The Unz Review • An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 TeasersiSteve Blog
Amy Harmon: Only 0.7% of Math Professors Are Black Due to Microaggressions! Reader: Coincidentally, Only 0.7% of Kids Who Score >=750 on Math SAT Are Black.
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeThanksLOLTroll
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

From iSteve commenter Gringo:

Amy Harmon at the NYT:

No one tallies the number of black mathematicians in those departments, but as best I can tell, there are 13. That comes to seven-tenths of 1 percent of the total -— perhaps as far as any job classification gets from accurately reflecting the share of black Americans in the general adult population, which stands at 13 percent.

It’s a fair bet that most math Ph.Ds. got 750 or above on the Math SAT. How do blacks do on the Math SAT? Of those who score 750 or above on the Math SAT, what proportion are black? How does this compare with Amy Harmon’s tally indicating that blacks comprise 0.7% of math professors ?Journal of Blacks in Higher Education: The Widening Racial Scoring Gap on the SAT College Admissions Test provides us with the answer.

In 2005, 153,132 African Americans took the SAT test. They made up 10.4 percent of all SAT test takers…

If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation’s most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

Blacks comprised 0.7% of those who scored 750 or above on the Math SAT, and also comprised 0.7% of Math professors. Looks to me as if there is no racial exclusion at all in doctoral level mathematics. On the contrary, Math SAT scores and blacks as math professors track very well.

This article was published in 2006, so the news has been out there for quite a while. It doesn’t say much for Amy Harmon’s information searching abilities that she is apparently unaware of it. One would think a journalist would be adept at information searching, but that doesn’t appear to be the case for Amy Harmon.

 
Hide 185 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
  1. Come on, Amy Harmon couldn’t do the math to figure that out EVEN IF she wanted to. To her they remain “hidden figures.”

    • Agree: AndrewR, Roderick Spode
    • Replies: @Mr. XYZ
    @Joe, Averaged

    How ironic, given the high average IQ of the ethnic group that she likely belongs to.

  2. What about African Math? We need to teach THAT!

    • Replies: @Anon7
    @Father O'Hara

    Exactly! What about the Lebombo bone, the world's oldest mathematical artifact? Found in a cave in Swaziland, it dates from 35,000 BC. It is a baboon fibula with exactly 29 marks cut uniformly into it. It is similar to calendar sticks in use by Bushmen today.

    The lunar month has 29.5 days; it could also be used to estimate menstrual cycle times, still an important topic in African math today.

    , @rufus
    @Father O'Hara

    Maff*

    , @Bruce County
    @Father O'Hara

    Yes I love that math... They use it mostly for getting the temperature just right in the cauldron for removing human flesh from bone. Later on it helped in the design and building of circular mud huts.

  3. Edray Goins, a professor of mathematics at Pomona College, gave up a professorship at Purdue after repeated racial slights.

    Purdue is MAGA Country! “Donald Trump won the election in Indiana with 56.9% of the vote. Hillary Clinton received 37.9% of the vote.” While Pomona is somewhere near the ocean in Hillary Country. Where there are still lots of shopping malls because East Asians and Subcons like them. Where you can see the preggo Chinese strolling around, here just to give birth, getting exercise.
    Did this oppressed mathematician land on his feet, or what?

    Slights means something above micro aggression….I am assuming.

  4. Anonymous [AKA "NoHomoInNoHo"] says:

    What a moron.

    What a persistent moron.

  5. • Replies: @eah
    @eah

    It’s a fair bet that most math Ph.Ds. got 750 or above on the Math SAT.

    You'll never convince people like Harmon that any gap -- the 750+ SAT gap, black math professor gap, whatever gap -- isn't due to 'institutional racism' etc -- the real problem is the dramatically low absolute level of achievement (ability) shown by a very large and growing fraction of the population (except it's basic arithmetic, not basic "math").

    So I've covered the big 3 metropolitan areas today: NYC, LA, Chicago/Illinois -- These places are pumping out people who can't do basic math. Think on the consequences.

  6. Her job is to help destroy our institutions of higher learning by replacing rational reasoning they always prized with a pseudo-religion that’s proven over the past century and a half an abject failure providing anyting it promises.

  7. @Father O'Hara
    What about African Math? We need to teach THAT!

    Replies: @Anon7, @rufus, @Bruce County

    Exactly! What about the Lebombo bone, the world’s oldest mathematical artifact? Found in a cave in Swaziland, it dates from 35,000 BC. It is a baboon fibula with exactly 29 marks cut uniformly into it. It is similar to calendar sticks in use by Bushmen today.

    The lunar month has 29.5 days; it could also be used to estimate menstrual cycle times, still an important topic in African math today.

  8. You forgot to adjust for the goodness factors:

    Blacks: 0.7% * goodness factor of 17 = 12%
    Whites: 60% * goodness factor of 0.5 = 30%
    Jews: 20% * goodness factor of 1.0 = 20%
    Asians: 20% * goodness factor of 1.5 = 30%
    Hispanics: 1.2% * goodness factor of 7 = 8%

    The figures on the right are what the percentages are supposed to be. That is why blacks and hispanics are underrepresented and the other groups are overrepresented.

  9. Amir Aczel found the earliest known representation of zero in a Cambodian jungle, and Terry Jones located the earliest known one of one in a temple in India, though he was locked out.

    FINDING ZERO
    A Mathematician’s Odyssey to Uncover the Origins of Numbers

    The Story of 1

    The earliest known something must be in Africa. How about the earliest known emigrant? “There goes the neighborhood! I’m crossing the isthmus, where the grass will be greener.”

    • Replies: @Logan
    @Reg Cæsar

    Not hardly. The Khmer zero is from AD 683. There is a Maya inscription using zero from 36 BC.

    , @El Dato
    @Reg Cæsar


    Amir Aczel found the earliest known representation of zero in a Cambodian jungle
     
    Indiana Jones and the Representation of Nothing!
    , @The Alarmist
    @Reg Cæsar


    "The earliest known something must be in Africa. How about the earliest known emigrant?"
     
    This is why paleontologists are brutally supressing any unearthed evidence of hominids outside of Africa that contradicts the Out of Africa theory.

    Replies: @El Dato

    , @Neil Templeton
    @Reg Cæsar

    Zero is the one great invention of the left, as it was developed to describe the intersection of leftist ideology with reality, i.e. the empty set. It also represents the lower limit of value to which the social system converges as those who believe in conserving value are liquidated.

  10. Well obviously the lack of black math PhD’s then must be the result of pre-collegiate microaggressions. Black students need to be kept away from white students and teachers in order to prevent these microaggressions. Or is it that black students need to be surrounded by white students in order to learn – white students, I guess, who must simply sit there in the classroom silently, like mannequins, in order to avoid microaggressions? So maybe we can just surround black students with white mannequins to help them learn better?

    Or maybe the microaggressions happen far earlier, like in the womb? Maybe black babies in the wombs of black bodies in Detroit are being microaggressed by white Trump voters in Idaho. If only we had a few more black math PhD’s to determine if that’s possible.

    • Replies: @Mr McKenna
    @Wilkey


    Black students need to be kept away from white students and teachers in order to prevent these microaggressions. Or is it that black students need to be surrounded by white students in order to learn – white students, I guess, who must simply sit there in the classroom silently, like mannequins, in order to avoid microaggressions? So maybe we can just surround black students with white mannequins to help them learn better?
     
    You're on the right track, believe it or not. What we need is for white people to act stupider, which is what the whole "step back" thing is about.

    Everywhere from the kindergarten class to the board meeting, when you, as a white male, have a god idea, the thing to do is to sit down, shut up, and act stupid. Everything else is aggression.

    Replies: @Mr McKenna

  11. Please stop bombarding us with hate facts. 12% of all Americans are black so until 12% of everything is black, the only explanation is racism.

    • Replies: @Barnard
    @Jack D

    They are willing to settle for 12%? My impression is there isn't a number that would satisfy them. Their percentage of the population has no influence on what they think they deserve.

    Replies: @ogunsiron

  12. Why would Amy Harmon use the share of Black Americans in the adult population of the USA as a point of comparison? Top university departments recruit internationally and will not reflect the demographics of the general American population. Isn’t this like complaining that the Yale faculty doesn’t reflect the demographics of New Haven, Connecticut?

  13. Ms. Harmon has a BA from Michigan in American Studies. Make of that what you will.

  14. @Wilkey
    Well obviously the lack of black math PhD's then must be the result of pre-collegiate microaggressions. Black students need to be kept away from white students and teachers in order to prevent these microaggressions. Or is it that black students need to be surrounded by white students in order to learn - white students, I guess, who must simply sit there in the classroom silently, like mannequins, in order to avoid microaggressions? So maybe we can just surround black students with white mannequins to help them learn better?

    Or maybe the microaggressions happen far earlier, like in the womb? Maybe black babies in the wombs of black bodies in Detroit are being microaggressed by white Trump voters in Idaho. If only we had a few more black math PhD's to determine if that's possible.

    Replies: @Mr McKenna

    Black students need to be kept away from white students and teachers in order to prevent these microaggressions. Or is it that black students need to be surrounded by white students in order to learn – white students, I guess, who must simply sit there in the classroom silently, like mannequins, in order to avoid microaggressions? So maybe we can just surround black students with white mannequins to help them learn better?

    You’re on the right track, believe it or not. What we need is for white people to act stupider, which is what the whole “step back” thing is about.

    Everywhere from the kindergarten class to the board meeting, when you, as a white male, have a god idea, the thing to do is to sit down, shut up, and act stupid. Everything else is aggression.

    • Replies: @Mr McKenna
    @Mr McKenna


    have a god idea
     
    See what I mean? Even my typos are megalomaniac.
    I'll step back now and let some WOKE WoCs have the floor.
  15. @Mr McKenna
    @Wilkey


    Black students need to be kept away from white students and teachers in order to prevent these microaggressions. Or is it that black students need to be surrounded by white students in order to learn – white students, I guess, who must simply sit there in the classroom silently, like mannequins, in order to avoid microaggressions? So maybe we can just surround black students with white mannequins to help them learn better?
     
    You're on the right track, believe it or not. What we need is for white people to act stupider, which is what the whole "step back" thing is about.

    Everywhere from the kindergarten class to the board meeting, when you, as a white male, have a god idea, the thing to do is to sit down, shut up, and act stupid. Everything else is aggression.

    Replies: @Mr McKenna

    have a god idea

    See what I mean? Even my typos are megalomaniac.
    I’ll step back now and let some WOKE WoCs have the floor.

  16. What percentile ranking would 750 be? I would expect virtually all math professors to score in the 99th percentile on the Math portion of the SAT.

  17. @Reg Cæsar
    Amir Aczel found the earliest known representation of zero in a Cambodian jungle, and Terry Jones located the earliest known one of one in a temple in India, though he was locked out.


    FINDING ZERO
    A Mathematician's Odyssey to Uncover the Origins of Numbers


    The Story of 1


    The earliest known something must be in Africa. How about the earliest known emigrant? "There goes the neighborhood! I'm crossing the isthmus, where the grass will be greener."

    Replies: @Logan, @El Dato, @The Alarmist, @Neil Templeton

    Not hardly. The Khmer zero is from AD 683. There is a Maya inscription using zero from 36 BC.

  18. What percent are Asian?

  19. @Jack D
    Please stop bombarding us with hate facts. 12% of all Americans are black so until 12% of everything is black, the only explanation is racism.

    Replies: @Barnard

    They are willing to settle for 12%? My impression is there isn’t a number that would satisfy them. Their percentage of the population has no influence on what they think they deserve.

    • Agree: bomag
    • Replies: @ogunsiron
    @Barnard

    The *minimum* acceptable percentage of blacks as members of a desirable group is more or less max( % black in neighborhood, % black in city, % black in state, % black in country, % black in the world). The min starts there.

    There is of course no max acceptable % of blacks in desirable categories.

    For undesirable categories there's a formula for the max acceptable percentage. Not too difficult to figure out.

  20. 750 on math is (as of 2017) good for 98th percentile of all students (estimated), 96th percentile of SAT takers (source: College Board). Sorry, but not 1 in 25 SAT takers can go on to get a math phd. That those two figures match demonstrates the pro-black bias.

    • Replies: @Thulean Friend
    @Alan Mercer

    Remember that not everyone is taking their PhD at an Ivy League school. UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD. You add the affirmative action factor and you're more than set. Plus far from every person with a 750 math score or above wants to be in academia.

    If you score that high as a black person, your options are virutally unlimited in the corporate sector. Lots of affirmative action jobs available for pretty smart blacks since there are so few to begin with. That scarcity must be affecting the PhD job market, too, for universities who want to diversify.

    Replies: @PhysicistDave

    , @Spangel
    @Alan Mercer

    Agreed. The math sat is graded on a steep curve since so many Asians don’t miss a single question. Missing even 1 question can take 40 points off or at least it could when I took it. Here and there you might see math PhD material getting less than an 800 because of a silly mistake like marking d when they meant to mark c, but I think they’d have to possess enough understanding to get every question right within the allotted time.

    The fact that .7% blacks have an sat 750 and above really just speaks to how muc affirmative action there must be in university math departments.

    , @International Jew
    @Alan Mercer


    Missing even 1 question can take 40 points off
     
    Mm, doubtful. Says here
    https://blog.prepscholar.com/how-is-the-sat-scored-scoring-charts
    you get 750 if you miss about five (out of the usual 58) math questions.

    Never mind becoming a math prof or even completing a PhD in math. For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT...) the math SAT is ridiculously easy. All of it. There's never a single question that's hard in any way. If you didn't get 800, you got 790 for some bit of carelessness. And I'm describing the state of affairs when I took it, which was before 1995, when it was made easier. You didn't even brag about your SAT (not that we were a modest bunch), you bragged about your exploits in the high school math league, if not in international competitions (where the easiest question is harder than the hardest question that's ever been seen on the SAT).

    To make a long story short, if only 0.7% of black kids can get 750 on the SAT (today's SAT — pre-1995 that was like a 700) then the fraction that has what it takes to become a math prof at a top-50 school is way lower.

    Replies: @Pericles, @Jack D, @res

    , @JimS
    @Alan Mercer

    Agreed. The test of reference is unequivocally the Putnam Competition, generally participated in by college seniors who are serious about mathematics. By my imperfect recollection, that test is comprised of a total of 12 questions, scored from 0-120, with roughly half of all takers scoring a 0.

    I highly doubt Putnam would provide a racial breakdown of performance. If it did, the obvious questions would be, what is the fraction of test takers that were black? In other words, how many are even of a level of mathematics such that they would even attempt the test? And the second question, how many blacks actually register a nonzero score?

    (Snarky question, to be disregarded in all serious discussion: Is the difference in numbers between those two answers exactly 13? For the record, I would seriously believe it to be much higher.)

  21. @Father O'Hara
    What about African Math? We need to teach THAT!

    Replies: @Anon7, @rufus, @Bruce County

    Maff*

  22. Since the renorming of the SAT, 750 corresponds to 99%tile. It’s an understatement to say math PhDs are in the top 1% (2.5 sigma). I’m guessing from personal experience that productive math grads are more like NFL players, from the 5 sigma level of talent. At that level, you are selecting from tiny sub populations of Europeans, East Asians, and Indians, only.

    • Replies: @Edward
    @JimB


    Since the renorming of the SAT, 750 corresponds to 99%tile. It’s an understatement to say math PhDs are in the top 1% (2.5 sigma). I’m guessing from personal experience that productive math grads are more like NFL players, from the 5 sigma level of talent. At that level, you are selecting from tiny sub populations of Europeans, East Asians, and Indians, only.
     
    5-sigma is too high. The Fields Medalist Richard Borcherds has a performance IQ (a measure of nonverbal ability) of 147 (top 0.09%) and a full-scale IQ of 137 (top 0.7%).

    https://www.tandfonline.com/doi/abs/10.1080/13554799908402743

    Performance IQ is only a proxy for mathematical ability, but if his nonverbal IQ is a bit over 3 SD above the mean, his mathematical ability is probably only going to be ~4SD above the mean, at most.

    Borcherds is probably on the low end of the ability distribution of Fields Medalists (Terence Tao reached the ceiling on the Stanford-Binet, with an IQ of 175). Even so, you're still right: it shouldn't be surprising that the only minority groups we see with good representation in math departments are Jewish-Americans (avg IQ 110-115, with their mathematical abilities probably even higher than this), Indian-Americans (avg IQ ~110) and East Asian-Americans (avg IQ ~106, with mathematical abilities probably higher than this).

    Replies: @Peter Johnson

  23. Ok, add to my list of reparations all blacks will now start with 750 on the SAT math test, however the perfect SAT math score will now be 1600. And unlimited #2 pencils.

    • Replies: @Mr. Anon
    @Buffalo Joe


    And unlimited #2 pencils.
     
    The very act of having to blacken in the ovals when answering the questions is tantamount to being exposed to black-face, and gives rise to stereotype-threat.

    Replies: @Buffalo Joe

  24. One would think a journalist would be adept at information searching, but that doesn’t appear to be the case for Amy Harmon.

    I suspect Gringo knows the answer to this conundrum. Harmon, like pretty much all NYT “journalists,” isn’t searching for any information that would contradict The Narrative. And she wouldn’t publish any counter-narrative facts that she found anyway.

    That’s why her articles are so embarassingly lame. There is no objective evidence supporting her claims. So she just makes false assertions and then collects quotes from people about their feelz on the subject.

    • Agree: Redneck farmer
    • Replies: @Joe, Averaged
    @Hypnotoad666

    Yes that’s what you get from American Studies majors. Didn’t journalism used to be a real profession?

    Replies: @Brutusale, @Gringo, @MBlanc46

  25. @Hypnotoad666

    One would think a journalist would be adept at information searching, but that doesn’t appear to be the case for Amy Harmon.
     
    I suspect Gringo knows the answer to this conundrum. Harmon, like pretty much all NYT "journalists," isn't searching for any information that would contradict The Narrative. And she wouldn't publish any counter-narrative facts that she found anyway.

    That's why her articles are so embarassingly lame. There is no objective evidence supporting her claims. So she just makes false assertions and then collects quotes from people about their feelz on the subject.

    Replies: @Joe, Averaged

    Yes that’s what you get from American Studies majors. Didn’t journalism used to be a real profession?

    • Replies: @Brutusale
    @Joe, Averaged

    Yes, but they were called reporters then. Or ink-stained wretches, whichever they preferred.

    , @Gringo
    @Joe, Averaged

    Yes that’s what you get from American Studies majors.
    I wonder if she took any math courses in college. I doubt it, as she talks social science lingo in discussing math, and very little about process of doing or understanding math.


    Didn’t journalism used to be a real profession?
    Four hometown peers went into journalism,perhaps inspired by Watergate- though two were on the high school paper years before Watergate. One was in my class; two were in my brother's class. They were very bright. Three were Merit Finalists. The other was bright enough to write a number of books and win some awards. Two have worked for big-name papers.

    Politics of these four are wide-ranging, but not as lefty as one would have predicted from our upbringing. One is far lefty. One is libertarian. One is rather apolitical- doesn't write on politics- and has gotten into trouble for some politically incorrect comments. The other, while liberal, is more of a 1960s liberal than a 2018 liberal.

    Amy Harmon is not on their level.

    , @MBlanc46
    @Joe, Averaged

    No. Journalists used to be high school grads who could write a bit and were very good at extracting information from people.

  26. Actually experienced nausea thinking about this one.

    1) A Field’s Medal or a 750 SAT score is not a reward that can be given to erase misperceptions about slavery, inferiority or microaggressions. They are recognition of and/or accomplishments resulting from talent and hard work.
    2) Any discussion of math disaparities needs to start with arithmetic disparities. I’ve had the good fortune of trying to teach whites and blacks eighth grade math after they have been thoroughly ruined by black seventh grade math teachers. The most basic problem that occurs frequently is that black teachers do not teach understanding. They only teach a procedure to get the ‘right answer’ which results in Cargo Cult math.
    3) Any discussion of higher math disparities needs to start with algebra and the abstractions needed to grasp and perform. The inability to deal with simple abstractions prevents any real progress amongst black math students.

    What’s scary about all this is the same deficiencies appear in social studies. The ability to make decisions regarding rights and responsibilities of blacks, whites, Asians, Mexicans etc. demands the ability to understand basic facts, to comprehend reality rather than to parrot what black leaders spout out, and to reason abstractly about these facts and the principles of civilization.

    OTOH, I see the manufactured crisis in Venezuela as admission from the PTB that all is not well in the Republic, meaning that at some time not too distant, insufficient food, oil etc will make it imperative for adults to reason effectively, to forge new social groups capable of self preservation, and to chart a path to a sustainable future. Constant pandering to lowered black expectations regarding capability, and phony awards and other meaningless symbols preclude blacks as a group from making this journey.

    Let us all mourn for them and then forget them.

    • Agree: ThreeCranes
    • Replies: @Buffalo Joe
    @James Speaks

    James, a bigger problem for blacks in America isn't abstraction, it is delusion. Nice comment.

    , @El Dato
    @James Speaks


    They only teach a procedure to get the ‘right answer’ which results in Cargo Cult math.
     
    That's astonishingly hard to break.

    I was educated according to a rather horribad "Bourbaki" approach (starting with set theory probably because Russell & Whithead decided to build math on this) and it was cargo cult all the way till I was 14 and really decided to get into it with 8+h/week of high school math. In an ideal world it would have been more because progress was slow.
  27. @Buffalo Joe
    Ok, add to my list of reparations all blacks will now start with 750 on the SAT math test, however the perfect SAT math score will now be 1600. And unlimited #2 pencils.

    Replies: @Mr. Anon

    And unlimited #2 pencils.

    The very act of having to blacken in the ovals when answering the questions is tantamount to being exposed to black-face, and gives rise to stereotype-threat.

    • Replies: @Buffalo Joe
    @Mr. Anon

    Mr. Anon, Ok, they can mark all the ovals with a white pencil.

  28. Black man in Texas ties up two white females and sets them on fire. CNN is uninterested.

    One of these “Beckys” should have called the police.

    The level of anti-white hostility from our institutions and MSM is staggering.

    https://www.dailymail.co.uk/news/article-6736455/Shocking-moment-man-douses-store-clerk-customer-lighter-fluid-sets-one-fire.html

  29. Anon[387] • Disclaimer says:

    Amy Harmon
    @amy_harmon

    I spoke to white and Asian math faculty for this story too. They say they DO care about diversity, the problem is the pipeline. And there is, it is true, a huge problem with the pipeline…. [I’ll try to finish this thread later…]

    1:12 PM – 23 Feb 2019

    This just ends up with the pressure being put on the undergraduate and graduate schools, and even high schools, to water down math programs so that mediocre black students are not weeded out. The sooner they are weeded out, the less student loan money they will be in debt for, and the more likely it is that they will find something that they are good at.

    Maya Trotz

    Replying to @amy_harmon

    Show the numbers. The pipeline is bigger at the undergraduate and graduate level, so that is a lame excuse. They are in industry and elsewhere, find them and hire them.

    She’s saying the the pipeline narrows as courses become more difficult and advanced … how lame, what an excuse! Why would that happen. Actually, the white and Asian pipeline narrows as the level goes up also. There is pipeline lameness across the board!

