.</p>

<p>I personally have an IQ that is measured (at age 8) to be above three standard deviations above the mean. This score, however, is merely average in a "highly gifted" magnet program such as North Hollywood HGM.</p>

<p>Furthermore, it's my opinion that if the SAT does not correlate very well with general intelligence, that the LSAT does so very strongly. It is pretty much like an IQ test minus the spatial portion.</p>

138

<p>One question to propose is => where do most people with such IQs go? They are more likely to go to top schools than average, but do most of them end up in state universities? I think that most people with such IQs go to college (as of 2000), but the number of people with such IQs is higher than the number of people who enroll in top universities.</p>

<p>==</p>

<p>Also, according to <a href="http://www.iqcomparisonsite.com/IQtable.aspx%5B/url%5D"&gt;http://www.iqcomparisonsite.com/IQtable.aspx&lt;/a>, 99.5th percentile corresponds to an IQ of approx 139.5.
As a rough comparison, the 99.5th percentile within a state's PSAT scores determines eligibility for National Merit Semifinalist status. We must also consider that the people who take the PSAT (how many high school students take it?) are more academically accomplished and college-focused than those who do not take the PSAT, and are likely to have higher IQs. Do most National Merit Semifinalists have such IQs? We don't know.</p>

<p>But clearly, PSAT scores are based on one setting. Many people have far higher SAT scores than their PSAT scores would otherwise indicate. Moreover, the PSAT math is sufficiently easy enough such that the highest scores differentiate those who have made stupid mistakes on a particular instance of the test, as opposed to those who don't (I'm referring to the 2004 PSAT, wherein a 76 means you missed 1 math question). The question is here - are tests like the AMC more reliable than the PSAT in the upper quartiles of student ability? Not necessarily, as a huge amount of variability comes in how many questions the student decides to guess on - many do not fully grasp the guessing penalty, and as a result, we see huge variability in AMC scores. AIME scores are even more variable - score drops are common, as well as score increases on the scale of 1-4, 1-5, 1-8, etc...</p>

<p>We also have to consider the population to which we're comparing the distribution to. Clearly, a person with an IQ of 130 in the United States would be at the 97.7th percentile, but he would have a much higher percentile in say, a random country in Africa (IQ differences between countries are well documented - the more important question is what does that mean?)</p>

139

<p>Also, does anyone know if medications like Adderall increase IQ scores in particular instances of tests? (for some students, it REALLY helps, for others, it doesn't, even though there is no difference in intelligence that it really confers) A number of students improve their SAT scores by hundreds of points with such medications (though I'm not quite sure how universal this trend is - clearly, Adderall helps some people far more than others - it may be related to the Yerkes-Dodson law, where those with low amounts of stimulation benefit A LOT, whereas those who already have adequate levels of it only become more anxious). If they do increase IQ scores, it only serves to indicate how IQ tests are indicative of a student's mood during the test.</p>

<p>A related question is - how related are IQ and SAT tests to measures of self-efficacy? If a person feels that he will do well in such tests, how much does that generally improve a student's score. Again, students with certain personalities are more likely to suffer from low self-efficacy than students with other personalities.</p>

<p>Another question to pose is - do they improve performance in timed tests more so than non-timed tests, and if so, then to what extent? And which types of students improve more with Adderall in timed tests, and do they improve more with Adderall in non-timed tests too? </p>

<p>For one thing, I hypothesize that Adderall doesn't help with knowledge-based non-IQ tests as much they help with non-knowledge-based IQ-related tests.</p>

<p>The core assumption to the reliability of SAT and IQ tests - is that they are relatively consistent for most students, provided that these students are performing to their maximum. Those who do not test at maximum ability may be that way because of (a) psychological factors particular to the instance, such as low attention span, (b) low self-efficacy, or (c) poor test-taking strategies. The pressing question is - how many students actually perform to their maximum? And even for those who are at maximum, how reliable are the IQ tests?</p>

<p>In that, tests like the AMC and AIME don't seem that reliable, given the huge amount of test score variability there is among certain students. But I think another question is - is variability higher when a substantial portion of the questions are difficult?</p>

140

<p>i just took an iq test (yes, im that bored) and my score is definitely around 30 points lower than when i took it last. has harvard made me stupider?</p>

<p>(dont pounce, im only kidding. but it is kinda sad. no more facebook!)</p>