    An old article from USNWR:

    Experts: ‘Weed Out’ Classes Are Killing STEM Achievement
    Classes designed to make students fail are making students switch majors at alarming rates, experts say.

    https://www.usnews.com/news/blogs/stem-education/2012/04/19/experts-weed-out-classes-are-killing-stem-achievement

    NOT ENOUGH AMERICAN students are showing interest in studying for degrees in science, technology, engineering and math, but what experts are more shocked by is the fact that colleges are throwing out the students who are interested.

    Nearly half of all students who begin studying for a STEM degree switch majors, according to several studies. “Weed-out” classes, curve grading and a lack of faculty involvement are to blame, experts said at a Bayer Corporation forum on STEM in higher education in Washington Wednesday.

    “We need to wash out the ‘weed-them-out orientation’ in the classroom,” says Mary Fox, co-director at the Center for Study of Women, Science and Technology at Georgia Tech. “That is not a hospitable climate for students, we have to teach students to move along rather than have them sink or swim.”

    Many veteran STEM professors believe science should be hard, and the course work isn’t something every student can do. For them, difficult freshman-year classes separate the cream of the crop.

    But in a country where more scientists are desperately needed, that culture needs to change, says Freeman Hrabowski, president of the University of Maryland Baltimore County. Hrabowski was named as one of TIME Magazine’s 100 most influential people earlier this week for the University’s success in graduating minority students in STEM.

    University of Maryland Baltimore County? Give me a break! Ain’t no talented STEM students coming out of Baltimore County ever. Be serious!

    Weeding Out Students In Stem

    http://gmufourthestate.com/2018/11/12/weeding-out-students-in-stem/

    About half of undergraduate STEM degree candidates leave their field before completing a bachelor’s degree. One of the reasons may be “weed-out” classes, which are typically large entry-level courses designed to be difficult so that students can decide whether they wish to continue in their field.

    “I think the big scare for STEM students is math,” said Guillen-Piazza. “[For example], calculus weeds people out. Many of my peers have complained about general and organic chemistry, and have switched to pursuing a liberal arts degree.”

    Most difficult STEM classes reside in freshmen, or entry level classes. Most people believe they design entry level classes to be difficult in order to weed out students who do not try enough, or are not passionate enough to go into the STEM field. A course notorious for failing students at Mason is Chemistry 211.

    “We do not design classes to be difficult or to weed out students,” said Dr. Rosenberg. “The College of Science is dedicated to the success of our students. Because these are intrinsically difficult classes, that is the nature of the material, not something we have created.”

  30. @Alan Mercer
    750 on math is (as of 2017) good for 98th percentile of all students (estimated), 96th percentile of SAT takers (source: College Board). Sorry, but not 1 in 25 SAT takers can go on to get a math phd. That those two figures match demonstrates the pro-black bias.

    Replies: @Thulean Friend, @Spangel, @International Jew, @JimS

    Remember that not everyone is taking their PhD at an Ivy League school. UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD. You add the affirmative action factor and you’re more than set. Plus far from every person with a 750 math score or above wants to be in academia.

    If you score that high as a black person, your options are virutally unlimited in the corporate sector. Lots of affirmative action jobs available for pretty smart blacks since there are so few to begin with. That scarcity must be affecting the PhD job market, too, for universities who want to diversify.

    • Replies: @PhysicistDave
    @Thulean Friend

    Thulean Friend wrote:


    UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD.
     
    No, I don't think so. As others have said, for those of us who are reasonably good at math, an 800 on the old SAT Math, before they made the scoring easier, was not very difficult.

    When I was at Caltech back in the '70s, we sort of looked down our noses at the students who were somehow admitted with less than an 800 on math: we figured they might be okay in biology.

    Yes, Caltech is harder than anywhere else except MIT, but I am talking undergrads, most of whom were not math majors. I would bet that a Ph.D. in math, even at UMass Amherst, requires being as good at math as the average BS chemical engineer from Caltech.

    Replies: @Thulean Friend

  31. @Joe, Averaged
    @Hypnotoad666

    Yes that’s what you get from American Studies majors. Didn’t journalism used to be a real profession?

    Replies: @Brutusale, @Gringo, @MBlanc46

    Yes, but they were called reporters then. Or ink-stained wretches, whichever they preferred.

  32. @Alan Mercer
    750 on math is (as of 2017) good for 98th percentile of all students (estimated), 96th percentile of SAT takers (source: College Board). Sorry, but not 1 in 25 SAT takers can go on to get a math phd. That those two figures match demonstrates the pro-black bias.

    Replies: @Thulean Friend, @Spangel, @International Jew, @JimS

    Agreed. The math sat is graded on a steep curve since so many Asians don’t miss a single question. Missing even 1 question can take 40 points off or at least it could when I took it. Here and there you might see math PhD material getting less than an 800 because of a silly mistake like marking d when they meant to mark c, but I think they’d have to possess enough understanding to get every question right within the allotted time.

    The fact that .7% blacks have an sat 750 and above really just speaks to how muc affirmative action there must be in university math departments.

  33. In 2015, around 25% fewer black kids took the test, but around 4x the 2005 number got above a 750. Could be the tyranny of tiny numbers, but I also thought black kids took the SAT at the same or higher rate than their genpop representation.

    • Replies: @Spangel
    @The Practical Conservative

    Is it just black Americans? Could be that many more promising Africans were invited to take the test. I once worked with a black African who was recruited by Harvard 2 years before high school graduation and put in a training program to prepare her to apply to Harvard.

    , @res
    @The Practical Conservative

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns "Unauthorized." It's available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    https://i0.wp.com/www.brookings.edu/wp-content/uploads/2017/01/ccf_20170201_reeves_2.png

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html


    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation's most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

     

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation's 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.
     
    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    Replies: @Gringo, @res, @Mr. XYZ, @res

  34. anon[214] • Disclaimer says:

    Amy Harmon and the Crazy Left are not interested in a more reasonable analysis of the .07% figure. We are dealing with a religious cult that worships black bodies and female penises. For these people reasoning about numbers is a devilish trick of the evil white patriarchy.
    Look at the history of the world’s great religions. There is no claim too far-fetched to be believed. Millions will believe it for thousands of years.
    The worship of black bodies and female penises is of course insane, but that has no bearing on whether it will have staying power as a religion.
    Whether this cult has staying power is unfortunately something we are going to find out.

  35. @The Practical Conservative
    In 2015, around 25% fewer black kids took the test, but around 4x the 2005 number got above a 750. Could be the tyranny of tiny numbers, but I also thought black kids took the SAT at the same or higher rate than their genpop representation.

    Replies: @Spangel, @res

    Is it just black Americans? Could be that many more promising Africans were invited to take the test. I once worked with a black African who was recruited by Harvard 2 years before high school graduation and put in a training program to prepare her to apply to Harvard.

  36. @Joe, Averaged
    @Hypnotoad666

    Yes that’s what you get from American Studies majors. Didn’t journalism used to be a real profession?

    Replies: @Brutusale, @Gringo, @MBlanc46

    Yes that’s what you get from American Studies majors.
    I wonder if she took any math courses in college. I doubt it, as she talks social science lingo in discussing math, and very little about process of doing or understanding math.

    Didn’t journalism used to be a real profession?
    Four hometown peers went into journalism,perhaps inspired by Watergate- though two were on the high school paper years before Watergate. One was in my class; two were in my brother’s class. They were very bright. Three were Merit Finalists. The other was bright enough to write a number of books and win some awards. Two have worked for big-name papers.

    Politics of these four are wide-ranging, but not as lefty as one would have predicted from our upbringing. One is far lefty. One is libertarian. One is rather apolitical- doesn’t write on politics- and has gotten into trouble for some politically incorrect comments. The other, while liberal, is more of a 1960s liberal than a 2018 liberal.

    Amy Harmon is not on their level.

  37. Anon[387] • Disclaimer says:

    A sociologist chimes in:

    https://familyinequality.wordpress.com/2019/02/20/white-children-are-2-7-times-more-likely-than-black-children-to-live-with-a-parent-who-has-a-phd/

    This guy was responsible for a statistic that Harmon used in her followup piece.

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD

    For a reflection Amy Harmon was working on, a followup to her article on the experience of Black mathematicians in American academia, I took a shot at the question: How many children have parents with PhDs?

    The result was the highlighted passage (17 words and a link!) in her piece:

    [all the racial biases that contribute to Black underrepresentation include] the well-documented racial disparities in public-school resources, the selection of students for gifted programs — and the fact that having a parent with a Ph.D. is helpful to getting one in math, while black children are less than half as likely as white children to live with such a parent.

    To get there: I used data from the U.S. Census Bureau via IPUMS.org: The 1990 5% Public Use Microdata Sample (decennial census); and the 2000, 2010, and 2017 American Community Surveys.

    I coded race/ethnicity into four mutually-exclusive categories: Single-race White, Black, and Asian/Pacific Islander (API); and Hispanic (including those of any race). I dropped from the analysis non-Hispanic children with multiple races reported, and American Indian / Alaska Natives (for whom about 0.5 percent lived with a PhD parent in 2017).

    IPUMS made a tool that attaches values of parents’ variables to children with whom they share a household. I used that to calculate the highest level of education of each child’s coresident parents. In the Census data, children may have up to two parents present (which may be of the same sex in 2010 and 2017). Children living with no parent in the household were not included.

    This let me calculate the percentage of children living (at the moment of the survey) with one or more parents who had a PhD. For each of the four groups the percentage of children living with a parent who has a PhD roughly doubled between 1990 and 2017. API children had the highest chance of living with a PhD parent, reaching 6.8 percent in 2017. The percentages for the other groups were: Whites, 2.7 percent; Blacks, 1.0 percent; and Hispanics, 0.7 percent:

    All that work resulted in a blog post in which searches for IQ, intelligence, cognitive, confound* come up with no results, except in a quote from Harmon:

    Some misguided people claim that there are not many black research mathematicians because African-Americans are not as intelligent as other races. These people, whom I have reported on for other stories in recent months, almost invariably use mathematical accomplishment as their yardstick for intelligence. They note that no individuals of African descent have won the Fields Medal, math’s equivalent of the Nobel Prize. They lack any genetic evidence to explain the gap in average I.Q. scores between white and black Americans that they cite as the basis of their belief, or reason to think that a genetic trait would be impervious to social or educational intervention, or that high I.Q. is key to math ability, which Timothy Gowers, a 1998 Fields medalist, has attributed largely to “the capacity to become obsessed with a math problem.”

    But I have been reporting on these topics for several years, and I am acutely aware that math prowess factors heavily into the popular conception of intelligence. There’s a vicious cycle at work: The lack of African-American representation in math can end up feeding pernicious biases, which in turn add to the many obstacles mathematically talented minorities face. Which was one more reason it seemed especially important to hold up to the light all the racial biases that contribute to that underrepresentation.

    If there is an IQ gap between blacks and whites, whatever the cause, that is sufficient to explain the lack of black math talent, is it not? If the gap is 100 percent environmental, it nevertheless explains the lack of black math talent and the lack of black math research academics.

    If the gap is environmental, say lead in water or fetal alcohol syndrome, then the environmental cause permanently affected the brain. Is she hinting that there is some environmental cause that would instantly liberate the brains of adult blacks if found, say Beyonce music overdosing? Even it that were true, you’d have to identify the cause and apply the fix before black math whizes could perform their whizziness. In the meantime they would be treading water in their jobs. Is she suggesting that blacks be hired as math professors in advance to have them in place when the real-soon-now discovery of the environmental cause of their dullness is uncovered?

    • Replies: @Buffalo Joe
    @Anon

    Anon, two thoughts, a lot of blacks have Doctorates in Education, does that count as a PhD.? And two, people doing the hiring really don't care if some one's parent or parents have PhDs, they are hiring the applicant.

    Replies: @Arclight

    , @Gringo
    @Anon

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD.

    My immediate reaction: so what? Not everyone wants to jump through the hoops to get a doctorate. And yes, jumping through hoops is what it involves. I didn't want to. I notice that Amy Harmon didn't want to.

    I would tend to discount doctorates in the field of Education. Because I taught for several years, I looked at the doctoral dissertations of some colleagues in Education- some who earned them after I left teaching.

    The doctoral dissertations I have read in Education were not about research that could be replicated, but about unique stories- call them narratives, if you will. I'm sure that there are some genuine research-quality dissertations done in Education, but those I have looked at were just stories.

    From yet another paper in the Journal of Blacks in Higher Education, we find out that in 2004, 41.3% of doctorates that blacks earned came in the field of Education. Whites earned 19.1% of their doctorates in Education.
    If you take out Education doctorates from the equation, you can say that "White children are ~3.5 times more likely than Black children to live with a parent who has a real Ph.D.- a doctorate not in the field of Education."

    .

    http://www.jbhe.com/news_views/50_black_doctoraldegrees.html

    Replies: @Anon

    , @Reg Cæsar
    @Anon


    White children are 2.7-times more likely
     
    As, not more. G's Louise!

    There's a logical difference.
  38. Anon[315] • Disclaimer says: • Website

    This has nothing to do with math, blacks, or anything. It’s all about her own schoolmarmish sense of virtue. “Look, I care about the right things for the right audience.” She is the product of conditioning by PC. (Or she knows the truth but is cynically telling lies to spread the blame around. After all, elite media and academia are totally controlled by Liberals. Even science departments too, though a bit less than social sciences. So, how come there aren’t enough blacks? By blaming all of society, those at the top can lessen the blame for themselves and just say WE ALL MUST TRY HARDER. Now, we know the truth. A lot of blacks can’t make the grade, and given black culture, even smart blacks are not all that interested in geek stuff. Why not just go into business and be a favored token pushed in front of the line?) It’s her stunt as Jane Goodall of social sciences.

  39. @James Speaks
    Actually experienced nausea thinking about this one.

    1) A Field's Medal or a 750 SAT score is not a reward that can be given to erase misperceptions about slavery, inferiority or microaggressions. They are recognition of and/or accomplishments resulting from talent and hard work.
    2) Any discussion of math disaparities needs to start with arithmetic disparities. I've had the good fortune of trying to teach whites and blacks eighth grade math after they have been thoroughly ruined by black seventh grade math teachers. The most basic problem that occurs frequently is that black teachers do not teach understanding. They only teach a procedure to get the 'right answer' which results in Cargo Cult math.
    3) Any discussion of higher math disparities needs to start with algebra and the abstractions needed to grasp and perform. The inability to deal with simple abstractions prevents any real progress amongst black math students.

    What's scary about all this is the same deficiencies appear in social studies. The ability to make decisions regarding rights and responsibilities of blacks, whites, Asians, Mexicans etc. demands the ability to understand basic facts, to comprehend reality rather than to parrot what black leaders spout out, and to reason abstractly about these facts and the principles of civilization.

    OTOH, I see the manufactured crisis in Venezuela as admission from the PTB that all is not well in the Republic, meaning that at some time not too distant, insufficient food, oil etc will make it imperative for adults to reason effectively, to forge new social groups capable of self preservation, and to chart a path to a sustainable future. Constant pandering to lowered black expectations regarding capability, and phony awards and other meaningless symbols preclude blacks as a group from making this journey.

    Let us all mourn for them and then forget them.

    Replies: @Buffalo Joe, @El Dato

    James, a bigger problem for blacks in America isn’t abstraction, it is delusion. Nice comment.

  40. @Mr. Anon
    @Buffalo Joe


    And unlimited #2 pencils.
     
    The very act of having to blacken in the ovals when answering the questions is tantamount to being exposed to black-face, and gives rise to stereotype-threat.

    Replies: @Buffalo Joe

    Mr. Anon, Ok, they can mark all the ovals with a white pencil.

  41. anon[207] • Disclaimer says: • Website

    This is my first post to Unz Review (although iSteve was kind enough once before to post something I’d written elsewhere online under my pseudonym A New Radical Centrism).

    Unfortunately, I come bearing some unpleasant personal news, and I’m afraid it involves AMY HARMON (I prefer to refer to her in ALL CAPS).

    After responding about an hour ago to a tweet by “genetics & social inequality” superstar Dr. Paige Harden, in which I breathlessly extolled the reportorial virtues, impartiality, and integrity of ace science journalist AMY HARMON, I was unceremoniously blocked by Dr. Harden.

    I won’t unnecessarily go into the details here (you can view them yourself at https://twitter.com/kph3k), but Dr. Harden had earlier tweeted the following:

    ‘Really want a science journalist to write a piece on how open science practices (preprints, open data) – while clearly a net plus for replicability and speed of scientific communication — have made it easier for eugenicist pseudoscience to thrive.”

    I responded to the tweet by enthusiastically recommending AMY HARMON for the task and then -– BLAM! — before you knew it, I was blocked!

    There’s no accounting for this kind of behavior.

    • Replies: @Anon
    @anon

    Give Paige Harden a break. I think she knows the score, and just wants to keep her job. She has a book coming out called The Genetic Lottery.

    https://www.unz.com/isteve/paige-hardens-upcoming-book-the-genetic-lottery/

    As far as I can tell, the premise is that our IQs are unearned, gained by random mitosis of our germ cells and random developmental processes in the womb.

    So high IQ people need to check their privilege and cooperate in the redistribution of the excess unearned income that they receive from their random luck.

    In other words, Dr. Harden's book will be saying that, okay, maybe blacks have low IQ, but so what, and we should give them free money in some sort of universal income scheme. She will not be literally saying that blacks have low IQs. More something along the lines of some people have IQs below 85 or 100 or whatever, and they should be supported. But really, if it isn't about blacks, would there really even be a book? Would anyone care about low IQ whites?

    , @Anon
    @anon


    ‘Really want a science journalist to write a piece on how open science practices (preprints, open data) – while clearly a net plus for replicability and speed of scientific communication — have made it easier for eugenicist pseudoscience to thrive.”

    I responded to the tweet by enthusiastically recommending AMY HARMON for the task and then -– BLAM! — before you knew it, I was blocked!
     
    I think that there is a "rest of the story" here, perhaps contained in the word "enthusiastically." Was your recommendation along the lines of, "Sounds like a job for that IDIOT New York Times so-called writer Amy Harmon, who should learn to code!!!!!"?
  42. @Anon
    A sociologist chimes in:

    https://familyinequality.wordpress.com/2019/02/20/white-children-are-2-7-times-more-likely-than-black-children-to-live-with-a-parent-who-has-a-phd/

    This guy was responsible for a statistic that Harmon used in her followup piece.

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD

    For a reflection Amy Harmon was working on, a followup to her article on the experience of Black mathematicians in American academia, I took a shot at the question: How many children have parents with PhDs?

    The result was the highlighted passage (17 words and a link!) in her piece:

    [all the racial biases that contribute to Black underrepresentation include] the well-documented racial disparities in public-school resources, the selection of students for gifted programs — and the fact that having a parent with a Ph.D. is helpful to getting one in math, while black children are less than half as likely as white children to live with such a parent.

    To get there: I used data from the U.S. Census Bureau via IPUMS.org: The 1990 5% Public Use Microdata Sample (decennial census); and the 2000, 2010, and 2017 American Community Surveys.

    I coded race/ethnicity into four mutually-exclusive categories: Single-race White, Black, and Asian/Pacific Islander (API); and Hispanic (including those of any race). I dropped from the analysis non-Hispanic children with multiple races reported, and American Indian / Alaska Natives (for whom about 0.5 percent lived with a PhD parent in 2017).

    IPUMS made a tool that attaches values of parents’ variables to children with whom they share a household. I used that to calculate the highest level of education of each child’s coresident parents. In the Census data, children may have up to two parents present (which may be of the same sex in 2010 and 2017). Children living with no parent in the household were not included.

    This let me calculate the percentage of children living (at the moment of the survey) with one or more parents who had a PhD. For each of the four groups the percentage of children living with a parent who has a PhD roughly doubled between 1990 and 2017. API children had the highest chance of living with a PhD parent, reaching 6.8 percent in 2017. The percentages for the other groups were: Whites, 2.7 percent; Blacks, 1.0 percent; and Hispanics, 0.7 percent:
     
    All that work resulted in a blog post in which searches for IQ, intelligence, cognitive, confound* come up with no results, except in a quote from Harmon:

    Some misguided people claim that there are not many black research mathematicians because African-Americans are not as intelligent as other races. These people, whom I have reported on for other stories in recent months, almost invariably use mathematical accomplishment as their yardstick for intelligence. They note that no individuals of African descent have won the Fields Medal, math’s equivalent of the Nobel Prize. They lack any genetic evidence to explain the gap in average I.Q. scores between white and black Americans that they cite as the basis of their belief, or reason to think that a genetic trait would be impervious to social or educational intervention, or that high I.Q. is key to math ability, which Timothy Gowers, a 1998 Fields medalist, has attributed largely to “the capacity to become obsessed with a math problem.”

    But I have been reporting on these topics for several years, and I am acutely aware that math prowess factors heavily into the popular conception of intelligence. There’s a vicious cycle at work: The lack of African-American representation in math can end up feeding pernicious biases, which in turn add to the many obstacles mathematically talented minorities face. Which was one more reason it seemed especially important to hold up to the light all the racial biases that contribute to that underrepresentation.
     
    If there is an IQ gap between blacks and whites, whatever the cause, that is sufficient to explain the lack of black math talent, is it not? If the gap is 100 percent environmental, it nevertheless explains the lack of black math talent and the lack of black math research academics.

    If the gap is environmental, say lead in water or fetal alcohol syndrome, then the environmental cause permanently affected the brain. Is she hinting that there is some environmental cause that would instantly liberate the brains of adult blacks if found, say Beyonce music overdosing? Even it that were true, you'd have to identify the cause and apply the fix before black math whizes could perform their whizziness. In the meantime they would be treading water in their jobs. Is she suggesting that blacks be hired as math professors in advance to have them in place when the real-soon-now discovery of the environmental cause of their dullness is uncovered?

    Replies: @Buffalo Joe, @Gringo, @Reg Cæsar

    Anon, two thoughts, a lot of blacks have Doctorates in Education, does that count as a PhD.? And two, people doing the hiring really don’t care if some one’s parent or parents have PhDs, they are hiring the applicant.

    • Replies: @Arclight
    @Buffalo Joe

    "Education" is one of the easiest bachelor's degrees to obtain, so I'd guess a PhD in that field is a pretty low bar as well - basically just requires the ability/desire to remain in school long enough. Given the number of blacks who are pastors of one variety or another, are there a bunch of doctorates of divinity in that 2.7 statistic? Seems surprising to me that the ratio is that low.

  43. Clearly all you commenters have been tainted by Hatefacts, and the only wise course had to ship you all to the gulag. White Hatefacts must be liquidated!!

  44. @The Practical Conservative
    In 2015, around 25% fewer black kids took the test, but around 4x the 2005 number got above a 750. Could be the tyranny of tiny numbers, but I also thought black kids took the SAT at the same or higher rate than their genpop representation.

    Replies: @Spangel, @res

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns “Unauthorized.” It’s available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html

    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation’s most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation’s 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.

    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    • Replies: @Gringo
    @res

    (did I miss the link above? pretty sure someone linked this in the other thread):

    I left out the link because when I supplied the link yesterday, comments that were made after my comment got published, while mine stayed in moderation. I left out the link in the hope my comment would get published. And it got published. Yes, I did publish a comment with the link earlier in the week.

    , @res
    @res

    I went back and looked at that Brookings report again. Turns out I had missed an important detail (emphasis mine).


    The College Board’s publicly available data provides data on racial composition at 50-point score intervals. We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos. (These estimates—which rely on conservative assumptions that maximize the number of high-scoring black students, are consistent with an older estimate from a 2005 paper in the Journal of Blacks in Higher Education, which found that only 244 black students scored above a 750 on the math section of the SAT.)
     
    Oh boy. If they consider "at most 1,000" consistent with 244 that says something. What's a factor of 4 between friends?

    Does anyone know where to find the publicly available "data on racial composition at 50-point score intervals"?

    Replies: @Gringo

    , @Mr. XYZ
    @res

    It seems like it would be a really good idea for the ultra-smart Blacks to have lots and lots of babies. That way, they would increase the odds of the entire Black population gradually looking more and more like them.

    Replies: @Reg Cæsar

    , @res
    @res

    The Brookings numerical analysis has a strong smell of fraud about it, so I decided to take a closer look at various numbers they gave and check them for consistency. Here are some salient points from the Brookings article and associated ETS documents. I believe these are all using the 2015 ETS data.


    We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.
     
    Table 7 of https://secure-media.collegeboard.org/digitalServices/pdf/sat/total-group-2015.pdf has scoring distributions and demographics

    The Brookings graphic (Race gaps at the tails, SAT math scores) I linked above has estimated percentiles scoring above 750.

    I combined those sources into a spreadsheet (with formulas in original appearing as computed values in CSV). It appears after the MORE tag as a CSV if anyone wants to take a look. There are some issues because Brookings lumps three different Latino groups together and ignores the "Other" and "No Response" categories which I estimate account for 10% of Math scores over 750.

    Using the Table 7 numbers I calculated the expected number of scores over 750 assuming a normal distribution. The black estimate was 140 (!). I suspect the distribution is fat tailed so I find the JBHE 244 number at least plausible (but would like to see details of how they computed it). But both of those numbers are only loosely related to "at most 1000." Also, the white estimate is low given the mean/SD. Odd.

    I suspect what Brookings did was make assumptions about the "Other" and "No Response" categories which favored the black estimates.

    In any case, I think my analysis shows their presentation was far enough from reality to be considered fraudulent. It would be interesting to see a better calculation using the full ETS data (which is quite possible). I wonder why that is not available?

    P.S. Here is another broken link from the Brookings article. Funny how the most informative links seem to be the ones that break.
    https://web.archive.org/web/20160310073125/https://cshe.berkeley.edu/sites/default/files/shared/publications/docs/ROPS.CSHE_.10.15.Geiser.RaceSAT.10.26.2015.pdf



    Race,N,Math Mean,Math SD,SDs for 750,Est prop 750,Est N 750,Est percent 750,Brookings Est N,Ratio of Estimates,Brookings percent 750 as N,Est percent 750
    American Indian,"10,031",482,107,2.50,0.0061,61,0.14%,,,,
    Asian,"211,238",598,127,1.20,0.1157,24437,55.06%,29570,0.83,26628,
    Black,"219,018",428,100,3.22,0.0006,140,0.32%,1000,0.14,888,
    Mexican,"130,026",457,98,2.99,0.0014,181,0.41%,967,0.19,894,
    Puerto Rican,"30,192",449,106,2.84,0.0023,68,0.15%,224,0.30,208,
    Other Hispanic,"162,655",457,107,2.74,0.0031,502,1.13%,1209,0.42,1118,
    White,"800,236",534,104,2.08,0.0189,15128,34.09%,"16,000",0.95,14646,
    Other ,"65,063",519,123,1.88,0.0302,1964,4.43%,,,,
    No Response,"70,062",492,134,1.93,0.0271,1898,4.28%,,,,
    Total,"1,698,521",511,120,1.99,0.0232,39412,88.81%,,,,
    ,,,,,,,,,,,
    Calculated Totals,"1,698,521",510.9,108.2,,,44380,,48970,0.91,44380,

    Replies: @Jack D, @dux.ie

  45. @Anon
    A sociologist chimes in:

    https://familyinequality.wordpress.com/2019/02/20/white-children-are-2-7-times-more-likely-than-black-children-to-live-with-a-parent-who-has-a-phd/

    This guy was responsible for a statistic that Harmon used in her followup piece.

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD

    For a reflection Amy Harmon was working on, a followup to her article on the experience of Black mathematicians in American academia, I took a shot at the question: How many children have parents with PhDs?

    The result was the highlighted passage (17 words and a link!) in her piece:

    [all the racial biases that contribute to Black underrepresentation include] the well-documented racial disparities in public-school resources, the selection of students for gifted programs — and the fact that having a parent with a Ph.D. is helpful to getting one in math, while black children are less than half as likely as white children to live with such a parent.

    To get there: I used data from the U.S. Census Bureau via IPUMS.org: The 1990 5% Public Use Microdata Sample (decennial census); and the 2000, 2010, and 2017 American Community Surveys.

    I coded race/ethnicity into four mutually-exclusive categories: Single-race White, Black, and Asian/Pacific Islander (API); and Hispanic (including those of any race). I dropped from the analysis non-Hispanic children with multiple races reported, and American Indian / Alaska Natives (for whom about 0.5 percent lived with a PhD parent in 2017).

    IPUMS made a tool that attaches values of parents’ variables to children with whom they share a household. I used that to calculate the highest level of education of each child’s coresident parents. In the Census data, children may have up to two parents present (which may be of the same sex in 2010 and 2017). Children living with no parent in the household were not included.

    This let me calculate the percentage of children living (at the moment of the survey) with one or more parents who had a PhD. For each of the four groups the percentage of children living with a parent who has a PhD roughly doubled between 1990 and 2017. API children had the highest chance of living with a PhD parent, reaching 6.8 percent in 2017. The percentages for the other groups were: Whites, 2.7 percent; Blacks, 1.0 percent; and Hispanics, 0.7 percent:
     
    All that work resulted in a blog post in which searches for IQ, intelligence, cognitive, confound* come up with no results, except in a quote from Harmon:

    Some misguided people claim that there are not many black research mathematicians because African-Americans are not as intelligent as other races. These people, whom I have reported on for other stories in recent months, almost invariably use mathematical accomplishment as their yardstick for intelligence. They note that no individuals of African descent have won the Fields Medal, math’s equivalent of the Nobel Prize. They lack any genetic evidence to explain the gap in average I.Q. scores between white and black Americans that they cite as the basis of their belief, or reason to think that a genetic trait would be impervious to social or educational intervention, or that high I.Q. is key to math ability, which Timothy Gowers, a 1998 Fields medalist, has attributed largely to “the capacity to become obsessed with a math problem.”

    But I have been reporting on these topics for several years, and I am acutely aware that math prowess factors heavily into the popular conception of intelligence. There’s a vicious cycle at work: The lack of African-American representation in math can end up feeding pernicious biases, which in turn add to the many obstacles mathematically talented minorities face. Which was one more reason it seemed especially important to hold up to the light all the racial biases that contribute to that underrepresentation.
     
    If there is an IQ gap between blacks and whites, whatever the cause, that is sufficient to explain the lack of black math talent, is it not? If the gap is 100 percent environmental, it nevertheless explains the lack of black math talent and the lack of black math research academics.

    If the gap is environmental, say lead in water or fetal alcohol syndrome, then the environmental cause permanently affected the brain. Is she hinting that there is some environmental cause that would instantly liberate the brains of adult blacks if found, say Beyonce music overdosing? Even it that were true, you'd have to identify the cause and apply the fix before black math whizes could perform their whizziness. In the meantime they would be treading water in their jobs. Is she suggesting that blacks be hired as math professors in advance to have them in place when the real-soon-now discovery of the environmental cause of their dullness is uncovered?

    Replies: @Buffalo Joe, @Gringo, @Reg Cæsar

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD.

    My immediate reaction: so what? Not everyone wants to jump through the hoops to get a doctorate. And yes, jumping through hoops is what it involves. I didn’t want to. I notice that Amy Harmon didn’t want to.

    I would tend to discount doctorates in the field of Education. Because I taught for several years, I looked at the doctoral dissertations of some colleagues in Education- some who earned them after I left teaching.

    The doctoral dissertations I have read in Education were not about research that could be replicated, but about unique stories- call them narratives, if you will. I’m sure that there are some genuine research-quality dissertations done in Education, but those I have looked at were just stories.

    From yet another paper in the Journal of Blacks in Higher Education, we find out that in 2004, 41.3% of doctorates that blacks earned came in the field of Education. Whites earned 19.1% of their doctorates in Education.
    If you take out Education doctorates from the equation, you can say that “White children are ~3.5 times more likely than Black children to live with a parent who has a real Ph.D.- a doctorate not in the field of Education.”

    .

    http://www.jbhe.com/news_views/50_black_doctoraldegrees.html

    • Replies: @Anon
    @Gringo

    I guess I didn't make my main point explicit enough: The sociologist's reasoning ignores that IQ and EA (parental educational attainment, e.g. the Ph.D. parent), as well as SES (socioeconomic status, e.g. salary and wealth), and many other things like lifespan and health, are all correlated, and you therefore end up with confounded and erroneous results if you assume that they are independent variables, as he did with parental EA and, by implication, child SES (which is a score that comprises income, EA itself, and job prestige). Another way to put it is that Ph.D.s are heritable. So of course blacks have fewer Ph.D.s if black parents have fewer Ph.D.s.

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the "books in the home" or "parents reading to the kids" theories were blown out of the water by twin and adoption studies, and it's a similar idea to "Ph.D.s in the home." Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment. They figure out how to get books. (One objection that has been raised is that twin and adoption studies have not included homes with butt-dumb parents ... they tend not to adopt, so you have to extrapolate down, not good statistically, but common sense tells you that the trend continues.)

    Replies: @Steve Sailer, @Gringo, @Gringo

  46. @res
    @The Practical Conservative

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns "Unauthorized." It's available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    https://i0.wp.com/www.brookings.edu/wp-content/uploads/2017/01/ccf_20170201_reeves_2.png

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html


    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation's most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

     

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation's 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.
     
    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    Replies: @Gringo, @res, @Mr. XYZ, @res

    (did I miss the link above? pretty sure someone linked this in the other thread):

    I left out the link because when I supplied the link yesterday, comments that were made after my comment got published, while mine stayed in moderation. I left out the link in the hope my comment would get published. And it got published. Yes, I did publish a comment with the link earlier in the week.

  47. Next week:

    Amy Harmon demolishes James Watson once and for all, using math to prove that 35% of blacks are smarter than the average white.

  48. @res
    @The Practical Conservative

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns "Unauthorized." It's available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    https://i0.wp.com/www.brookings.edu/wp-content/uploads/2017/01/ccf_20170201_reeves_2.png

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html


    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation's most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

     

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation's 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.
     
    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    Replies: @Gringo, @res, @Mr. XYZ, @res

    I went back and looked at that Brookings report again. Turns out I had missed an important detail (emphasis mine).

    The College Board’s publicly available data provides data on racial composition at 50-point score intervals. We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos. (These estimates—which rely on conservative assumptions that maximize the number of high-scoring black students, are consistent with an older estimate from a 2005 paper in the Journal of Blacks in Higher Education, which found that only 244 black students scored above a 750 on the math section of the SAT.)

    Oh boy. If they consider “at most 1,000” consistent with 244 that says something. What’s a factor of 4 between friends?

    Does anyone know where to find the publicly available “data on racial composition at 50-point score intervals”?

    • Replies: @Gringo
    @res

    Oh boy. If they consider “at most 1,000” consistent with 244 that says something. What’s a factor of 4 between friends?

    Brookings informs us that whites form a decreasing percentage at higher test intervals.
    650-700 62% whites
    700-75o 54% whites
    750-800 33% whites

    But blacks do not form a decreasing percentage at higher test intervals.
    650-700 2% blacks
    700-75o 2% blacks
    750-800 2% blacks

    That looks to me like cooked data, which is implied in "at most."

  49. Wow, finally. I’m no longer the only true race-ist.

    https://twitter.com/SenatorNot/status/1074513299121807360

  50. @Joe, Averaged
    @Hypnotoad666

    Yes that’s what you get from American Studies majors. Didn’t journalism used to be a real profession?

    Replies: @Brutusale, @Gringo, @MBlanc46

    No. Journalists used to be high school grads who could write a bit and were very good at extracting information from people.

  51. Average black Math SAT is 425 and SD is 101. I guess we’ll never know why there aren’t more black mathematicians.

  52. OT

    A question people who call Jussie Smollett “Jussie” ought to ask themselves (kleine Ergänzung: he duct taped two women, one the clerk the other a customer, poured inflammable liquid on them, and ignited it):

    Meanwhile in news you didn’t hear, a black man robbed a gas station, tied up the white clerk, poured gas on her and set her on fire. Was this extreme level of hate motivated by Jussie Smollett’s lies?

    • Replies: @Nathan
    @eah

    Oh my God. Aggravated assault and aggravated robbery??? What's pathetic in this story are the light charges brought against obvious attempted murder and terrorism. The man should hang.

    , @El Dato
    @eah


    Was this extreme level of hate motivated by Jussie Smollett’s lies?
     
    Who knows what goes on in a gangbanger's head? Could 5.56 mm find out?

    The US are really sub-saharan in the fun level department now.
  53. I’m trying to rent an apartment on Craigslist. I find reading comprehension skills useful in replying(or not) to an applicant.

  54. @eah
    NY -- State Education Department Releases Spring 2018 Grades 3-8 Ela & Math Assessment Results

    https://i.postimg.cc/44p3Tj44/ny-2018.png

    Replies: @eah

    It’s a fair bet that most math Ph.Ds. got 750 or above on the Math SAT.

    You’ll never convince people like Harmon that any gap — the 750+ SAT gap, black math professor gap, whatever gap — isn’t due to ‘institutional racism’ etc — the real problem is the dramatically low absolute level of achievement (ability) shown by a very large and growing fraction of the population (except it’s basic arithmetic, not basic “math”).

    So I’ve covered the big 3 metropolitan areas today: NYC, LA, Chicago/Illinois — These places are pumping out people who can’t do basic math. Think on the consequences.

  55. Hey, hey, hey! No fair crunching the numbers on how many people actually achieve top scores on college admissions tests. Pretty soon you’ll notice that there aren’t enough top scores to back up the admissions claims of elite colleges. Can’t have that.

    • Replies: @res
    @Nathan


    Pretty soon you’ll notice that there aren’t enough top scores to back up the admissions claims of elite colleges.
     
    Sounds plausible. Do you have an analysis to back that up?

    Replies: @Nathan

  56. @eah
    OT

    A question people who call Jussie Smollett "Jussie" ought to ask themselves (kleine Ergänzung: he duct taped two women, one the clerk the other a customer, poured inflammable liquid on them, and ignited it):

    Meanwhile in news you didn't hear, a black man robbed a gas station, tied up the white clerk, poured gas on her and set her on fire. Was this extreme level of hate motivated by Jussie Smollett's lies?

    https://media.easttexasmatters.com/nxsglobal/myeasttex/photo/2019/02/21/Palestine_Robbery_Surveillance_Video_0_74202909_ver1.0_1280_720.jpg

    Replies: @Nathan, @El Dato

    Oh my God. Aggravated assault and aggravated robbery??? What’s pathetic in this story are the light charges brought against obvious attempted murder and terrorism. The man should hang.

  57. 0.7% is higher than I’d expect.

    I don’t want to troll, it really is higher than I’d expect. 1 in 150 math profs is black? Amazing. Never seen one, ever.

  58. @Joe, Averaged
    Come on, Amy Harmon couldn’t do the math to figure that out EVEN IF she wanted to. To her they remain “hidden figures.”

    Replies: @Mr. XYZ

    How ironic, given the high average IQ of the ethnic group that she likely belongs to.

  59. @res
    @The Practical Conservative

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns "Unauthorized." It's available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    https://i0.wp.com/www.brookings.edu/wp-content/uploads/2017/01/ccf_20170201_reeves_2.png

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html


    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation's most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

     

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation's 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.
     
    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    Replies: @Gringo, @res, @Mr. XYZ, @res

    It seems like it would be a really good idea for the ultra-smart Blacks to have lots and lots of babies. That way, they would increase the odds of the entire Black population gradually looking more and more like them.

    • Replies: @Reg Cæsar
    @Mr. XYZ


    It seems like it would be a really good idea for the ultra-smart Blacks to have lots and lots of babies.
     
    I don't know about "ultra-smart" ones, but the competent ones, the middle- and upper-middle-class ones, have about the lowest TFR of any subset of the population. Well, any straight, not disabled subset.

    It's like they can put everything into their careers, or into their kids, and do fine. But not both.
  60. Only 0.7 blacks score > 750 on SAT math

    And I bet every single one of them got into all the Ivies, MIT, Caltech. These are the black geniuses who are going to send us to Mars!

    • Replies: @Foreign Expert
    @IQ stuff

    I bet that the 0.7% were in fact half black.

  61. @Anon
    A sociologist chimes in:

    https://familyinequality.wordpress.com/2019/02/20/white-children-are-2-7-times-more-likely-than-black-children-to-live-with-a-parent-who-has-a-phd/

    This guy was responsible for a statistic that Harmon used in her followup piece.

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD

    For a reflection Amy Harmon was working on, a followup to her article on the experience of Black mathematicians in American academia, I took a shot at the question: How many children have parents with PhDs?

    The result was the highlighted passage (17 words and a link!) in her piece:

    [all the racial biases that contribute to Black underrepresentation include] the well-documented racial disparities in public-school resources, the selection of students for gifted programs — and the fact that having a parent with a Ph.D. is helpful to getting one in math, while black children are less than half as likely as white children to live with such a parent.

    To get there: I used data from the U.S. Census Bureau via IPUMS.org: The 1990 5% Public Use Microdata Sample (decennial census); and the 2000, 2010, and 2017 American Community Surveys.

    I coded race/ethnicity into four mutually-exclusive categories: Single-race White, Black, and Asian/Pacific Islander (API); and Hispanic (including those of any race). I dropped from the analysis non-Hispanic children with multiple races reported, and American Indian / Alaska Natives (for whom about 0.5 percent lived with a PhD parent in 2017).

    IPUMS made a tool that attaches values of parents’ variables to children with whom they share a household. I used that to calculate the highest level of education of each child’s coresident parents. In the Census data, children may have up to two parents present (which may be of the same sex in 2010 and 2017). Children living with no parent in the household were not included.

    This let me calculate the percentage of children living (at the moment of the survey) with one or more parents who had a PhD. For each of the four groups the percentage of children living with a parent who has a PhD roughly doubled between 1990 and 2017. API children had the highest chance of living with a PhD parent, reaching 6.8 percent in 2017. The percentages for the other groups were: Whites, 2.7 percent; Blacks, 1.0 percent; and Hispanics, 0.7 percent:
     
    All that work resulted in a blog post in which searches for IQ, intelligence, cognitive, confound* come up with no results, except in a quote from Harmon:

    Some misguided people claim that there are not many black research mathematicians because African-Americans are not as intelligent as other races. These people, whom I have reported on for other stories in recent months, almost invariably use mathematical accomplishment as their yardstick for intelligence. They note that no individuals of African descent have won the Fields Medal, math’s equivalent of the Nobel Prize. They lack any genetic evidence to explain the gap in average I.Q. scores between white and black Americans that they cite as the basis of their belief, or reason to think that a genetic trait would be impervious to social or educational intervention, or that high I.Q. is key to math ability, which Timothy Gowers, a 1998 Fields medalist, has attributed largely to “the capacity to become obsessed with a math problem.”

    But I have been reporting on these topics for several years, and I am acutely aware that math prowess factors heavily into the popular conception of intelligence. There’s a vicious cycle at work: The lack of African-American representation in math can end up feeding pernicious biases, which in turn add to the many obstacles mathematically talented minorities face. Which was one more reason it seemed especially important to hold up to the light all the racial biases that contribute to that underrepresentation.
     
    If there is an IQ gap between blacks and whites, whatever the cause, that is sufficient to explain the lack of black math talent, is it not? If the gap is 100 percent environmental, it nevertheless explains the lack of black math talent and the lack of black math research academics.

    If the gap is environmental, say lead in water or fetal alcohol syndrome, then the environmental cause permanently affected the brain. Is she hinting that there is some environmental cause that would instantly liberate the brains of adult blacks if found, say Beyonce music overdosing? Even it that were true, you'd have to identify the cause and apply the fix before black math whizes could perform their whizziness. In the meantime they would be treading water in their jobs. Is she suggesting that blacks be hired as math professors in advance to have them in place when the real-soon-now discovery of the environmental cause of their dullness is uncovered?

    Replies: @Buffalo Joe, @Gringo, @Reg Cæsar

    White children are 2.7-times more likely

    As, not more. G’s Louise!

    There’s a logical difference.

  62. @Mr. XYZ
    @res

    It seems like it would be a really good idea for the ultra-smart Blacks to have lots and lots of babies. That way, they would increase the odds of the entire Black population gradually looking more and more like them.

    Replies: @Reg Cæsar

    It seems like it would be a really good idea for the ultra-smart Blacks to have lots and lots of babies.

    I don’t know about “ultra-smart” ones, but the competent ones, the middle- and upper-middle-class ones, have about the lowest TFR of any subset of the population. Well, any straight, not disabled subset.

    It’s like they can put everything into their careers, or into their kids, and do fine. But not both.

  63. Anon[387] • Disclaimer says:
    @Gringo
    @Anon

    White children are 2.7-times more likely than Black children to live with a parent who has a PhD.

    My immediate reaction: so what? Not everyone wants to jump through the hoops to get a doctorate. And yes, jumping through hoops is what it involves. I didn't want to. I notice that Amy Harmon didn't want to.

    I would tend to discount doctorates in the field of Education. Because I taught for several years, I looked at the doctoral dissertations of some colleagues in Education- some who earned them after I left teaching.

    The doctoral dissertations I have read in Education were not about research that could be replicated, but about unique stories- call them narratives, if you will. I'm sure that there are some genuine research-quality dissertations done in Education, but those I have looked at were just stories.

    From yet another paper in the Journal of Blacks in Higher Education, we find out that in 2004, 41.3% of doctorates that blacks earned came in the field of Education. Whites earned 19.1% of their doctorates in Education.
    If you take out Education doctorates from the equation, you can say that "White children are ~3.5 times more likely than Black children to live with a parent who has a real Ph.D.- a doctorate not in the field of Education."

    .

    http://www.jbhe.com/news_views/50_black_doctoraldegrees.html

    Replies: @Anon

    I guess I didn’t make my main point explicit enough: The sociologist’s reasoning ignores that IQ and EA (parental educational attainment, e.g. the Ph.D. parent), as well as SES (socioeconomic status, e.g. salary and wealth), and many other things like lifespan and health, are all correlated, and you therefore end up with confounded and erroneous results if you assume that they are independent variables, as he did with parental EA and, by implication, child SES (which is a score that comprises income, EA itself, and job prestige). Another way to put it is that Ph.D.s are heritable. So of course blacks have fewer Ph.D.s if black parents have fewer Ph.D.s.

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the “books in the home” or “parents reading to the kids” theories were blown out of the water by twin and adoption studies, and it’s a similar idea to “Ph.D.s in the home.” Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment. They figure out how to get books. (One objection that has been raised is that twin and adoption studies have not included homes with butt-dumb parents … they tend not to adopt, so you have to extrapolate down, not good statistically, but common sense tells you that the trend continues.)

    • Replies: @Steve Sailer
    @Anon

    I suspect it's pretty likely that parents' jobs have a sizable influence on what jobs kids go into. A lot of military officers are the children of a military officer, a lot of college professors are the children of a college professor. Some of that is genetic, but another part is simply that you can imagine yourself doing that job because you saw somebody in the family do it.

    Replies: @Anon, @Anonymous

    , @Gringo
    @Anon

    I wasn't disagreeing with you. I was simply adding my reaction to the "2.7 times" statement. I find it especially silly that Amy Harmon, who doesn't have a Ph.D. and probably never tried to get one, is kvetching about this. I doubt she has any idea at all what hoop jumping it takes to get a doctorate.

    I am reminded of my sojourn in the teaching profession. I took an Education course where a female student was kvetching about the low percentage of female students in some engineering department. My sister worked as a Chem E for 25 years, which informs me that gender isn't the main issue in becoming an engineer. The main issue is that bright students need to spend 60 hours a week in class and in studying to get an engineering degree. That female Education student had neither the intellect nor the work ethic to make it as an engineer. As such, her complaint was absurd.

    If students want to put up with the hoop jumping to get a doctorate, more power to them. I see no need to complain about those who do not so choose. It is their choice not to pursue a doctorate.

    , @Gringo
    @Anon

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the “books in the home” or “parents reading to the kids” theories were blown out of the water by twin and adoption studies, and it’s a similar idea to “Ph.D.s in the home.” Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment.

    It also works the other way : not as bright kids adopted into the homes of bright parents. What I personally know about adoption is anecdote, not a research finding, but I will relate it nonetheless. I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.

    Replies: @Anon, @Jack D

  64. @Alan Mercer
    750 on math is (as of 2017) good for 98th percentile of all students (estimated), 96th percentile of SAT takers (source: College Board). Sorry, but not 1 in 25 SAT takers can go on to get a math phd. That those two figures match demonstrates the pro-black bias.

    Replies: @Thulean Friend, @Spangel, @International Jew, @JimS

    Missing even 1 question can take 40 points off

    Mm, doubtful. Says here
    https://blog.prepscholar.com/how-is-the-sat-scored-scoring-charts
    you get 750 if you miss about five (out of the usual 58) math questions.

    Never mind becoming a math prof or even completing a PhD in math. For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT…) the math SAT is ridiculously easy. All of it. There’s never a single question that’s hard in any way. If you didn’t get 800, you got 790 for some bit of carelessness. And I’m describing the state of affairs when I took it, which was before 1995, when it was made easier. You didn’t even brag about your SAT (not that we were a modest bunch), you bragged about your exploits in the high school math league, if not in international competitions (where the easiest question is harder than the hardest question that’s ever been seen on the SAT).

    To make a long story short, if only 0.7% of black kids can get 750 on the SAT (today’s SAT — pre-1995 that was like a 700) then the fraction that has what it takes to become a math prof at a top-50 school is way lower.

    • Replies: @Pericles
    @International Jew



    For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT…) the math SAT is ridiculously easy.

     

    MIT is good, don't know about Harvard. It's more about the hobnobbing, isn't it? Princeton has awarded at least one PhD in math to a black woman, who wrote a highly amusing thesis. (Princeton is trash.)
    , @Jack D
    @International Jew

    The exact # translation between raw score and scaled score differs with each version of the test because they "curve" each one to fit the normal distribution. As more and more Asians have gotten better and better at the test, in some versions even 1 question wrong is a 780, there being no way to get a 790. The raw to scale translation given in that blog post are illustrative and YMMV. Maybe there is some version of the test where 1 wrong is a 760 - I dunno but it's not impossible.

    But otherwise I agree with you. The 25th percentile Math SAT at MIT is 770 and it would be even higher if they didn't have AA for blacks and women. I would guess that virtually all of the Asian male applicants have 800s (or as you say, maybe one wrong thru some careless mistake). A 770+ Math score is just a gatekeeping device for a place like MIT and they have to separate the wheat from the chaff in that pool using other means. They admit around 1,500 of the 21,000 applications that they get. Maybe 1/3 or 1/4 of those applications are Hail Mary plays with sub- 770 Math SATs and they can trash most of those on the 1st reading (at least from white and Asian applicants) but that still leaves them with something like ten 770+ applicants for every person they admit (probably a significant chunk of all the 770+ SAT scorers apply to MIT as one of their choices) so they have to look at other things beyond just the SATs.

    , @res
    @International Jew


    if only 0.7% of black kids can get 750 on the SAT (today’s SAT — pre-1995 that was like a 700)
     
    Math didn't change that much (e.g. old 750 = new 760). Verbal was the test that was really dumbed down (e.g. old 730 = new 800).

    Conversion chart: https://www.greenes.com/html/convert.htm

    P.S. I suspect the test scoring has been made even less rigorous since then through the various changes over the years since 1996, but good evidence to support that is scarce.
  65. @Anon
    @Gringo

    I guess I didn't make my main point explicit enough: The sociologist's reasoning ignores that IQ and EA (parental educational attainment, e.g. the Ph.D. parent), as well as SES (socioeconomic status, e.g. salary and wealth), and many other things like lifespan and health, are all correlated, and you therefore end up with confounded and erroneous results if you assume that they are independent variables, as he did with parental EA and, by implication, child SES (which is a score that comprises income, EA itself, and job prestige). Another way to put it is that Ph.D.s are heritable. So of course blacks have fewer Ph.D.s if black parents have fewer Ph.D.s.

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the "books in the home" or "parents reading to the kids" theories were blown out of the water by twin and adoption studies, and it's a similar idea to "Ph.D.s in the home." Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment. They figure out how to get books. (One objection that has been raised is that twin and adoption studies have not included homes with butt-dumb parents ... they tend not to adopt, so you have to extrapolate down, not good statistically, but common sense tells you that the trend continues.)

    Replies: @Steve Sailer, @Gringo, @Gringo

    I suspect it’s pretty likely that parents’ jobs have a sizable influence on what jobs kids go into. A lot of military officers are the children of a military officer, a lot of college professors are the children of a college professor. Some of that is genetic, but another part is simply that you can imagine yourself doing that job because you saw somebody in the family do it.

    • Replies: @Anon
    @Steve Sailer

    In Japan a surprising number of doctors have a doctor parent, more so I think than in the U.S.

    My theory is that in Japan you go right into medical school after high school, like pharmacy school in the U.S. So it's not a career path decision that you can put off too long. You pretty much have to have made the decision by the time you enter high school at the latest, and without the guidance of a physician parent a lot of early teens are not going to be thinking about stuff like this.

    In the U.S. on the other hand if you're a STEM major with strong grades you can wait until you've almost graduated college to choose a medical career.

    And there are a few medical schools that have slots for non-STEM people, although I expect this is almost the harder route, given the few number of such slots:

    https://www.nytimes.com/2010/07/30/nyregion/30medschools.html


    For generations of pre-med students, three things have been as certain as death and taxes: organic chemistry, physics and the Medical College Admission Test, known by its dread-inducing acronym, the MCAT.

    So it came as a total shock to Elizabeth Adler when she discovered, through a singer in her favorite a cappella group at Brown University, that one of the nation’s top medical schools admits a small number of students every year who have skipped all three requirements.

    Until then, despite being the daughter of a physician, she said, “I was kind of thinking medical school was not the right track for me.”
     
    , @Anonymous
    @Steve Sailer

    I have a good sample of elite math grad students. A lot of the (American) boys were the children of profs, often in other fields. But the girls! A full half had fathers who were math/physics profs.

    Replies: @Steve Sailer

  66. A lot of military officers are the children of a military officer…

    As are a lot of, well, anything-but-military types. Eg, Jim Morrison and Kris Kristofferson.

    You can go all the way, or all the other way. Or part way– more anodyne military brat acts were America and Katrina and the Waves, both of whom came out of USAF bases in England.

    • Replies: @William Badwhite
    @Reg Cæsar

    Kristofferson also served in the Army as an officer. He was a helicopter pilot.

    Replies: @Reg Cæsar, @Hibernian

  67. @Alan Mercer
    750 on math is (as of 2017) good for 98th percentile of all students (estimated), 96th percentile of SAT takers (source: College Board). Sorry, but not 1 in 25 SAT takers can go on to get a math phd. That those two figures match demonstrates the pro-black bias.

    Replies: @Thulean Friend, @Spangel, @International Jew, @JimS

    Agreed. The test of reference is unequivocally the Putnam Competition, generally participated in by college seniors who are serious about mathematics. By my imperfect recollection, that test is comprised of a total of 12 questions, scored from 0-120, with roughly half of all takers scoring a 0.

    I highly doubt Putnam would provide a racial breakdown of performance. If it did, the obvious questions would be, what is the fraction of test takers that were black? In other words, how many are even of a level of mathematics such that they would even attempt the test? And the second question, how many blacks actually register a nonzero score?

    (Snarky question, to be disregarded in all serious discussion: Is the difference in numbers between those two answers exactly 13? For the record, I would seriously believe it to be much higher.)

  68. @Thulean Friend
    @Alan Mercer

    Remember that not everyone is taking their PhD at an Ivy League school. UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD. You add the affirmative action factor and you're more than set. Plus far from every person with a 750 math score or above wants to be in academia.

    If you score that high as a black person, your options are virutally unlimited in the corporate sector. Lots of affirmative action jobs available for pretty smart blacks since there are so few to begin with. That scarcity must be affecting the PhD job market, too, for universities who want to diversify.

    Replies: @PhysicistDave

    Thulean Friend wrote:

    UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD.

    No, I don’t think so. As others have said, for those of us who are reasonably good at math, an 800 on the old SAT Math, before they made the scoring easier, was not very difficult.

    When I was at Caltech back in the ’70s, we sort of looked down our noses at the students who were somehow admitted with less than an 800 on math: we figured they might be okay in biology.

    Yes, Caltech is harder than anywhere else except MIT, but I am talking undergrads, most of whom were not math majors. I would bet that a Ph.D. in math, even at UMass Amherst, requires being as good at math as the average BS chemical engineer from Caltech.

    • Replies: @Thulean Friend
    @PhysicistDave


    I would bet that a Ph.D. in math, even at UMass Amherst, requires being as good at math as the average BS chemical engineer from Caltech.
     
    UMass Amherst is still more like a 2nd tier school, remember that there are 3rd and 4th tier ones putting out math PhDs too. I'd highly doubt that having a perfect math SAT score at those colleges would be required for a math PhD, especially given the affirmative action factor. You're too blinded by the fact that you went to what is arguably the US' most competitive university.
  69. @IQ stuff

    Only 0.7 blacks score > 750 on SAT math
     
    And I bet every single one of them got into all the Ivies, MIT, Caltech. These are the black geniuses who are going to send us to Mars!

    Replies: @Foreign Expert

    I bet that the 0.7% were in fact half black.

  70. It’s even less than 0.7% if you exclude Black immigrants.

    Anecdotal evidence: In the year I got my Math PhD, 1100 doctoral degrees in Math were awarded. Of those, two were awarded to Blacks. By accident, I knew both of them. One was an immigrant from Nigeria, the other was from Belize.

    • Replies: @Triumph104
    @Undocumented Shopper

    Is Arlie Petters the guy from Belize? He identifies as being "two or more races", although he is always referred to as black or African-American.

    Former NFL player John Urschel is scheduled to earn his math PhD from MIT this spring. He was born in Winnipeg, Canada and has a black mother and white father, but you wouldn't know that from his media coverage.

    Replies: @Undocumented Shopper, @Buffalo Joe

    , @Jack D
    @Undocumented Shopper

    I think this is true in general of that whole 750+ SAT group. Actual Moochelle type American slave descendants (other than light skinned mulattoes from way back - Louisiana Creoles, etc.) are almost completely missing from that group. Instead you have blacks with one white parent, Caribbean blacks, Nigerian Igbos, etc.

    Replies: @Undocumented Shopper

  71. NYT:

    NASA Renames Facility After Katherine Johnson of ‘Hidden Figures’ Fame
    By Elisha Brown
    Feb. 23, 2019

    NASA on Friday officially renamed a facility in West Virginia after Katherine Johnson, an African-American mathematician and centenarian whose barrier-breaking career was depicted in the film “Hidden Figures.”

    NASA’s Katherine Johnson Independent Verification and Validation Facility in Fairmont, W.Va. The program housed at the facility monitors the software used to track high-profile NASA missions.

    • Replies: @El Dato
    @MEH 0910


    NASA has redesignated its Independent Verification and Validation (IV&V) Facility in Fairmont, West Virginia, as the Katherine Johnson Independent Verification and Validation Facility, in honor of the West Virginia native and NASA "hidden figure."

    ....

    President Donald Trump signed into law in December an act of Congress calling for the redesignation. The facility's program contributes to the safety and success of NASA's highest-profile missions by assuring that mission software performs correctly. IV&V now is in the process of planning a rededication ceremony.

     

    Why the hell not. IV&V sounds a bit evil.

    Ms. Johnson, who turned 100 in August, “remains in awe and honored by” the accolades she has received, Joylette Hylick, one of Ms. Johnson’s daughters, said on Friday. Ms. Hylick said her mother “can’t imagine why people would want to honor her for just doing a good job.”

    NASA’s decision to name the facility for Ms. Johnson is not the first time she has been celebrated on the national stage. In 2015, President Barack Obama awarded her the Presidential Medal of Freedom, the nation’s highest civilian honor.

    Ms. Johnson was not always respected. Called “colored computers,” she and other black women who worked in NASA’s computing pool more than a half-century ago were separated from their white colleagues while they calculated trajectories for the Apollo missions and other programs.
     
    Colored computers -- definitely not standard "office beige"!
  72. @Reg Cæsar
    Amir Aczel found the earliest known representation of zero in a Cambodian jungle, and Terry Jones located the earliest known one of one in a temple in India, though he was locked out.


    FINDING ZERO
    A Mathematician's Odyssey to Uncover the Origins of Numbers


    The Story of 1


    The earliest known something must be in Africa. How about the earliest known emigrant? "There goes the neighborhood! I'm crossing the isthmus, where the grass will be greener."

    Replies: @Logan, @El Dato, @The Alarmist, @Neil Templeton

    Amir Aczel found the earliest known representation of zero in a Cambodian jungle

    Indiana Jones and the Representation of Nothing!

  73. @eah
    OT

    A question people who call Jussie Smollett "Jussie" ought to ask themselves (kleine Ergänzung: he duct taped two women, one the clerk the other a customer, poured inflammable liquid on them, and ignited it):

    Meanwhile in news you didn't hear, a black man robbed a gas station, tied up the white clerk, poured gas on her and set her on fire. Was this extreme level of hate motivated by Jussie Smollett's lies?

    https://media.easttexasmatters.com/nxsglobal/myeasttex/photo/2019/02/21/Palestine_Robbery_Surveillance_Video_0_74202909_ver1.0_1280_720.jpg

    Replies: @Nathan, @El Dato

    Was this extreme level of hate motivated by Jussie Smollett’s lies?

    Who knows what goes on in a gangbanger’s head? Could 5.56 mm find out?

    The US are really sub-saharan in the fun level department now.

  74. @James Speaks
    Actually experienced nausea thinking about this one.

    1) A Field's Medal or a 750 SAT score is not a reward that can be given to erase misperceptions about slavery, inferiority or microaggressions. They are recognition of and/or accomplishments resulting from talent and hard work.
    2) Any discussion of math disaparities needs to start with arithmetic disparities. I've had the good fortune of trying to teach whites and blacks eighth grade math after they have been thoroughly ruined by black seventh grade math teachers. The most basic problem that occurs frequently is that black teachers do not teach understanding. They only teach a procedure to get the 'right answer' which results in Cargo Cult math.
    3) Any discussion of higher math disparities needs to start with algebra and the abstractions needed to grasp and perform. The inability to deal with simple abstractions prevents any real progress amongst black math students.

    What's scary about all this is the same deficiencies appear in social studies. The ability to make decisions regarding rights and responsibilities of blacks, whites, Asians, Mexicans etc. demands the ability to understand basic facts, to comprehend reality rather than to parrot what black leaders spout out, and to reason abstractly about these facts and the principles of civilization.

    OTOH, I see the manufactured crisis in Venezuela as admission from the PTB that all is not well in the Republic, meaning that at some time not too distant, insufficient food, oil etc will make it imperative for adults to reason effectively, to forge new social groups capable of self preservation, and to chart a path to a sustainable future. Constant pandering to lowered black expectations regarding capability, and phony awards and other meaningless symbols preclude blacks as a group from making this journey.

    Let us all mourn for them and then forget them.

    Replies: @Buffalo Joe, @El Dato

    They only teach a procedure to get the ‘right answer’ which results in Cargo Cult math.

    That’s astonishingly hard to break.

    I was educated according to a rather horribad “Bourbaki” approach (starting with set theory probably because Russell & Whithead decided to build math on this) and it was cargo cult all the way till I was 14 and really decided to get into it with 8+h/week of high school math. In an ideal world it would have been more because progress was slow.

  75. @Reg Cæsar
    Amir Aczel found the earliest known representation of zero in a Cambodian jungle, and Terry Jones located the earliest known one of one in a temple in India, though he was locked out.


    FINDING ZERO
    A Mathematician's Odyssey to Uncover the Origins of Numbers


    The Story of 1


    The earliest known something must be in Africa. How about the earliest known emigrant? "There goes the neighborhood! I'm crossing the isthmus, where the grass will be greener."

    Replies: @Logan, @El Dato, @The Alarmist, @Neil Templeton

    “The earliest known something must be in Africa. How about the earliest known emigrant?”

    This is why paleontologists are brutally supressing any unearthed evidence of hominids outside of Africa that contradicts the Out of Africa theory.

    • Replies: @El Dato
    @The Alarmist

    It's all those UN paleontologists falling out of windows and committing suicide in parked cars?

    "Professor Bonner, he's a real killer!"

  76. @Undocumented Shopper
    It's even less than 0.7% if you exclude Black immigrants.

    Anecdotal evidence: In the year I got my Math PhD, 1100 doctoral degrees in Math were awarded. Of those, two were awarded to Blacks. By accident, I knew both of them. One was an immigrant from Nigeria, the other was from Belize.

    Replies: @Triumph104, @Jack D

    Is Arlie Petters the guy from Belize? He identifies as being “two or more races”, although he is always referred to as black or African-American.

    Former NFL player John Urschel is scheduled to earn his math PhD from MIT this spring. He was born in Winnipeg, Canada and has a black mother and white father, but you wouldn’t know that from his media coverage.

    • Replies: @Undocumented Shopper
    @Triumph104

    Yes, one of those two was Arlie Petters.
    The discrepancy may be due to the fact that in the United States people who have some black ancestry are referred to as "Black". Perhaps Belize does this differently?

    , @Buffalo Joe
    @Triumph104

    Tri, John Urschel graduated from the local Jesuit Prep school, his father is a surgeon and his mother is a nurse. Young Urschel quit the NFL after three productive years as a lineman. He was quoted as saying he didn't want to hurt his brain. His mother used to tell him after every HS, college and pro game....."You don't need to do this." Play football that is.

  77. Earliest example of eating albinos to obtain powers.

  78. @The Alarmist
    @Reg Cæsar


    "The earliest known something must be in Africa. How about the earliest known emigrant?"
     
    This is why paleontologists are brutally supressing any unearthed evidence of hominids outside of Africa that contradicts the Out of Africa theory.

    Replies: @El Dato

    It’s all those UN paleontologists falling out of windows and committing suicide in parked cars?

    “Professor Bonner, he’s a real killer!”

  79. @International Jew
    @Alan Mercer


    Missing even 1 question can take 40 points off
     
    Mm, doubtful. Says here
    https://blog.prepscholar.com/how-is-the-sat-scored-scoring-charts
    you get 750 if you miss about five (out of the usual 58) math questions.

    Never mind becoming a math prof or even completing a PhD in math. For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT...) the math SAT is ridiculously easy. All of it. There's never a single question that's hard in any way. If you didn't get 800, you got 790 for some bit of carelessness. And I'm describing the state of affairs when I took it, which was before 1995, when it was made easier. You didn't even brag about your SAT (not that we were a modest bunch), you bragged about your exploits in the high school math league, if not in international competitions (where the easiest question is harder than the hardest question that's ever been seen on the SAT).

    To make a long story short, if only 0.7% of black kids can get 750 on the SAT (today's SAT — pre-1995 that was like a 700) then the fraction that has what it takes to become a math prof at a top-50 school is way lower.

    Replies: @Pericles, @Jack D, @res

    For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT…) the math SAT is ridiculously easy.

    MIT is good, don’t know about Harvard. It’s more about the hobnobbing, isn’t it? Princeton has awarded at least one PhD in math to a black woman, who wrote a highly amusing thesis. (Princeton is trash.)

  80. @Undocumented Shopper
    It's even less than 0.7% if you exclude Black immigrants.

    Anecdotal evidence: In the year I got my Math PhD, 1100 doctoral degrees in Math were awarded. Of those, two were awarded to Blacks. By accident, I knew both of them. One was an immigrant from Nigeria, the other was from Belize.

    Replies: @Triumph104, @Jack D

    I think this is true in general of that whole 750+ SAT group. Actual Moochelle type American slave descendants (other than light skinned mulattoes from way back – Louisiana Creoles, etc.) are almost completely missing from that group. Instead you have blacks with one white parent, Caribbean blacks, Nigerian Igbos, etc.

    • Replies: @Undocumented Shopper
    @Jack D

    Apparently, 27% of Black undergrads are immigrants from Africa or the Caribbean.
    There is tension between them and those who descend from Southern slaves.

    At Cornell, demands were made:
    “We demand that Cornell admissions come up with a plan to actively increase the presence of underrepresented black students on this campus. We define underrepresented black students as black Americans who have several generations (more than two) in this country.
 The black student population at Cornell disproportionately represents international or first-generation African or Caribbean students. While these students have a right to flourish at Cornell, there is a lack of investment in black students whose families were affected directly by the African Holocaust in America. Cornell must work to actively support students whose families have been impacted for generations by white supremacy and American fascism.”
    https://www.insidehighered.com/admissions/article/2017/10/09/cornell-students-revive-debate-whom-colleges-should-count-black

  81. @Triumph104
    @Undocumented Shopper

    Is Arlie Petters the guy from Belize? He identifies as being "two or more races", although he is always referred to as black or African-American.

    Former NFL player John Urschel is scheduled to earn his math PhD from MIT this spring. He was born in Winnipeg, Canada and has a black mother and white father, but you wouldn't know that from his media coverage.

    Replies: @Undocumented Shopper, @Buffalo Joe

    Yes, one of those two was Arlie Petters.
    The discrepancy may be due to the fact that in the United States people who have some black ancestry are referred to as “Black”. Perhaps Belize does this differently?

  82. @Jack D
    @Undocumented Shopper

    I think this is true in general of that whole 750+ SAT group. Actual Moochelle type American slave descendants (other than light skinned mulattoes from way back - Louisiana Creoles, etc.) are almost completely missing from that group. Instead you have blacks with one white parent, Caribbean blacks, Nigerian Igbos, etc.

    Replies: @Undocumented Shopper

    Apparently, 27% of Black undergrads are immigrants from Africa or the Caribbean.
    There is tension between them and those who descend from Southern slaves.

    At Cornell, demands were made:
    “We demand that Cornell admissions come up with a plan to actively increase the presence of underrepresented black students on this campus. We define underrepresented black students as black Americans who have several generations (more than two) in this country.
 The black student population at Cornell disproportionately represents international or first-generation African or Caribbean students. While these students have a right to flourish at Cornell, there is a lack of investment in black students whose families were affected directly by the African Holocaust in America. Cornell must work to actively support students whose families have been impacted for generations by white supremacy and American fascism.”
    https://www.insidehighered.com/admissions/article/2017/10/09/cornell-students-revive-debate-whom-colleges-should-count-black

  83. @MEH 0910
    NYT:

    NASA Renames Facility After Katherine Johnson of ‘Hidden Figures’ Fame
    By Elisha Brown
    Feb. 23, 2019

    NASA on Friday officially renamed a facility in West Virginia after Katherine Johnson, an African-American mathematician and centenarian whose barrier-breaking career was depicted in the film “Hidden Figures.”
     
    https://static01.nyt.com/images/2019/02/23/science/23xp-nasa/merlin_151082844_794a02bb-8998-48aa-9779-e15f13cc2f0b-jumbo.jpg

    NASA’s Katherine Johnson Independent Verification and Validation Facility in Fairmont, W.Va. The program housed at the facility monitors the software used to track high-profile NASA missions.
     

    Replies: @El Dato

    NASA has redesignated its Independent Verification and Validation (IV&V) Facility in Fairmont, West Virginia, as the Katherine Johnson Independent Verification and Validation Facility, in honor of the West Virginia native and NASA “hidden figure.”

    ….

    President Donald Trump signed into law in December an act of Congress calling for the redesignation. The facility’s program contributes to the safety and success of NASA’s highest-profile missions by assuring that mission software performs correctly. IV&V now is in the process of planning a rededication ceremony.

    Why the hell not. IV&V sounds a bit evil.

    Ms. Johnson, who turned 100 in August, “remains in awe and honored by” the accolades she has received, Joylette Hylick, one of Ms. Johnson’s daughters, said on Friday. Ms. Hylick said her mother “can’t imagine why people would want to honor her for just doing a good job.”

    NASA’s decision to name the facility for Ms. Johnson is not the first time she has been celebrated on the national stage. In 2015, President Barack Obama awarded her the Presidential Medal of Freedom, the nation’s highest civilian honor.

    Ms. Johnson was not always respected. Called “colored computers,” she and other black women who worked in NASA’s computing pool more than a half-century ago were separated from their white colleagues while they calculated trajectories for the Apollo missions and other programs.

    Colored computers — definitely not standard “office beige”!

  84. On a previous thread commenter utu gave the following numbers:

    Purdue Math Dept. Faculty Count

    33 China
    27 American (White)
    6 Germany
    3 American (Jewish)
    3 South Asia
    3 Romania
    3 S. America/Spain
    2 American (Black)
    2 Hungary
    2 Italy
    1 France
    1 Poland
    1 Bulgaria
    1 Armenia
    1 Iran
    1 Korea
    1 Japan

    Assuming those numbers are roughly correct, only about a third (31 out of 91) of Purdue math profs are white or black Americans. Probably the foreign percentage is higher among the younger profs.
    So, it’s pretty dumb to think that the “correct” percentage of blacks among math professors equals the black percentage of the American population.

    Also, Camilla Benbow has gotten data on what happens with kids who test at the 1-in-10,000 level of smart, versus kids at the 1-in-100 level of smart. The super-smart kids have much higher rates of getting STEM Ph.D.s and becoming successful researchers. The black percentage among the super-smart kids is almost certainly much lower than the black percentage among the pretty smart kids.

  85. @Buffalo Joe
    @Anon

    Anon, two thoughts, a lot of blacks have Doctorates in Education, does that count as a PhD.? And two, people doing the hiring really don't care if some one's parent or parents have PhDs, they are hiring the applicant.

    Replies: @Arclight

    “Education” is one of the easiest bachelor’s degrees to obtain, so I’d guess a PhD in that field is a pretty low bar as well – basically just requires the ability/desire to remain in school long enough. Given the number of blacks who are pastors of one variety or another, are there a bunch of doctorates of divinity in that 2.7 statistic? Seems surprising to me that the ratio is that low.

  86. @res
    @The Practical Conservative

    This article has more 2015 information (and links to underlying ETS documents):
    https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

    13% of 2015 SAT test takers were black, 47% white.

    The ETS percentile distributions document link returns "Unauthorized." It's available at
    https://web.archive.org/web/20170215214339/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-gender-ethnicity-2015.pdf

    iSteve wrote about that Brookings article in February 2017: https://www.unz.com/isteve/brookings-race-gaps-in-sat-math-scores-are-as-big-as-ever/
    I linked that PDF in a comment there. Interesting that it went missing later that year. Coincidence?

    Does anyone know the original source for this graphic?

    https://i0.wp.com/www.brookings.edu/wp-content/uploads/2017/01/ccf_20170201_reeves_2.png

    It is not in the PDF I linked above which has the same title as the stated source.

    For reference, here is the The Journal of Blacks in Higher Education article referenced (did I miss the link above? pretty sure someone linked this in the other thread): http://www.jbhe.com/features/49_college_admissions-test.html


    On the math SAT, only 0.7 percent of all black test takers scored at least 700 compared to 6.3 percent of all white test takers. Thus, whites were nine times as likely as blacks to score 700 or above on the math SAT. Overall, there were 45 times as many whites as blacks who scored 700 or above on the math SAT.

    If we raise the top-scoring threshold to students scoring 750 or above on both the math and verbal SAT — a level equal to the mean score of students entering the nation's most selective colleges such as Harvard, Princeton, and CalTech — we find that in the entire country 244 blacks scored 750 or above on the math SAT and 363 black students scored 750 or above on the verbal portion of the test. Nationwide, 33,841 students scored at least 750 on the math test and 30,479 scored at least 750 on the verbal SAT. Therefore, black students made up 0.7 percent of the test takers who scored 750 or above on the math test and 1.2 percent of all test takers who scored 750 or above on the verbal section.

     

    Their conclusion:

    In a race-neutral competition for the approximately 50,000 places for first-year students at the nation's 25 top-ranked universities, high-scoring blacks would be buried by a huge mountain of high-scoring non-black students. Today, under prevailing affirmative action admissions policies, there are about 3,000 black first-year students matriculating at these 25 high-ranking universities, about 6 percent of all first-year students at these institutions. But if these schools operated under a strict race-neutral admissions policy where SAT scores were the most important qualifying yardstick, these universities could fill their freshman classes almost exclusively with students who score at the very top of the SAT scoring scale. As shown previously, black students make up at best between 1 and 2 percent of these high-scoring groups.
     
    P.S. Some 2016 ETS data: https://reports.collegeboard.org/pdf/total-group-2016.pdf
    Some 2018 data: https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    Replies: @Gringo, @res, @Mr. XYZ, @res

    The Brookings numerical analysis has a strong smell of fraud about it, so I decided to take a closer look at various numbers they gave and check them for consistency. Here are some salient points from the Brookings article and associated ETS documents. I believe these are all using the 2015 ETS data.

    We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.

    Table 7 of https://secure-media.collegeboard.org/digitalServices/pdf/sat/total-group-2015.pdf has scoring distributions and demographics

    The Brookings graphic (Race gaps at the tails, SAT math scores) I linked above has estimated percentiles scoring above 750.

    I combined those sources into a spreadsheet (with formulas in original appearing as computed values in CSV). It appears after the MORE tag as a CSV if anyone wants to take a look. There are some issues because Brookings lumps three different Latino groups together and ignores the “Other” and “No Response” categories which I estimate account for 10% of Math scores over 750.

    Using the Table 7 numbers I calculated the expected number of scores over 750 assuming a normal distribution. The black estimate was 140 (!). I suspect the distribution is fat tailed so I find the JBHE 244 number at least plausible (but would like to see details of how they computed it). But both of those numbers are only loosely related to “at most 1000.” Also, the white estimate is low given the mean/SD. Odd.

    I suspect what Brookings did was make assumptions about the “Other” and “No Response” categories which favored the black estimates.

    In any case, I think my analysis shows their presentation was far enough from reality to be considered fraudulent. It would be interesting to see a better calculation using the full ETS data (which is quite possible). I wonder why that is not available?

    P.S. Here is another broken link from the Brookings article. Funny how the most informative links seem to be the ones that break.
    https://web.archive.org/web/20160310073125/https://cshe.berkeley.edu/sites/default/files/shared/publications/docs/ROPS.CSHE_.10.15.Geiser.RaceSAT.10.26.2015.pdf

    [MORE]

    Race,N,Math Mean,Math SD,SDs for 750,Est prop 750,Est N 750,Est percent 750,Brookings Est N,Ratio of Estimates,Brookings percent 750 as N,Est percent 750
    American Indian,”10,031″,482,107,2.50,0.0061,61,0.14%,,,,
    Asian,”211,238″,598,127,1.20,0.1157,24437,55.06%,29570,0.83,26628,
    Black,”219,018″,428,100,3.22,0.0006,140,0.32%,1000,0.14,888,
    Mexican,”130,026″,457,98,2.99,0.0014,181,0.41%,967,0.19,894,
    Puerto Rican,”30,192″,449,106,2.84,0.0023,68,0.15%,224,0.30,208,
    Other Hispanic,”162,655″,457,107,2.74,0.0031,502,1.13%,1209,0.42,1118,
    White,”800,236″,534,104,2.08,0.0189,15128,34.09%,”16,000″,0.95,14646,
    Other ,”65,063″,519,123,1.88,0.0302,1964,4.43%,,,,
    No Response,”70,062″,492,134,1.93,0.0271,1898,4.28%,,,,
    Total,”1,698,521″,511,120,1.99,0.0232,39412,88.81%,,,,
    ,,,,,,,,,,,
    Calculated Totals,”1,698,521″,510.9,108.2,,,44380,,48970,0.91,44380,

    • Replies: @Jack D
    @res

    Everyone keeps talking about black and white but the lede is getting buried:


    16,000 whites and 29,570 Asians scored above a 750,
     
    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it's not because all the Asian are cheating).

    In other words, close to 2/3 of all 750+ SAT scorers are Asian (despite Asians being less than 6% of the US population). Asians are the new Jews. So why isn't MIT 60% Asian instead of 30% Asian?

    Replies: @res, @res

    , @dux.ie
    @res

    Your calculations assumed that the distributions are perfect bell curve which might not be the case. The College Board's data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors. The results are close to that from Brookings, especially for large Pct75, e.g. """16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.""" compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.

    sum(N75)=53700

    Eth:Pct75 Nsub N75 Pct Brookings
    Bla: 1 219018 2190 4.08% 2%
    Nam: 1 10031 100 0.19%
    Asi: 14 211238 29573 55.07% 60%
    Mex: 1 130026 1300 2.42%
    Pue: 1 30192 302 0.56%
    Lat: 1 162655 1627 3.03% 5%
    Whi: 2 800236 16005 29.80% 33%
    Oth: 4 65063 2603 4.85%

    Replies: @res

  87. @International Jew
    @Alan Mercer


    Missing even 1 question can take 40 points off
     
    Mm, doubtful. Says here
    https://blog.prepscholar.com/how-is-the-sat-scored-scoring-charts
    you get 750 if you miss about five (out of the usual 58) math questions.

    Never mind becoming a math prof or even completing a PhD in math. For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT...) the math SAT is ridiculously easy. All of it. There's never a single question that's hard in any way. If you didn't get 800, you got 790 for some bit of carelessness. And I'm describing the state of affairs when I took it, which was before 1995, when it was made easier. You didn't even brag about your SAT (not that we were a modest bunch), you bragged about your exploits in the high school math league, if not in international competitions (where the easiest question is harder than the hardest question that's ever been seen on the SAT).

    To make a long story short, if only 0.7% of black kids can get 750 on the SAT (today's SAT — pre-1995 that was like a 700) then the fraction that has what it takes to become a math prof at a top-50 school is way lower.

    Replies: @Pericles, @Jack D, @res

    The exact # translation between raw score and scaled score differs with each version of the test because they “curve” each one to fit the normal distribution. As more and more Asians have gotten better and better at the test, in some versions even 1 question wrong is a 780, there being no way to get a 790. The raw to scale translation given in that blog post are illustrative and YMMV. Maybe there is some version of the test where 1 wrong is a 760 – I dunno but it’s not impossible.

    But otherwise I agree with you. The 25th percentile Math SAT at MIT is 770 and it would be even higher if they didn’t have AA for blacks and women. I would guess that virtually all of the Asian male applicants have 800s (or as you say, maybe one wrong thru some careless mistake). A 770+ Math score is just a gatekeeping device for a place like MIT and they have to separate the wheat from the chaff in that pool using other means. They admit around 1,500 of the 21,000 applications that they get. Maybe 1/3 or 1/4 of those applications are Hail Mary plays with sub- 770 Math SATs and they can trash most of those on the 1st reading (at least from white and Asian applicants) but that still leaves them with something like ten 770+ applicants for every person they admit (probably a significant chunk of all the 770+ SAT scorers apply to MIT as one of their choices) so they have to look at other things beyond just the SATs.

  88. @International Jew
    @Alan Mercer


    Missing even 1 question can take 40 points off
     
    Mm, doubtful. Says here
    https://blog.prepscholar.com/how-is-the-sat-scored-scoring-charts
    you get 750 if you miss about five (out of the usual 58) math questions.

    Never mind becoming a math prof or even completing a PhD in math. For anyone with a hope of successfully completing a math BA at a top American college (Harvard, Princeton, MIT...) the math SAT is ridiculously easy. All of it. There's never a single question that's hard in any way. If you didn't get 800, you got 790 for some bit of carelessness. And I'm describing the state of affairs when I took it, which was before 1995, when it was made easier. You didn't even brag about your SAT (not that we were a modest bunch), you bragged about your exploits in the high school math league, if not in international competitions (where the easiest question is harder than the hardest question that's ever been seen on the SAT).

    To make a long story short, if only 0.7% of black kids can get 750 on the SAT (today's SAT — pre-1995 that was like a 700) then the fraction that has what it takes to become a math prof at a top-50 school is way lower.

    Replies: @Pericles, @Jack D, @res

    if only 0.7% of black kids can get 750 on the SAT (today’s SAT — pre-1995 that was like a 700)

    Math didn’t change that much (e.g. old 750 = new 760). Verbal was the test that was really dumbed down (e.g. old 730 = new 800).

    Conversion chart: https://www.greenes.com/html/convert.htm

    P.S. I suspect the test scoring has been made even less rigorous since then through the various changes over the years since 1996, but good evidence to support that is scarce.

  89. @res
    @res

    The Brookings numerical analysis has a strong smell of fraud about it, so I decided to take a closer look at various numbers they gave and check them for consistency. Here are some salient points from the Brookings article and associated ETS documents. I believe these are all using the 2015 ETS data.


    We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.
     
    Table 7 of https://secure-media.collegeboard.org/digitalServices/pdf/sat/total-group-2015.pdf has scoring distributions and demographics

    The Brookings graphic (Race gaps at the tails, SAT math scores) I linked above has estimated percentiles scoring above 750.

    I combined those sources into a spreadsheet (with formulas in original appearing as computed values in CSV). It appears after the MORE tag as a CSV if anyone wants to take a look. There are some issues because Brookings lumps three different Latino groups together and ignores the "Other" and "No Response" categories which I estimate account for 10% of Math scores over 750.

    Using the Table 7 numbers I calculated the expected number of scores over 750 assuming a normal distribution. The black estimate was 140 (!). I suspect the distribution is fat tailed so I find the JBHE 244 number at least plausible (but would like to see details of how they computed it). But both of those numbers are only loosely related to "at most 1000." Also, the white estimate is low given the mean/SD. Odd.

    I suspect what Brookings did was make assumptions about the "Other" and "No Response" categories which favored the black estimates.

    In any case, I think my analysis shows their presentation was far enough from reality to be considered fraudulent. It would be interesting to see a better calculation using the full ETS data (which is quite possible). I wonder why that is not available?

    P.S. Here is another broken link from the Brookings article. Funny how the most informative links seem to be the ones that break.
    https://web.archive.org/web/20160310073125/https://cshe.berkeley.edu/sites/default/files/shared/publications/docs/ROPS.CSHE_.10.15.Geiser.RaceSAT.10.26.2015.pdf



    Race,N,Math Mean,Math SD,SDs for 750,Est prop 750,Est N 750,Est percent 750,Brookings Est N,Ratio of Estimates,Brookings percent 750 as N,Est percent 750
    American Indian,"10,031",482,107,2.50,0.0061,61,0.14%,,,,
    Asian,"211,238",598,127,1.20,0.1157,24437,55.06%,29570,0.83,26628,
    Black,"219,018",428,100,3.22,0.0006,140,0.32%,1000,0.14,888,
    Mexican,"130,026",457,98,2.99,0.0014,181,0.41%,967,0.19,894,
    Puerto Rican,"30,192",449,106,2.84,0.0023,68,0.15%,224,0.30,208,
    Other Hispanic,"162,655",457,107,2.74,0.0031,502,1.13%,1209,0.42,1118,
    White,"800,236",534,104,2.08,0.0189,15128,34.09%,"16,000",0.95,14646,
    Other ,"65,063",519,123,1.88,0.0302,1964,4.43%,,,,
    No Response,"70,062",492,134,1.93,0.0271,1898,4.28%,,,,
    Total,"1,698,521",511,120,1.99,0.0232,39412,88.81%,,,,
    ,,,,,,,,,,,
    Calculated Totals,"1,698,521",510.9,108.2,,,44380,,48970,0.91,44380,

    Replies: @Jack D, @dux.ie

    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).

    In other words, close to 2/3 of all 750+ SAT scorers are Asian (despite Asians being less than 6% of the US population). Asians are the new Jews. So why isn’t MIT 60% Asian instead of 30% Asian?

    • Replies: @res
    @Jack D


    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,
     

     
    I'm curious how accurate those numbers really are. My distributional analysis gives 24437 Asians and 15128 whites. But I suspect most of the Other and No Response groups contribution of about 4000 is Asian (or part) so maybe they are close.

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).
     
    Agreed. You and I both know that is how the different distributions work further into the tail.

    So why isn’t MIT 60% Asian instead of 30% Asian?
     
    Perhaps because things other than SAT math matter? (for real) Caltech is a good comparison at 40% Asian (which is probably a more realistic reflection of ability, but still skewed low): http://www.registrar.caltech.edu/academics/enrollment

    But I think an even better explanation is being caught between the rock of black (etc.) underrepresentation and the hard place of Asian ability. Of course as various analyses have shown, white gentiles are the ones really getting crushed between both of those plus Jewish ability.

    Replies: @Jack D

    , @res
    @Jack D


    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,
     

     
    I'm curious how accurate those numbers really are. My distributional analysis gives 24437 Asians and 15128 whites. But I suspect most of the Other and No Response groups contribution of about 4000 is Asian (or part) so maybe they are close.

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).
     
    Agreed. You and I both know that is how the different distributions work further into the tail.

    So why isn’t MIT 60% Asian instead of 30% Asian?
     
    Perhaps because things other than SAT math matter? (for real) Caltech is a good comparison at 40% Asian (which is probably a more realistic reflection of ability, but still skewed low): http://www.registrar.caltech.edu/academics/enrollment

    But I think an even better explanation is being caught between the rock of black (etc.) underrepresentation and the hard place of Asian ability. Of course as various analyses have shown, white gentiles are the ones really getting crushed between both of those plus Jewish ability.
  90. @Nathan
    Hey, hey, hey! No fair crunching the numbers on how many people actually achieve top scores on college admissions tests. Pretty soon you'll notice that there aren't enough top scores to back up the admissions claims of elite colleges. Can't have that.

    Replies: @res

    Pretty soon you’ll notice that there aren’t enough top scores to back up the admissions claims of elite colleges.

    Sounds plausible. Do you have an analysis to back that up?

    • Replies: @Nathan
    @res

    Yes, but I don't want to put the whole excel file up. It's enough to check out Harvard, and extrapolate from there:

    https://college.harvard.edu/admissions/admissions-statistics

    Harvard admitted 2,024 undergraduate students, and claims that their 75th percentile on the SAT is a perfect 800 in the reading and math sections. So 506 students in the admitted class have either a perfect reading or math score. So far, so good. The problem is when you add up all the other colleges that claim similar admissions stats. Yale, Princeton, most of the Ivy league. MIT, Caltech... sure. Washington University in St. Louis?? No freaking way.

    Harvard is probably not lying. Washington University in St. Louis almost certainly is. If you aced the SAT, would you want to go to some no-name recognition school in America's worst (or 2nd worst) city?

    Replies: @dux.ie, @Jack D

  91. @Jack D
    @res

    Everyone keeps talking about black and white but the lede is getting buried:


    16,000 whites and 29,570 Asians scored above a 750,
     
    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it's not because all the Asian are cheating).

    In other words, close to 2/3 of all 750+ SAT scorers are Asian (despite Asians being less than 6% of the US population). Asians are the new Jews. So why isn't MIT 60% Asian instead of 30% Asian?

    Replies: @res, @res

    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,

    I’m curious how accurate those numbers really are. My distributional analysis gives 24437 Asians and 15128 whites. But I suspect most of the Other and No Response groups contribution of about 4000 is Asian (or part) so maybe they are close.

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).

    Agreed. You and I both know that is how the different distributions work further into the tail.

    So why isn’t MIT 60% Asian instead of 30% Asian?

    Perhaps because things other than SAT math matter? (for real) Caltech is a good comparison at 40% Asian (which is probably a more realistic reflection of ability, but still skewed low): http://www.registrar.caltech.edu/academics/enrollment

    But I think an even better explanation is being caught between the rock of black (etc.) underrepresentation and the hard place of Asian ability. Of course as various analyses have shown, white gentiles are the ones really getting crushed between both of those plus Jewish ability.

    • Replies: @Jack D
    @res


    white gentiles are the ones really getting crushed
     
    It depends on whether you go by share of the general population or share of top SAT scorers. Whites are now 36% of MIT admits, which makes them underrepresented vs. their share of the US population. But if you are looking at 750+ scorers (who are really the relevant pool for MIT), they are over-represented. Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones "getting crushed" is subjective and not clear cut.

    I agree with you that Cal Tech (or Berkeley - 43% Asian) is probably the more realistic benchmark but then MIT is still 10+ percentage points low. Most of this comes from taking blacks and Latinos, of which there are far more than should be there.

    Replies: @res

  92. @Jack D
    @res

    Everyone keeps talking about black and white but the lede is getting buried:


    16,000 whites and 29,570 Asians scored above a 750,
     
    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it's not because all the Asian are cheating).

    In other words, close to 2/3 of all 750+ SAT scorers are Asian (despite Asians being less than 6% of the US population). Asians are the new Jews. So why isn't MIT 60% Asian instead of 30% Asian?

    Replies: @res, @res

    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,

    I’m curious how accurate those numbers really are. My distributional analysis gives 24437 Asians and 15128 whites. But I suspect most of the Other and No Response groups contribution of about 4000 is Asian (or part) so maybe they are close.

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).

    Agreed. You and I both know that is how the different distributions work further into the tail.

    So why isn’t MIT 60% Asian instead of 30% Asian?

    Perhaps because things other than SAT math matter? (for real) Caltech is a good comparison at 40% Asian (which is probably a more realistic reflection of ability, but still skewed low): http://www.registrar.caltech.edu/academics/enrollment

    But I think an even better explanation is being caught between the rock of black (etc.) underrepresentation and the hard place of Asian ability. Of course as various analyses have shown, white gentiles are the ones really getting crushed between both of those plus Jewish ability.

  93. @res
    @Jack D


    Everyone keeps talking about black and white but the lede is getting buried:

    16,000 whites and 29,570 Asians scored above a 750,
     

     
    I'm curious how accurate those numbers really are. My distributional analysis gives 24437 Asians and 15128 whites. But I suspect most of the Other and No Response groups contribution of about 4000 is Asian (or part) so maybe they are close.

    (And I would bet that if you filter up to 770 or 800, the Asian/white imbalance is even greater, and no it’s not because all the Asian are cheating).
     
    Agreed. You and I both know that is how the different distributions work further into the tail.

    So why isn’t MIT 60% Asian instead of 30% Asian?
     
    Perhaps because things other than SAT math matter? (for real) Caltech is a good comparison at 40% Asian (which is probably a more realistic reflection of ability, but still skewed low): http://www.registrar.caltech.edu/academics/enrollment

    But I think an even better explanation is being caught between the rock of black (etc.) underrepresentation and the hard place of Asian ability. Of course as various analyses have shown, white gentiles are the ones really getting crushed between both of those plus Jewish ability.

    Replies: @Jack D

    white gentiles are the ones really getting crushed

    It depends on whether you go by share of the general population or share of top SAT scorers. Whites are now 36% of MIT admits, which makes them underrepresented vs. their share of the US population. But if you are looking at 750+ scorers (who are really the relevant pool for MIT), they are over-represented. Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones “getting crushed” is subjective and not clear cut.

    I agree with you that Cal Tech (or Berkeley – 43% Asian) is probably the more realistic benchmark but then MIT is still 10+ percentage points low. Most of this comes from taking blacks and Latinos, of which there are far more than should be there.

    • Replies: @res
    @Jack D

    Which source(s) are you using for the MIT numbers? For the Class of 2022 I see https://mitadmissions.org/apply/process/profile/
    which says 49% white (and 13% Jewish does seem plausible, if perhaps low depending on definition). It also says 37% Asian American (so not quite as far off as you said), 18% Hispanic, and 10% black.

    It's hard to be precise about all of this given the lack of data and general difficulty of coming up with a common definition of Jewish (and other categories) which can be applied against what data is available. Ron Unz's articles on this are the best reference I know and make clear there is an issue. Even if the exact magnitude is arguable.


    Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones “getting crushed” is subjective and not clear cut.
     
    Fair enough. I did overstate that a bit.

    P.S. For anyone interested in this, there is a fair amount of discussion of the relative M/V/S profiles of different groups (relevant to broadening this discussion beyond math) in the comments of https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/
    For example, this from Ron Unz: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2738509
    and this gold box comment from JLK: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2740587

    Replies: @Jack D

  94. @Jack D
    @res


    white gentiles are the ones really getting crushed
     
    It depends on whether you go by share of the general population or share of top SAT scorers. Whites are now 36% of MIT admits, which makes them underrepresented vs. their share of the US population. But if you are looking at 750+ scorers (who are really the relevant pool for MIT), they are over-represented. Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones "getting crushed" is subjective and not clear cut.

    I agree with you that Cal Tech (or Berkeley - 43% Asian) is probably the more realistic benchmark but then MIT is still 10+ percentage points low. Most of this comes from taking blacks and Latinos, of which there are far more than should be there.

    Replies: @res

    Which source(s) are you using for the MIT numbers? For the Class of 2022 I see https://mitadmissions.org/apply/process/profile/
    which says 49% white (and 13% Jewish does seem plausible, if perhaps low depending on definition). It also says 37% Asian American (so not quite as far off as you said), 18% Hispanic, and 10% black.

    It’s hard to be precise about all of this given the lack of data and general difficulty of coming up with a common definition of Jewish (and other categories) which can be applied against what data is available. Ron Unz’s articles on this are the best reference I know and make clear there is an issue. Even if the exact magnitude is arguable.

    Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones “getting crushed” is subjective and not clear cut.

    Fair enough. I did overstate that a bit.

    P.S. For anyone interested in this, there is a fair amount of discussion of the relative M/V/S profiles of different groups (relevant to broadening this discussion beyond math) in the comments of https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/
    For example, this from Ron Unz: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2738509
    and this gold box comment from JLK: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2740587

    • Replies: @Jack D
    @res

    I don't remember now what link I looked at and don't feel like finding them again but the link you gave looks good. 13% Jewish for MIT is very high - MIT is not a particularly Jewish place - although Jews are not unknown there, they are not particularly common either, at least in relation to some of the other Ivies.

    Hillel (known for overstating Jewish %'s) claims 7% of undergrads:

    https://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology

    (They claim 11% for Harvard and 17% for Penn).


    Jews nowadays are not that interested in STEM subjects - they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for - American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.

    Replies: @res, @Hibernian

  95. Facts inconvenient to The Narrative will always be ignored.

  96. @res
    @Jack D

    Which source(s) are you using for the MIT numbers? For the Class of 2022 I see https://mitadmissions.org/apply/process/profile/
    which says 49% white (and 13% Jewish does seem plausible, if perhaps low depending on definition). It also says 37% Asian American (so not quite as far off as you said), 18% Hispanic, and 10% black.

    It's hard to be precise about all of this given the lack of data and general difficulty of coming up with a common definition of Jewish (and other categories) which can be applied against what data is available. Ron Unz's articles on this are the best reference I know and make clear there is an issue. Even if the exact magnitude is arguable.


    Now SAT score is not by itself the be-all and end-all, but whether they (or the Asians) are the ones “getting crushed” is subjective and not clear cut.
     
    Fair enough. I did overstate that a bit.

    P.S. For anyone interested in this, there is a fair amount of discussion of the relative M/V/S profiles of different groups (relevant to broadening this discussion beyond math) in the comments of https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/
    For example, this from Ron Unz: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2738509
    and this gold box comment from JLK: https://www.unz.com/jthompson/swanning-about-fooled-by-algebra/#comment-2740587

    Replies: @Jack D

    I don’t remember now what link I looked at and don’t feel like finding them again but the link you gave looks good. 13% Jewish for MIT is very high – MIT is not a particularly Jewish place – although Jews are not unknown there, they are not particularly common either, at least in relation to some of the other Ivies.

    Hillel (known for overstating Jewish %’s) claims 7% of undergrads:

    https://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology

    (They claim 11% for Harvard and 17% for Penn).

    Jews nowadays are not that interested in STEM subjects – they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for – American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.

    • Replies: @res
    @Jack D


    MIT is not a particularly Jewish place – although Jews are not unknown there, they are not particularly common either
     
    Perhaps you are right (depending on definition of "particularly Jewish place", 3-4x population percentage is significant, but probably in ballpark of ability advantage). My knowledge of MIT is old. Asians have certainly increased at MIT over the years (eating into Jewish %). Let's look some more. ~5 years ago Hillel said 9% Jewish: https://web.archive.org/web/20130828044534/http://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology
    I am having trouble finding older data. Here is a 2007 compilation of Hillel data (again, 9%): https://talk.collegeconfidential.com/college-search-selection/310901-jewish-enrollment-a-graph.html

    Probably some perceptual bias due to Jewish overrepresentation in assorted campus subgroups as well.

    at least in relation to some of the other Ivies.
     
    Definitely agreed about that point.

    Jews nowadays are not that interested in STEM subjects – they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for – American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.
     
    I knew this, but thinking about it some more emphasizes just how much you are right with that whole paragraph. It is a very sensible response given the relative M/V/S profiles of Asians and Jews and the upsurge of Asians in the US. Not to mention the financial incentives you mention.

    That's a bit sad. As much as high V low M people can annoy me, losing high V AND high M Jews from STEM could be a problem.
    , @Hibernian
    @Jack D

    Finance has a significant (albeit relatively low powered) STEM component.

    Replies: @Jack D

  97. @Anon
    @Gringo

    I guess I didn't make my main point explicit enough: The sociologist's reasoning ignores that IQ and EA (parental educational attainment, e.g. the Ph.D. parent), as well as SES (socioeconomic status, e.g. salary and wealth), and many other things like lifespan and health, are all correlated, and you therefore end up with confounded and erroneous results if you assume that they are independent variables, as he did with parental EA and, by implication, child SES (which is a score that comprises income, EA itself, and job prestige). Another way to put it is that Ph.D.s are heritable. So of course blacks have fewer Ph.D.s if black parents have fewer Ph.D.s.

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the "books in the home" or "parents reading to the kids" theories were blown out of the water by twin and adoption studies, and it's a similar idea to "Ph.D.s in the home." Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment. They figure out how to get books. (One objection that has been raised is that twin and adoption studies have not included homes with butt-dumb parents ... they tend not to adopt, so you have to extrapolate down, not good statistically, but common sense tells you that the trend continues.)

    Replies: @Steve Sailer, @Gringo, @Gringo

    I wasn’t disagreeing with you. I was simply adding my reaction to the “2.7 times” statement. I find it especially silly that Amy Harmon, who doesn’t have a Ph.D. and probably never tried to get one, is kvetching about this. I doubt she has any idea at all what hoop jumping it takes to get a doctorate.

    I am reminded of my sojourn in the teaching profession. I took an Education course where a female student was kvetching about the low percentage of female students in some engineering department. My sister worked as a Chem E for 25 years, which informs me that gender isn’t the main issue in becoming an engineer. The main issue is that bright students need to spend 60 hours a week in class and in studying to get an engineering degree. That female Education student had neither the intellect nor the work ethic to make it as an engineer. As such, her complaint was absurd.

    If students want to put up with the hoop jumping to get a doctorate, more power to them. I see no need to complain about those who do not so choose. It is their choice not to pursue a doctorate.

  98. @Anon
    @Gringo

    I guess I didn't make my main point explicit enough: The sociologist's reasoning ignores that IQ and EA (parental educational attainment, e.g. the Ph.D. parent), as well as SES (socioeconomic status, e.g. salary and wealth), and many other things like lifespan and health, are all correlated, and you therefore end up with confounded and erroneous results if you assume that they are independent variables, as he did with parental EA and, by implication, child SES (which is a score that comprises income, EA itself, and job prestige). Another way to put it is that Ph.D.s are heritable. So of course blacks have fewer Ph.D.s if black parents have fewer Ph.D.s.

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the "books in the home" or "parents reading to the kids" theories were blown out of the water by twin and adoption studies, and it's a similar idea to "Ph.D.s in the home." Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment. They figure out how to get books. (One objection that has been raised is that twin and adoption studies have not included homes with butt-dumb parents ... they tend not to adopt, so you have to extrapolate down, not good statistically, but common sense tells you that the trend continues.)

    Replies: @Steve Sailer, @Gringo, @Gringo

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the “books in the home” or “parents reading to the kids” theories were blown out of the water by twin and adoption studies, and it’s a similar idea to “Ph.D.s in the home.” Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment.

    It also works the other way : not as bright kids adopted into the homes of bright parents. What I personally know about adoption is anecdote, not a research finding, but I will relate it nonetheless. I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.

    • Replies: @Anon
    @Gringo


    I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.
     
    Statistically the correlation between parental and child intelligence is indisputable and very strong, but there can be quite a difference in the intelligence between siblings (and this is true in my own family). Mitchell in Innate goes into the details. In childhood adopted siblings' IQ correlate at 0.25 and biological siblings at 0.60. But the adopted correlation disappears in adulthood (and biological shrinks to 0.5 or less). Identical twins are higher, maybe 0.8.
    , @Jack D
    @Gringo

    I have mentioned before that my wife (who works in the educational field) has observed this same situation often - the adopted child of average or below average intelligence (which is most of them - nowadays high IQ people tend not to put their kids up for adoption) is adopted into a high achievement wealthy family where all the children are expected to attend a private university, preferably one of Ivy League calibre or at least in the top 50. This leads to a lot of stress as the kid is really not set up to operate at this level. If he had stayed with his birth family they would have been fine with the kid becoming a mechanic or a beautician but in his adoptive families circles this would be seen as a shocking failure.

    Vice versa (which used to happen more in the past when high IQ college girls were not supposed to be single mothers and abortion was not available) such as Steve Jobs tends to work out better as there are enough resources in our society for bright and ambitious blue collar kids to find their way without a lot of practical support from their families (as long as they get emotional support). Also that being bright means that you don't need a lot of practical support - you figure this stuff out on your own. In the "Three Identical Strangers" movie, after the brothers were reunited, they spent the most time at the home of the "blue collar" adoptee, whose parents (especially the father) was the most emotionally supportive. The "blue collar" brother was also the one who was the most mentally intact of the three. For kids of high intelligence, a loving home is perhaps more important than having really smart parents.

  99. @Jack D
    @res

    I don't remember now what link I looked at and don't feel like finding them again but the link you gave looks good. 13% Jewish for MIT is very high - MIT is not a particularly Jewish place - although Jews are not unknown there, they are not particularly common either, at least in relation to some of the other Ivies.

    Hillel (known for overstating Jewish %'s) claims 7% of undergrads:

    https://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology

    (They claim 11% for Harvard and 17% for Penn).


    Jews nowadays are not that interested in STEM subjects - they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for - American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.

    Replies: @res, @Hibernian

    MIT is not a particularly Jewish place – although Jews are not unknown there, they are not particularly common either

    Perhaps you are right (depending on definition of “particularly Jewish place”, 3-4x population percentage is significant, but probably in ballpark of ability advantage). My knowledge of MIT is old. Asians have certainly increased at MIT over the years (eating into Jewish %). Let’s look some more. ~5 years ago Hillel said 9% Jewish: https://web.archive.org/web/20130828044534/http://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology
    I am having trouble finding older data. Here is a 2007 compilation of Hillel data (again, 9%): https://talk.collegeconfidential.com/college-search-selection/310901-jewish-enrollment-a-graph.html

    Probably some perceptual bias due to Jewish overrepresentation in assorted campus subgroups as well.

    at least in relation to some of the other Ivies.

    Definitely agreed about that point.

    Jews nowadays are not that interested in STEM subjects – they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for – American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.

    I knew this, but thinking about it some more emphasizes just how much you are right with that whole paragraph. It is a very sensible response given the relative M/V/S profiles of Asians and Jews and the upsurge of Asians in the US. Not to mention the financial incentives you mention.

    That’s a bit sad. As much as high V low M people can annoy me, losing high V AND high M Jews from STEM could be a problem.

  100. @res
    @Nathan


    Pretty soon you’ll notice that there aren’t enough top scores to back up the admissions claims of elite colleges.
     
    Sounds plausible. Do you have an analysis to back that up?

    Replies: @Nathan

    Yes, but I don’t want to put the whole excel file up. It’s enough to check out Harvard, and extrapolate from there:

    https://college.harvard.edu/admissions/admissions-statistics

    Harvard admitted 2,024 undergraduate students, and claims that their 75th percentile on the SAT is a perfect 800 in the reading and math sections. So 506 students in the admitted class have either a perfect reading or math score. So far, so good. The problem is when you add up all the other colleges that claim similar admissions stats. Yale, Princeton, most of the Ivy league. MIT, Caltech… sure. Washington University in St. Louis?? No freaking way.

    Harvard is probably not lying. Washington University in St. Louis almost certainly is. If you aced the SAT, would you want to go to some no-name recognition school in America’s worst (or 2nd worst) city?

    • Replies: @dux.ie
    @Nathan

    > The problem is when you add up all the other colleges that claim similar admissions stats.

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai's dataset with only SAT75. IQlike75 is separately calculated. MIT no. Princeton no. WUST no.

    Rank|SAT75|SATV75|SATM75|IQlike75|Inst
    1|1600|800|800|140.4|California Institute of Technology
    1|1600|800|800|140.4|Yale University
    1|1600|800|800|140.4|University of Chicago
    1|1600|800|800|140.4|Harvard University

    5|1590|790|800|139.7|Massachusetts Institute of Technology
    5|1590|790|800|139.7|Princeton University
    5|1590|790|800|139.7|Columbia University in the City of New York
    5|1590|790|800|139.7|Vanderbilt University
    9|1580|780|800|138.9|Harvey Mudd College
    9|1580|780|800|138.9|Stanford University
    9|1580|780|800|138.9|Franklin W Olin College of Engineering
    12|1570|770|800|138.2|Johns Hopkins University
    12|1570|770|800|138.2|Washington University in St Louis
    12|1570|770|800|138.2|Duke University
    12|1570|770|800|138.2|University of Pennsylvania
    12|1570|780|790|138.2|Brown University
    12|1570|770|800|138.2|Rice University
    18|1560|760|800|137.5|Northwestern University
    18|1560|780|780|137.5|Dartmouth College
    19|1555|775|780|137.1|Amherst College
    20|1550|750|800|136.8|Carnegie Mellon University
    21|1540|750|790|136.0|University of California-Berkeley
    51|1470|710|760|131.0|University of California-Los Angeles

    > If you aced the SAT, would you want to go to some no-name recognition school

    Top universities rejected more than accepted students with perfect scores (especially from some ethnic groups) to make room for the subpar applicants. There are self reported cases of those with perfect scores of 1600 in limbo wrt 7 tier1 universities and were on the waiting lists of another (he/she will be smart to start applying to some tier2 immediately), or with near perfect SAT scores on being immediately rejected wrt 7 tier1 universities (just below the 75th percentile for Berkeley) and settled for a tier2, e.g. from CollegeData.com dataset,

    ID|State|Class|Status|SAT|Inst
    AAA|MI|2022|Applied|1600|Brown
    AAA|MI|2022|Applied|1600|Cornell
    AAA|MI|2022|Applied|1600|Duke
    AAA|MI|2022|Applied|1600|Princeton
    AAA|MI|2022|Applied|1600|Stanford
    AAA|MI|2022|Applied|1600|UChicago
    AAA|MI|2022|Applied|1600|Yale
    AAA|MI|2022|Wait-Listed|1600|Caltech
    ---
    BBB|CA|2022|Denied|1520|Berkeley
    BBB|CA|2022|Denied|1520|Cornell
    BBB|CA|2022|Denied|1520|Duke
    BBB|CA|2022|Denied|1520|Harvard
    BBB|CA|2022|Denied|1520|Princeton
    BBB|CA|2022|Denied|1520|Stanford
    BBB|CA|2022|Denied|1520|UChicago
    BBB|CA|2022|Will Attend|1520|UCLA

    Replies: @Nathan, @res, @Calvin Hobbes

    , @Jack D
    @Nathan

    There are plenty to go around. Something like 15,000 kids each year (1%) score 800 on math (of which perhap 2/3 are Asian and maybe a couple of hundred are black(ish). This is enough to supply 300 to each of the top 50 schools.

    Replies: @Nathan

  101. @Reg Cæsar
    Amir Aczel found the earliest known representation of zero in a Cambodian jungle, and Terry Jones located the earliest known one of one in a temple in India, though he was locked out.


    FINDING ZERO
    A Mathematician's Odyssey to Uncover the Origins of Numbers


    The Story of 1


    The earliest known something must be in Africa. How about the earliest known emigrant? "There goes the neighborhood! I'm crossing the isthmus, where the grass will be greener."

    Replies: @Logan, @El Dato, @The Alarmist, @Neil Templeton

    Zero is the one great invention of the left, as it was developed to describe the intersection of leftist ideology with reality, i.e. the empty set. It also represents the lower limit of value to which the social system converges as those who believe in conserving value are liquidated.

  102. @Triumph104
    @Undocumented Shopper

    Is Arlie Petters the guy from Belize? He identifies as being "two or more races", although he is always referred to as black or African-American.

    Former NFL player John Urschel is scheduled to earn his math PhD from MIT this spring. He was born in Winnipeg, Canada and has a black mother and white father, but you wouldn't know that from his media coverage.

    Replies: @Undocumented Shopper, @Buffalo Joe

    Tri, John Urschel graduated from the local Jesuit Prep school, his father is a surgeon and his mother is a nurse. Young Urschel quit the NFL after three productive years as a lineman. He was quoted as saying he didn’t want to hurt his brain. His mother used to tell him after every HS, college and pro game…..”You don’t need to do this.” Play football that is.

  103. Chances of this fact causing Amy Harmon to reconsider her position: 0.7%.

    • Replies: @Spangel
    @Malcolm X-Lax

    I theory, refuting Amy Harmon’s position doesn’t require approaching the topic of whether or not the race iq gap is genetic. It merely requires strong proof that the gap exists. There is tremendous documentation of the existence of the black white iq gap on average and how iq follows a bell curve distribution. We also have year over year documentation of average sat scores and the ratio of math scores over 750, which follows a bell curve distribution for whites and blacks. We also see the percentage of sat test takers who are black is 14%, in proportion with the overall population. This indicates at least that there isn’t anything to be gained by encouraging more blacks to take the sat.

    Perhaps Amy Harmon thinks whites benefit from sat coaching. But note that the black sat takers who score over 700 math is only around 1100. So assuming we send all of those to tutors and expensive sat prep, even if every 700 to 750 math sat black were to somehow bump up into the 750+ range, blacks would still be around 3% of that level. And that’s an impossible scenario- nothing known could take every 700 math aptitude black and make them a 750. In essence, there is no foreseeable way to increase the percentage of blacks with high enough math aptitude to be phds in math such that black mathematicians would be anything but severely underrepresented. As only .7% of 750+ sat math takers are black at this time, and as that score is still realistically too low to become a math PhD most of the time, it would seem there is no way to increase the percentage of competent black math PhD candidates or professors in the pipeline.

  104. It’s interesting that all of her anecdotes of “educated racism” involve only white racists. What? No anecdotes involving Indian or Chinese racists mathematicians? Of course, we no how woke Asians in STEM fields are so I guess this comes as no surprise.

  105. @Malcolm X-Lax
    Chances of this fact causing Amy Harmon to reconsider her position: 0.7%.

    Replies: @Spangel

    I theory, refuting Amy Harmon’s position doesn’t require approaching the topic of whether or not the race iq gap is genetic. It merely requires strong proof that the gap exists. There is tremendous documentation of the existence of the black white iq gap on average and how iq follows a bell curve distribution. We also have year over year documentation of average sat scores and the ratio of math scores over 750, which follows a bell curve distribution for whites and blacks. We also see the percentage of sat test takers who are black is 14%, in proportion with the overall population. This indicates at least that there isn’t anything to be gained by encouraging more blacks to take the sat.

    Perhaps Amy Harmon thinks whites benefit from sat coaching. But note that the black sat takers who score over 700 math is only around 1100. So assuming we send all of those to tutors and expensive sat prep, even if every 700 to 750 math sat black were to somehow bump up into the 750+ range, blacks would still be around 3% of that level. And that’s an impossible scenario- nothing known could take every 700 math aptitude black and make them a 750. In essence, there is no foreseeable way to increase the percentage of blacks with high enough math aptitude to be phds in math such that black mathematicians would be anything but severely underrepresented. As only .7% of 750+ sat math takers are black at this time, and as that score is still realistically too low to become a math PhD most of the time, it would seem there is no way to increase the percentage of competent black math PhD candidates or professors in the pipeline.

  106. If affirmative action due to slavery guilt were the end of it, it would not be enough to destroy the country. We could handle that much loss of proficiency and decency, and furnish that much custodial support. But it has been deemed insufficient redemption and reparation; to atone, it was and is also deemed necessary to open the borders, and that put us over the tipping point. By the way, I don’t hear the author demanding proportional employment in the airline cockpit, plumbers’ union, and such other things she uses.

    • Replies: @ic1000
    @SafeNow

    > I don’t hear Amy Harmon demanding proportional employment in the airline cockpit

    marginally competent pilots flying the planes she and her daughter ride in

    > plumbers’ union

    more or less proficient plumbers doing the pipes in her place out in the Hamptons

    > and such other things she uses.

    bottom-of-their-class critical-care physicians managing the complex cases when it's life or death for her family members

    All sorts of compelling, heart-wrenching special pleading as to why the circumstances faced by Amy and her loved ones make that sort of reasoning inapplicable. Nomenklatura, not apparatchik.

    Replies: @Reg Cæsar

  107. @Reg Cæsar

    A lot of military officers are the children of a military officer...
     
    As are a lot of, well, anything-but-military types. Eg, Jim Morrison and Kris Kristofferson.

    You can go all the way, or all the other way. Or part way-- more anodyne military brat acts were America and Katrina and the Waves, both of whom came out of USAF bases in England.

    Replies: @William Badwhite

    Kristofferson also served in the Army as an officer. He was a helicopter pilot.

    • Replies: @Reg Cæsar
    @William Badwhite

    He was also born before John McCain.

    , @Hibernian
    @William Badwhite

    Additionally he taught English at West Point for a while.

  108. We’ve already reached the point of no return. A Ph.D (and for that matter, even a college degree) meant, in my parents’ time, that a person possessed higher-than-average intelligence and a strong desire to learn for the joy of it. In our Current Year, there are so-called academic programs which exist solely to allow blacks to get a degree. The programs themselves have no bearing, influence, or connection to or on reality. And oops, it looks like I’m late for my “African-American LBTGQW Foundations in America” Ph.D seminar.

  109. @SafeNow
    If affirmative action due to slavery guilt were the end of it, it would not be enough to destroy the country. We could handle that much loss of proficiency and decency, and furnish that much custodial support. But it has been deemed insufficient redemption and reparation; to atone, it was and is also deemed necessary to open the borders, and that put us over the tipping point. By the way, I don’t hear the author demanding proportional employment in the airline cockpit, plumbers’ union, and such other things she uses.

    Replies: @ic1000

    > I don’t hear Amy Harmon demanding proportional employment in the airline cockpit

    marginally competent pilots flying the planes she and her daughter ride in

    > plumbers’ union

    more or less proficient plumbers doing the pipes in her place out in the Hamptons

    > and such other things she uses.

    bottom-of-their-class critical-care physicians managing the complex cases when it’s life or death for her family members

    All sorts of compelling, heart-wrenching special pleading as to why the circumstances faced by Amy and her loved ones make that sort of reasoning inapplicable. Nomenklatura, not apparatchik.

    • Replies: @Reg Cæsar
    @ic1000

    Remember that actress who was murdered by the Peruvian illegal she and her husband hired to fix their shower? Adrienne Somethingorother, age 40.

    Gee, had they only been willing to pay an American...

  110. @Father O'Hara
    What about African Math? We need to teach THAT!

    Replies: @Anon7, @rufus, @Bruce County

    Yes I love that math… They use it mostly for getting the temperature just right in the cauldron for removing human flesh from bone. Later on it helped in the design and building of circular mud huts.

  111. @Nathan
    @res

    Yes, but I don't want to put the whole excel file up. It's enough to check out Harvard, and extrapolate from there:

    https://college.harvard.edu/admissions/admissions-statistics

    Harvard admitted 2,024 undergraduate students, and claims that their 75th percentile on the SAT is a perfect 800 in the reading and math sections. So 506 students in the admitted class have either a perfect reading or math score. So far, so good. The problem is when you add up all the other colleges that claim similar admissions stats. Yale, Princeton, most of the Ivy league. MIT, Caltech... sure. Washington University in St. Louis?? No freaking way.

    Harvard is probably not lying. Washington University in St. Louis almost certainly is. If you aced the SAT, would you want to go to some no-name recognition school in America's worst (or 2nd worst) city?

    Replies: @dux.ie, @Jack D

    > The problem is when you add up all the other colleges that claim similar admissions stats.

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai’s dataset with only SAT75. IQlike75 is separately calculated. MIT no. Princeton no. WUST no.

    Rank|SAT75|SATV75|SATM75|IQlike75|Inst
    1|1600|800|800|140.4|California Institute of Technology
    1|1600|800|800|140.4|Yale University
    1|1600|800|800|140.4|University of Chicago
    1|1600|800|800|140.4|Harvard University

    5|1590|790|800|139.7|Massachusetts Institute of Technology
    5|1590|790|800|139.7|Princeton University
    5|1590|790|800|139.7|Columbia University in the City of New York
    5|1590|790|800|139.7|Vanderbilt University
    9|1580|780|800|138.9|Harvey Mudd College
    9|1580|780|800|138.9|Stanford University
    9|1580|780|800|138.9|Franklin W Olin College of Engineering
    12|1570|770|800|138.2|Johns Hopkins University
    12|1570|770|800|138.2|Washington University in St Louis
    12|1570|770|800|138.2|Duke University
    12|1570|770|800|138.2|University of Pennsylvania
    12|1570|780|790|138.2|Brown University
    12|1570|770|800|138.2|Rice University
    18|1560|760|800|137.5|Northwestern University
    18|1560|780|780|137.5|Dartmouth College
    19|1555|775|780|137.1|Amherst College
    20|1550|750|800|136.8|Carnegie Mellon University
    21|1540|750|790|136.0|University of California-Berkeley
    51|1470|710|760|131.0|University of California-Los Angeles

    > If you aced the SAT, would you want to go to some no-name recognition school

    Top universities rejected more than accepted students with perfect scores (especially from some ethnic groups) to make room for the subpar applicants. There are self reported cases of those with perfect scores of 1600 in limbo wrt 7 tier1 universities and were on the waiting lists of another (he/she will be smart to start applying to some tier2 immediately), or with near perfect SAT scores on being immediately rejected wrt 7 tier1 universities (just below the 75th percentile for Berkeley) and settled for a tier2, e.g. from CollegeData.com dataset,

    ID|State|Class|Status|SAT|Inst
    AAA|MI|2022|Applied|1600|Brown
    AAA|MI|2022|Applied|1600|Cornell
    AAA|MI|2022|Applied|1600|Duke
    AAA|MI|2022|Applied|1600|Princeton
    AAA|MI|2022|Applied|1600|Stanford
    AAA|MI|2022|Applied|1600|UChicago
    AAA|MI|2022|Applied|1600|Yale
    AAA|MI|2022|Wait-Listed|1600|Caltech

    BBB|CA|2022|Denied|1520|Berkeley
    BBB|CA|2022|Denied|1520|Cornell
    BBB|CA|2022|Denied|1520|Duke
    BBB|CA|2022|Denied|1520|Harvard
    BBB|CA|2022|Denied|1520|Princeton
    BBB|CA|2022|Denied|1520|Stanford
    BBB|CA|2022|Denied|1520|UChicago
    BBB|CA|2022|Will Attend|1520|UCLA

    • Replies: @Nathan
    @dux.ie

    Thank you! The 75th percentile of Cal Tech, Chicago, Harvard, and Yale works out to be 1,770 students for the class of 220. There are only about 500 perfect SAT scores last year. There were 504 in 2015:

    https://web.archive.org/web/20170106113421/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

    Your table says explicitly that the 75th percentile is a *composite* 1600.

    Replies: @Jack D, @res

    , @res
    @dux.ie


    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai’s dataset with only SAT75.
     
    I wish you could find that link. It would be very helpful.

    This is the closest thing I found. It has the SAT 25-75 range for many universities.
    https://www.iecaonline.com/wp-content/uploads/2018/01/College-fall-2017-Reality-Check-SAT-25-75-college-Class-of-2020.pdf

    P.S. Some possibly relevant Wai papers: https://pdfs.semanticscholar.org/6fd7/de502146be112ef32d75f6814bea212d8455.pdf
    https://www.psychologytoday.com/files/attachments/56143/wai-americas-elite-2013.pdf
    https://www.journalofexpertise.org/articles/JoE_2018_1_1_Wai_Perina_Mar3.pdf
    , @Calvin Hobbes
    @dux.ie

    I know of several applicants who were rejected by Harvard after getting IMO gold medals and who were subsequently Putnam fellows for MIT.

    Here’s a similar case:

    https://blog.tanyakhovanova.com/2009/07/what-does-it-take-to-get-accepted-by-harvard-or-princeton/

    Replies: @Calvin Hobbes, @dux.ie

  112. Amy Harmon is quite right. Relative to an expected representation due to their intelligence plus affirmative action help they get for attaining professorial positions, blacks are inordinately underrepresented, really. And yes Virginia, it’s because of microaggressions by whites.

    Whites have a More Robust Hypothalamic-Pituitary-Adrenal Axis Response to a Psychological Stressor than Blacks

    I remember reading accounts of combat by white Rhodesians in which they noted that it was surprisingly common for blacks to be shot multiple times and keep fighting. There is a price to be paid for everything.

  113. @Nathan
    @res

    Yes, but I don't want to put the whole excel file up. It's enough to check out Harvard, and extrapolate from there:

    https://college.harvard.edu/admissions/admissions-statistics

    Harvard admitted 2,024 undergraduate students, and claims that their 75th percentile on the SAT is a perfect 800 in the reading and math sections. So 506 students in the admitted class have either a perfect reading or math score. So far, so good. The problem is when you add up all the other colleges that claim similar admissions stats. Yale, Princeton, most of the Ivy league. MIT, Caltech... sure. Washington University in St. Louis?? No freaking way.

    Harvard is probably not lying. Washington University in St. Louis almost certainly is. If you aced the SAT, would you want to go to some no-name recognition school in America's worst (or 2nd worst) city?

    Replies: @dux.ie, @Jack D

    There are plenty to go around. Something like 15,000 kids each year (1%) score 800 on math (of which perhap 2/3 are Asian and maybe a couple of hundred are black(ish). This is enough to supply 300 to each of the top 50 schools.

    • Replies: @Nathan
    @Jack D

    College/ Number Accepted/ 75 Percentile/
    CalTech/ 541/ 135
    Yale/ 2,229/ 557
    Chicago/ 2,348/ 587
    Harvard/ 1962/ 491
    MIT/ 1464/ 366
    Princeton/ 1941/ 485
    Columbia/ 2214/ 554
    Vanderbilt/ 3298/ 825
    Harvey Mudd/ 629/ 157
    Stanford/ 2071/ 518
    Olin Engineering/ 90/ 23
    Johns Hopkins/ 3234/ 809
    Washington STL/ 4708/ 1177
    Duke/ 3219/ 805
    Penn/ 3731/ 933
    Rice/ 2328/ 582
    Northwestern/ 3396/ 849
    Carnegie Mellon/ 4509/ 1127
    Total: 10978

    There you go. You need 10,978 800 Math Scores to go with @dux.ie's very helpful table of admissions data. I SHOULD have included all the schools. You say "about 1%" score an 800 on the math portion every year. I don't think so. Missing one question most years will drop you down not to a 790, but usually a 780. Schools that claim a 75th percentile of 790 or 780 are effectively claiming a portion of 800 scores.

    The College Board has been very UNHELPFUL by not providing me the simple number of people who scored 800 in math in 2018. From past data, it may be as low as .006, or lower.

  114. Anon[214] • Disclaimer says:
    @Gringo
    @Anon

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the “books in the home” or “parents reading to the kids” theories were blown out of the water by twin and adoption studies, and it’s a similar idea to “Ph.D.s in the home.” Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment.

    It also works the other way : not as bright kids adopted into the homes of bright parents. What I personally know about adoption is anecdote, not a research finding, but I will relate it nonetheless. I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.

    Replies: @Anon, @Jack D

    I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.

    Statistically the correlation between parental and child intelligence is indisputable and very strong, but there can be quite a difference in the intelligence between siblings (and this is true in my own family). Mitchell in Innate goes into the details. In childhood adopted siblings’ IQ correlate at 0.25 and biological siblings at 0.60. But the adopted correlation disappears in adulthood (and biological shrinks to 0.5 or less). Identical twins are higher, maybe 0.8.

  115. Anon[214] • Disclaimer says:
    @Steve Sailer
    @Anon

    I suspect it's pretty likely that parents' jobs have a sizable influence on what jobs kids go into. A lot of military officers are the children of a military officer, a lot of college professors are the children of a college professor. Some of that is genetic, but another part is simply that you can imagine yourself doing that job because you saw somebody in the family do it.

    Replies: @Anon, @Anonymous

    In Japan a surprising number of doctors have a doctor parent, more so I think than in the U.S.

    My theory is that in Japan you go right into medical school after high school, like pharmacy school in the U.S. So it’s not a career path decision that you can put off too long. You pretty much have to have made the decision by the time you enter high school at the latest, and without the guidance of a physician parent a lot of early teens are not going to be thinking about stuff like this.

    In the U.S. on the other hand if you’re a STEM major with strong grades you can wait until you’ve almost graduated college to choose a medical career.

    And there are a few medical schools that have slots for non-STEM people, although I expect this is almost the harder route, given the few number of such slots:

    https://www.nytimes.com/2010/07/30/nyregion/30medschools.html

    For generations of pre-med students, three things have been as certain as death and taxes: organic chemistry, physics and the Medical College Admission Test, known by its dread-inducing acronym, the MCAT.

    So it came as a total shock to Elizabeth Adler when she discovered, through a singer in her favorite a cappella group at Brown University, that one of the nation’s top medical schools admits a small number of students every year who have skipped all three requirements.

    Until then, despite being the daughter of a physician, she said, “I was kind of thinking medical school was not the right track for me.”

  116. @res
    @res

    I went back and looked at that Brookings report again. Turns out I had missed an important detail (emphasis mine).


    The College Board’s publicly available data provides data on racial composition at 50-point score intervals. We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos. (These estimates—which rely on conservative assumptions that maximize the number of high-scoring black students, are consistent with an older estimate from a 2005 paper in the Journal of Blacks in Higher Education, which found that only 244 black students scored above a 750 on the math section of the SAT.)
     
    Oh boy. If they consider "at most 1,000" consistent with 244 that says something. What's a factor of 4 between friends?

    Does anyone know where to find the publicly available "data on racial composition at 50-point score intervals"?

    Replies: @Gringo

    Oh boy. If they consider “at most 1,000” consistent with 244 that says something. What’s a factor of 4 between friends?

    Brookings informs us that whites form a decreasing percentage at higher test intervals.
    650-700 62% whites
    700-75o 54% whites
    750-800 33% whites

    But blacks do not form a decreasing percentage at higher test intervals.
    650-700 2% blacks
    700-75o 2% blacks
    750-800 2% blacks

    That looks to me like cooked data, which is implied in “at most.”

  117. @William Badwhite
    @Reg Cæsar

    Kristofferson also served in the Army as an officer. He was a helicopter pilot.

    Replies: @Reg Cæsar, @Hibernian

    He was also born before John McCain.

  118. @ic1000
    @SafeNow

    > I don’t hear Amy Harmon demanding proportional employment in the airline cockpit

    marginally competent pilots flying the planes she and her daughter ride in

    > plumbers’ union

    more or less proficient plumbers doing the pipes in her place out in the Hamptons

    > and such other things she uses.

    bottom-of-their-class critical-care physicians managing the complex cases when it's life or death for her family members

    All sorts of compelling, heart-wrenching special pleading as to why the circumstances faced by Amy and her loved ones make that sort of reasoning inapplicable. Nomenklatura, not apparatchik.

    Replies: @Reg Cæsar

    Remember that actress who was murdered by the Peruvian illegal she and her husband hired to fix their shower? Adrienne Somethingorother, age 40.

    Gee, had they only been willing to pay an American…

  119. @res
    @res

    The Brookings numerical analysis has a strong smell of fraud about it, so I decided to take a closer look at various numbers they gave and check them for consistency. Here are some salient points from the Brookings article and associated ETS documents. I believe these are all using the 2015 ETS data.


    We estimate that in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.
     
    Table 7 of https://secure-media.collegeboard.org/digitalServices/pdf/sat/total-group-2015.pdf has scoring distributions and demographics

    The Brookings graphic (Race gaps at the tails, SAT math scores) I linked above has estimated percentiles scoring above 750.

    I combined those sources into a spreadsheet (with formulas in original appearing as computed values in CSV). It appears after the MORE tag as a CSV if anyone wants to take a look. There are some issues because Brookings lumps three different Latino groups together and ignores the "Other" and "No Response" categories which I estimate account for 10% of Math scores over 750.

    Using the Table 7 numbers I calculated the expected number of scores over 750 assuming a normal distribution. The black estimate was 140 (!). I suspect the distribution is fat tailed so I find the JBHE 244 number at least plausible (but would like to see details of how they computed it). But both of those numbers are only loosely related to "at most 1000." Also, the white estimate is low given the mean/SD. Odd.

    I suspect what Brookings did was make assumptions about the "Other" and "No Response" categories which favored the black estimates.

    In any case, I think my analysis shows their presentation was far enough from reality to be considered fraudulent. It would be interesting to see a better calculation using the full ETS data (which is quite possible). I wonder why that is not available?

    P.S. Here is another broken link from the Brookings article. Funny how the most informative links seem to be the ones that break.
    https://web.archive.org/web/20160310073125/https://cshe.berkeley.edu/sites/default/files/shared/publications/docs/ROPS.CSHE_.10.15.Geiser.RaceSAT.10.26.2015.pdf



    Race,N,Math Mean,Math SD,SDs for 750,Est prop 750,Est N 750,Est percent 750,Brookings Est N,Ratio of Estimates,Brookings percent 750 as N,Est percent 750
    American Indian,"10,031",482,107,2.50,0.0061,61,0.14%,,,,
    Asian,"211,238",598,127,1.20,0.1157,24437,55.06%,29570,0.83,26628,
    Black,"219,018",428,100,3.22,0.0006,140,0.32%,1000,0.14,888,
    Mexican,"130,026",457,98,2.99,0.0014,181,0.41%,967,0.19,894,
    Puerto Rican,"30,192",449,106,2.84,0.0023,68,0.15%,224,0.30,208,
    Other Hispanic,"162,655",457,107,2.74,0.0031,502,1.13%,1209,0.42,1118,
    White,"800,236",534,104,2.08,0.0189,15128,34.09%,"16,000",0.95,14646,
    Other ,"65,063",519,123,1.88,0.0302,1964,4.43%,,,,
    No Response,"70,062",492,134,1.93,0.0271,1898,4.28%,,,,
    Total,"1,698,521",511,120,1.99,0.0232,39412,88.81%,,,,
    ,,,,,,,,,,,
    Calculated Totals,"1,698,521",510.9,108.2,,,44380,,48970,0.91,44380,

    Replies: @Jack D, @dux.ie

    Your calculations assumed that the distributions are perfect bell curve which might not be the case. The College Board’s data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors. The results are close to that from Brookings, especially for large Pct75, e.g. “””16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.””” compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.

    sum(N75)=53700

    Eth:Pct75 Nsub N75 Pct Brookings
    Bla: 1 219018 2190 4.08% 2%
    Nam: 1 10031 100 0.19%
    Asi: 14 211238 29573 55.07% 60%
    Mex: 1 130026 1300 2.42%
    Pue: 1 30192 302 0.56%
    Lat: 1 162655 1627 3.03% 5%
    Whi: 2 800236 16005 29.80% 33%
    Oth: 4 65063 2603 4.85%

    • Replies: @res
    @dux.ie


    Your calculations assumed that the distributions are perfect bell curve which might not be the case.
     
    Right. That was why I explicitly said: "assuming a normal distribution." And later qualified "I suspect the distribution is fat tailed." And then backed off to the JBHE number rather than my own estimate (roughly a factor of two difference).

    How often do you discuss your assumptions so clearly?

    The College Board’s data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors.
     
    Right. I linked that document above. The problem is that for blacks (the case under consideration), 650 was 98th percentile, 700 was 99th percentile, and 750 (the case of interest) was 99+. Rendering the rounding off errors much less acceptable than the normal assumption. I included the other cases to give perspective on how accurate my methodology seemed to be.

    FWIW, my guess is their percentiles are integer rounded (so 98.5 - 99.49 is 99, and > 99.5 is 99+). If anyone knows otherwise, please give a reference. That would turn your 2190 blacks scoring over 750 into 1095 which is much closer to reality (but what's a factor of 2 error between friends? not to mention the additional factor of 4 which is likely, so you are probably only a factor of 8 off for the black case. nice work. perhaps that explains why I chose to do things the other way?). Thinking about this, that's probably how Brookings got "not more than 1,000." Not some elaborate calculation involving "Other" and "No Response" like I speculated.

    The results are close to that from Brookings, especially for large Pct75, e.g. “””16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.””” compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.
     
    I find some of your comments interesting (they would be even better if you included your data sources), but that you post that black number (2190) without apparently realizing those "large rounding off errors" make the analysis unusable for blacks make me question the validity of anything you write. I suspect the reason your numbers agree so well with Brookings is that they did the same calculations except for your error with 99+ with blacks and Brookings ignoring the other groups in their aggregate percentage.

    In addition, you only looked at one of the three Latino groups--Mexican and Puerto Rican being the other two. As far as I can tell Brookings aggregated them given the percentages they quoted (I also noted this in my comment, did you miss that?). That seems like another rather substantial error (leaving out half of all Latinos, both groups having 750 at the 99+ percentile). Perhaps you should consider either putting down the stones or moving out of your glass house?

    P.S. I went back and double checked. From the Brookings article:

    (The College Board has separate categories for “Mexican or Mexican American” and “Other Hispanic, Latino, or Latin American.” We have combined them under the term Latino.)
     
    So it looks like they ignored the poorly performing (lowest mean of the Latino groups) Puerto Ricans (about 30k people, relative to 130k Mexicans). Funny definition of Latino they have that excludes Puerto Ricans.

    Also, anyone have any idea why ETS left "No Response" out of their percentile tables? That group accounts for about the same proportions as "Other" (i.e. a significant--about 5%--proportion of >750 scorers).

    Replies: @dux.ie

  120. Amy Harmon is obviously correct: Wakanda Math is the answer to any science-y question… why, Wakanda’s quadroon lady-scientists think of all the smart things that the white scientists ought to have done (especially to get the stone out of Vision’s head, apparently).

    OK, two Marvel references is too many… however in my defence I spent half the day berating my 8 year old nephew whose world-view had been infected by the chanting that the Wakandans apparently do in some Marvel movie he was watching on his iThing.

    He cried, and probably still thinks I’m the “horrible uncle”… but better that he is inoculated against nonsense now, than that he suffers the consequences of ignorance later. (I disabused him about Santa a few years back… he got over it; someone’s gotta do it… his mother’s a textbook Delta in the Huxley taxonomy).

  121. Harmon is the typical Jewish nepotism hire. There’s no way that she obtained that position through merit.

  122. @Jack D
    @res

    I don't remember now what link I looked at and don't feel like finding them again but the link you gave looks good. 13% Jewish for MIT is very high - MIT is not a particularly Jewish place - although Jews are not unknown there, they are not particularly common either, at least in relation to some of the other Ivies.

    Hillel (known for overstating Jewish %'s) claims 7% of undergrads:

    https://www.hillel.org/college-guide/list/record/massachusetts-institute-of-technology

    (They claim 11% for Harvard and 17% for Penn).


    Jews nowadays are not that interested in STEM subjects - they have all but disappeared from the math competitions, etc. The real money is in finance, law, politics, etc. and math is the kind of hard grinding work that immigrant kids are willing to put in their 10,000 hours for - American Jews are now several generations out of the ghetto and the taste of desperate poverty and need to please your immigrant parents who have sacrificed everything so you could go to college has worn off. As I said before, Asians are the new Jews.

    Replies: @res, @Hibernian

    Finance has a significant (albeit relatively low powered) STEM component.

    • Replies: @Jack D
    @Hibernian

    Askenazi brains have not disappeared (although they increasingly will as there are fewer and fewer full blooded Ashkenazis due to intermarriage - Jussie is an extreme example) but they have re-directed their energies toward different pursuits.

    John Adams noted the generational nature of intellectual pursuits even 240 years ago:


    I must study Politicks and War that my sons may have liberty to study Painting and Poetry Mathematicks and Philosophy. My sons ought to study Mathematicks and Philosophy, Geography, natural History, Naval Architecture, navigation, Commerce and Agriculture, in order to give their Children a right to study Painting, Poetry, Musick, Architecture, Statuary, Tapestry and Porcelaine

     

    A lot of Jews have passed the Mathematicks and Commerce generation and are up to the Poetry and Musick stage.

    Replies: @Jack D

  123. @William Badwhite
    @Reg Cæsar

    Kristofferson also served in the Army as an officer. He was a helicopter pilot.

    Replies: @Reg Cæsar, @Hibernian

    Additionally he taught English at West Point for a while.

  124. Anon[214] • Disclaimer says:
    @anon
    This is my first post to Unz Review (although iSteve was kind enough once before to post something I'd written elsewhere online under my pseudonym A New Radical Centrism).

    Unfortunately, I come bearing some unpleasant personal news, and I'm afraid it involves AMY HARMON (I prefer to refer to her in ALL CAPS).

    After responding about an hour ago to a tweet by "genetics & social inequality" superstar Dr. Paige Harden, in which I breathlessly extolled the reportorial virtues, impartiality, and integrity of ace science journalist AMY HARMON, I was unceremoniously blocked by Dr. Harden.

    I won't unnecessarily go into the details here (you can view them yourself at https://twitter.com/kph3k), but Dr. Harden had earlier tweeted the following:

    'Really want a science journalist to write a piece on how open science practices (preprints, open data) - while clearly a net plus for replicability and speed of scientific communication — have made it easier for eugenicist pseudoscience to thrive.”

    I responded to the tweet by enthusiastically recommending AMY HARMON for the task and then -– BLAM! -- before you knew it, I was blocked!

    There's no accounting for this kind of behavior.

    Replies: @Anon, @Anon

    Give Paige Harden a break. I think she knows the score, and just wants to keep her job. She has a book coming out called The Genetic Lottery.

    https://www.unz.com/isteve/paige-hardens-upcoming-book-the-genetic-lottery/

    As far as I can tell, the premise is that our IQs are unearned, gained by random mitosis of our germ cells and random developmental processes in the womb.

    So high IQ people need to check their privilege and cooperate in the redistribution of the excess unearned income that they receive from their random luck.

    In other words, Dr. Harden’s book will be saying that, okay, maybe blacks have low IQ, but so what, and we should give them free money in some sort of universal income scheme. She will not be literally saying that blacks have low IQs. More something along the lines of some people have IQs below 85 or 100 or whatever, and they should be supported. But really, if it isn’t about blacks, would there really even be a book? Would anyone care about low IQ whites?

  125. Anon[589] • Disclaimer says:
    @anon
    This is my first post to Unz Review (although iSteve was kind enough once before to post something I'd written elsewhere online under my pseudonym A New Radical Centrism).

    Unfortunately, I come bearing some unpleasant personal news, and I'm afraid it involves AMY HARMON (I prefer to refer to her in ALL CAPS).

    After responding about an hour ago to a tweet by "genetics & social inequality" superstar Dr. Paige Harden, in which I breathlessly extolled the reportorial virtues, impartiality, and integrity of ace science journalist AMY HARMON, I was unceremoniously blocked by Dr. Harden.

    I won't unnecessarily go into the details here (you can view them yourself at https://twitter.com/kph3k), but Dr. Harden had earlier tweeted the following:

    'Really want a science journalist to write a piece on how open science practices (preprints, open data) - while clearly a net plus for replicability and speed of scientific communication — have made it easier for eugenicist pseudoscience to thrive.”

    I responded to the tweet by enthusiastically recommending AMY HARMON for the task and then -– BLAM! -- before you knew it, I was blocked!

    There's no accounting for this kind of behavior.

    Replies: @Anon, @Anon

    ‘Really want a science journalist to write a piece on how open science practices (preprints, open data) – while clearly a net plus for replicability and speed of scientific communication — have made it easier for eugenicist pseudoscience to thrive.”

    I responded to the tweet by enthusiastically recommending AMY HARMON for the task and then -– BLAM! — before you knew it, I was blocked!

    I think that there is a “rest of the story” here, perhaps contained in the word “enthusiastically.” Was your recommendation along the lines of, “Sounds like a job for that IDIOT New York Times so-called writer Amy Harmon, who should learn to code!!!!!”?

  126. @Jack D
    @Nathan

    There are plenty to go around. Something like 15,000 kids each year (1%) score 800 on math (of which perhap 2/3 are Asian and maybe a couple of hundred are black(ish). This is enough to supply 300 to each of the top 50 schools.

    Replies: @Nathan

    College/ Number Accepted/ 75 Percentile/
    CalTech/ 541/ 135
    Yale/ 2,229/ 557
    Chicago/ 2,348/ 587
    Harvard/ 1962/ 491
    MIT/ 1464/ 366
    Princeton/ 1941/ 485
    Columbia/ 2214/ 554
    Vanderbilt/ 3298/ 825
    Harvey Mudd/ 629/ 157
    Stanford/ 2071/ 518
    Olin Engineering/ 90/ 23
    Johns Hopkins/ 3234/ 809
    Washington STL/ 4708/ 1177
    Duke/ 3219/ 805
    Penn/ 3731/ 933
    Rice/ 2328/ 582
    Northwestern/ 3396/ 849
    Carnegie Mellon/ 4509/ 1127
    Total: 10978

    There you go. You need 10,978 800 Math Scores to go with ’s very helpful table of admissions data. I SHOULD have included all the schools. You say “about 1%” score an 800 on the math portion every year. I don’t think so. Missing one question most years will drop you down not to a 790, but usually a 780. Schools that claim a 75th percentile of 790 or 780 are effectively claiming a portion of 800 scores.

    The College Board has been very UNHELPFUL by not providing me the simple number of people who scored 800 in math in 2018. From past data, it may be as low as .006, or lower.

  127. @dux.ie
    @Nathan

    > The problem is when you add up all the other colleges that claim similar admissions stats.

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai's dataset with only SAT75. IQlike75 is separately calculated. MIT no. Princeton no. WUST no.

    Rank|SAT75|SATV75|SATM75|IQlike75|Inst
    1|1600|800|800|140.4|California Institute of Technology
    1|1600|800|800|140.4|Yale University
    1|1600|800|800|140.4|University of Chicago
    1|1600|800|800|140.4|Harvard University

    5|1590|790|800|139.7|Massachusetts Institute of Technology
    5|1590|790|800|139.7|Princeton University
    5|1590|790|800|139.7|Columbia University in the City of New York
    5|1590|790|800|139.7|Vanderbilt University
    9|1580|780|800|138.9|Harvey Mudd College
    9|1580|780|800|138.9|Stanford University
    9|1580|780|800|138.9|Franklin W Olin College of Engineering
    12|1570|770|800|138.2|Johns Hopkins University
    12|1570|770|800|138.2|Washington University in St Louis
    12|1570|770|800|138.2|Duke University
    12|1570|770|800|138.2|University of Pennsylvania
    12|1570|780|790|138.2|Brown University
    12|1570|770|800|138.2|Rice University
    18|1560|760|800|137.5|Northwestern University
    18|1560|780|780|137.5|Dartmouth College
    19|1555|775|780|137.1|Amherst College
    20|1550|750|800|136.8|Carnegie Mellon University
    21|1540|750|790|136.0|University of California-Berkeley
    51|1470|710|760|131.0|University of California-Los Angeles

    > If you aced the SAT, would you want to go to some no-name recognition school

    Top universities rejected more than accepted students with perfect scores (especially from some ethnic groups) to make room for the subpar applicants. There are self reported cases of those with perfect scores of 1600 in limbo wrt 7 tier1 universities and were on the waiting lists of another (he/she will be smart to start applying to some tier2 immediately), or with near perfect SAT scores on being immediately rejected wrt 7 tier1 universities (just below the 75th percentile for Berkeley) and settled for a tier2, e.g. from CollegeData.com dataset,

    ID|State|Class|Status|SAT|Inst
    AAA|MI|2022|Applied|1600|Brown
    AAA|MI|2022|Applied|1600|Cornell
    AAA|MI|2022|Applied|1600|Duke
    AAA|MI|2022|Applied|1600|Princeton
    AAA|MI|2022|Applied|1600|Stanford
    AAA|MI|2022|Applied|1600|UChicago
    AAA|MI|2022|Applied|1600|Yale
    AAA|MI|2022|Wait-Listed|1600|Caltech
    ---
    BBB|CA|2022|Denied|1520|Berkeley
    BBB|CA|2022|Denied|1520|Cornell
    BBB|CA|2022|Denied|1520|Duke
    BBB|CA|2022|Denied|1520|Harvard
    BBB|CA|2022|Denied|1520|Princeton
    BBB|CA|2022|Denied|1520|Stanford
    BBB|CA|2022|Denied|1520|UChicago
    BBB|CA|2022|Will Attend|1520|UCLA

    Replies: @Nathan, @res, @Calvin Hobbes

    Thank you! The 75th percentile of Cal Tech, Chicago, Harvard, and Yale works out to be 1,770 students for the class of 220. There are only about 500 perfect SAT scores last year. There were 504 in 2015:

    https://web.archive.org/web/20170106113421/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

    Your table says explicitly that the 75th percentile is a *composite* 1600.

    • Replies: @Jack D
    @Nathan

    Most colleges will take your highest score on each subsection if you took the SAT more than once (which many students do) and they will convert ACT scores to SAT. Only 500 people got 1600s at the same sitting but the number of people who got an 800 on either section is much higher and to that you can add the # of people who got 36's on ACT English or Math.

    Replies: @Nathan

    , @res
    @Nathan

    Here are the corresponding pages for the 2015 subtests:
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-crit-reading-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-writing-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-mathematics-2015.pdf

    16668 Math 800s in 2015
    9906 Critical Reading 800s in 2015 (CR is equivalent to Verbal in 2 subtest years).

    1,698,521 total test takers if you like percentages. So Math 800 is about top 1%, Verbal 800 a little above 0.5%.

    So assuming those numbers are similar to 2018, your analysis showing 10,978 800 Math Scores accounts for about 2/3 of them. That actually sounds possible to me.

    Replies: @Nathan

  128. @Nathan
    @dux.ie

    Thank you! The 75th percentile of Cal Tech, Chicago, Harvard, and Yale works out to be 1,770 students for the class of 220. There are only about 500 perfect SAT scores last year. There were 504 in 2015:

    https://web.archive.org/web/20170106113421/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

    Your table says explicitly that the 75th percentile is a *composite* 1600.

    Replies: @Jack D, @res

    Most colleges will take your highest score on each subsection if you took the SAT more than once (which many students do) and they will convert ACT scores to SAT. Only 500 people got 1600s at the same sitting but the number of people who got an 800 on either section is much higher and to that you can add the # of people who got 36’s on ACT English or Math.

    • Replies: @Nathan
    @Jack D

    That's just cheating on everyone's part.

    Replies: @Jack D

  129. I have a suspicion, but no proof, that the New York Times comments editor stopped the comments thread on this article after twenty comments, when the comments started veering in a dangerously truthful direction. If so, it is a bit of black-arts censorship to block public information on HBD. Am I imagining this? The comments thread was hot and then just stopped suddenly and never re-started, soon after the article appeared.

    • Replies: @Gringo
    @Peter Johnson

    Most of the NYT comments support the "progressive" views of the typical NYT article. Many and perhaps most of the comments debunked the "progressive" tone of Amy Harmon's article, so you may have a point. Perhaps they stop reviewing posted comments for publication after x number of days.

    BTW, they have added several more comments.
    My initial comment, after 24 hours, had not gotten posted.
    Guessing that this was because I had embedded my link, I posted my comment with a "naked" link. Within a short amount of time, my comment was posted.

  130. @Jack D
    @Nathan

    Most colleges will take your highest score on each subsection if you took the SAT more than once (which many students do) and they will convert ACT scores to SAT. Only 500 people got 1600s at the same sitting but the number of people who got an 800 on either section is much higher and to that you can add the # of people who got 36's on ACT English or Math.

    Replies: @Nathan

    That’s just cheating on everyone’s part.

    • Replies: @Jack D
    @Nathan

    Maybe so (though I don't agree) but it helps to explain how HYP etc. can claim that their 75th %ile is 1600 when only a few hundred per year get 1600 AT ONE SITTING.

    Replies: @res

  131. @JimB
    Since the renorming of the SAT, 750 corresponds to 99%tile. It’s an understatement to say math PhDs are in the top 1% (2.5 sigma). I’m guessing from personal experience that productive math grads are more like NFL players, from the 5 sigma level of talent. At that level, you are selecting from tiny sub populations of Europeans, East Asians, and Indians, only.

    Replies: @Edward

    Since the renorming of the SAT, 750 corresponds to 99%tile. It’s an understatement to say math PhDs are in the top 1% (2.5 sigma). I’m guessing from personal experience that productive math grads are more like NFL players, from the 5 sigma level of talent. At that level, you are selecting from tiny sub populations of Europeans, East Asians, and Indians, only.

    5-sigma is too high. The Fields Medalist Richard Borcherds has a performance IQ (a measure of nonverbal ability) of 147 (top 0.09%) and a full-scale IQ of 137 (top 0.7%).

    https://www.tandfonline.com/doi/abs/10.1080/13554799908402743

    Performance IQ is only a proxy for mathematical ability, but if his nonverbal IQ is a bit over 3 SD above the mean, his mathematical ability is probably only going to be ~4SD above the mean, at most.

    Borcherds is probably on the low end of the ability distribution of Fields Medalists (Terence Tao reached the ceiling on the Stanford-Binet, with an IQ of 175). Even so, you’re still right: it shouldn’t be surprising that the only minority groups we see with good representation in math departments are Jewish-Americans (avg IQ 110-115, with their mathematical abilities probably even higher than this), Indian-Americans (avg IQ ~110) and East Asian-Americans (avg IQ ~106, with mathematical abilities probably higher than this).

    • Replies: @Peter Johnson
    @Edward

    One important correction to your comments -- as David Reich has shown, Indians (from India) are best not viewed as a single ethnic group but rather as a collection of quite distinct ethnic groups. Some of these ethnic groups have high average IQ and others do not. The country is unusual in that respect.

    Replies: @Edward

  132. @Hibernian
    @Jack D

    Finance has a significant (albeit relatively low powered) STEM component.

    Replies: @Jack D

    Askenazi brains have not disappeared (although they increasingly will as there are fewer and fewer full blooded Ashkenazis due to intermarriage – Jussie is an extreme example) but they have re-directed their energies toward different pursuits.

    John Adams noted the generational nature of intellectual pursuits even 240 years ago:

    I must study Politicks and War that my sons may have liberty to study Painting and Poetry Mathematicks and Philosophy. My sons ought to study Mathematicks and Philosophy, Geography, natural History, Naval Architecture, navigation, Commerce and Agriculture, in order to give their Children a right to study Painting, Poetry, Musick, Architecture, Statuary, Tapestry and Porcelaine

    A lot of Jews have passed the Mathematicks and Commerce generation and are up to the Poetry and Musick stage.

    • Replies: @Jack D
    @Jack D

    PS Adams forgot to say what happens to the kids of the parents who study Statuary and Porcelaine. The usual answer is "from shirtsleeves to shirtsleeves in three generations" - it's back to square one. Although the Adams family has distinguished descendants to this day.

  133. @Jack D
    @Hibernian

    Askenazi brains have not disappeared (although they increasingly will as there are fewer and fewer full blooded Ashkenazis due to intermarriage - Jussie is an extreme example) but they have re-directed their energies toward different pursuits.

    John Adams noted the generational nature of intellectual pursuits even 240 years ago:


    I must study Politicks and War that my sons may have liberty to study Painting and Poetry Mathematicks and Philosophy. My sons ought to study Mathematicks and Philosophy, Geography, natural History, Naval Architecture, navigation, Commerce and Agriculture, in order to give their Children a right to study Painting, Poetry, Musick, Architecture, Statuary, Tapestry and Porcelaine

     

    A lot of Jews have passed the Mathematicks and Commerce generation and are up to the Poetry and Musick stage.

    Replies: @Jack D

    PS Adams forgot to say what happens to the kids of the parents who study Statuary and Porcelaine. The usual answer is “from shirtsleeves to shirtsleeves in three generations” – it’s back to square one. Although the Adams family has distinguished descendants to this day.

  134. @Gringo
    @Anon

    There are not enough Ph.D.s around for this to be susceptible to twin and adoption studies, but for instance, the “books in the home” or “parents reading to the kids” theories were blown out of the water by twin and adoption studies, and it’s a similar idea to “Ph.D.s in the home.” Kids with smart parents, who read and have books (and have Ph.D.s more than dumb parents) read books and become smart themselves, even if they were adopted into a home of dumb parents. They manage to create their own environment.

    It also works the other way : not as bright kids adopted into the homes of bright parents. What I personally know about adoption is anecdote, not a research finding, but I will relate it nonetheless. I know some examples of bright parents with terminal degrees who adopted children of normal intelligence. It did not turn out well for the adopted children. Part of the problem was that both parents and children had difficulty adjusting to the intelligence differences between them.

    Replies: @Anon, @Jack D

    I have mentioned before that my wife (who works in the educational field) has observed this same situation often – the adopted child of average or below average intelligence (which is most of them – nowadays high IQ people tend not to put their kids up for adoption) is adopted into a high achievement wealthy family where all the children are expected to attend a private university, preferably one of Ivy League calibre or at least in the top 50. This leads to a lot of stress as the kid is really not set up to operate at this level. If he had stayed with his birth family they would have been fine with the kid becoming a mechanic or a beautician but in his adoptive families circles this would be seen as a shocking failure.

    Vice versa (which used to happen more in the past when high IQ college girls were not supposed to be single mothers and abortion was not available) such as Steve Jobs tends to work out better as there are enough resources in our society for bright and ambitious blue collar kids to find their way without a lot of practical support from their families (as long as they get emotional support). Also that being bright means that you don’t need a lot of practical support – you figure this stuff out on your own. In the “Three Identical Strangers” movie, after the brothers were reunited, they spent the most time at the home of the “blue collar” adoptee, whose parents (especially the father) was the most emotionally supportive. The “blue collar” brother was also the one who was the most mentally intact of the three. For kids of high intelligence, a loving home is perhaps more important than having really smart parents.

  135. So what is the big deal? I was a math major in college and most of the abstract math at higher levels is useless in the real world.

  136. @Peter Johnson
    I have a suspicion, but no proof, that the New York Times comments editor stopped the comments thread on this article after twenty comments, when the comments started veering in a dangerously truthful direction. If so, it is a bit of black-arts censorship to block public information on HBD. Am I imagining this? The comments thread was hot and then just stopped suddenly and never re-started, soon after the article appeared.

    Replies: @Gringo

    Most of the NYT comments support the “progressive” views of the typical NYT article. Many and perhaps most of the comments debunked the “progressive” tone of Amy Harmon’s article, so you may have a point. Perhaps they stop reviewing posted comments for publication after x number of days.

    BTW, they have added several more comments.
    My initial comment, after 24 hours, had not gotten posted.
    Guessing that this was because I had embedded my link, I posted my comment with a “naked” link. Within a short amount of time, my comment was posted.

  137. @dux.ie
    @res

    Your calculations assumed that the distributions are perfect bell curve which might not be the case. The College Board's data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors. The results are close to that from Brookings, especially for large Pct75, e.g. """16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.""" compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.

    sum(N75)=53700

    Eth:Pct75 Nsub N75 Pct Brookings
    Bla: 1 219018 2190 4.08% 2%
    Nam: 1 10031 100 0.19%
    Asi: 14 211238 29573 55.07% 60%
    Mex: 1 130026 1300 2.42%
    Pue: 1 30192 302 0.56%
    Lat: 1 162655 1627 3.03% 5%
    Whi: 2 800236 16005 29.80% 33%
    Oth: 4 65063 2603 4.85%

    Replies: @res

    Your calculations assumed that the distributions are perfect bell curve which might not be the case.

    Right. That was why I explicitly said: “assuming a normal distribution.” And later qualified “I suspect the distribution is fat tailed.” And then backed off to the JBHE number rather than my own estimate (roughly a factor of two difference).

    How often do you discuss your assumptions so clearly?

    The College Board’s data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors.

    Right. I linked that document above. The problem is that for blacks (the case under consideration), 650 was 98th percentile, 700 was 99th percentile, and 750 (the case of interest) was 99+. Rendering the rounding off errors much less acceptable than the normal assumption. I included the other cases to give perspective on how accurate my methodology seemed to be.

    FWIW, my guess is their percentiles are integer rounded (so 98.5 – 99.49 is 99, and > 99.5 is 99+). If anyone knows otherwise, please give a reference. That would turn your 2190 blacks scoring over 750 into 1095 which is much closer to reality (but what’s a factor of 2 error between friends? not to mention the additional factor of 4 which is likely, so you are probably only a factor of 8 off for the black case. nice work. perhaps that explains why I chose to do things the other way?). Thinking about this, that’s probably how Brookings got “not more than 1,000.” Not some elaborate calculation involving “Other” and “No Response” like I speculated.

    The results are close to that from Brookings, especially for large Pct75, e.g. “””16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.””” compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.

    I find some of your comments interesting (they would be even better if you included your data sources), but that you post that black number (2190) without apparently realizing those “large rounding off errors” make the analysis unusable for blacks make me question the validity of anything you write. I suspect the reason your numbers agree so well with Brookings is that they did the same calculations except for your error with 99+ with blacks and Brookings ignoring the other groups in their aggregate percentage.

    In addition, you only looked at one of the three Latino groups–Mexican and Puerto Rican being the other two. As far as I can tell Brookings aggregated them given the percentages they quoted (I also noted this in my comment, did you miss that?). That seems like another rather substantial error (leaving out half of all Latinos, both groups having 750 at the 99+ percentile). Perhaps you should consider either putting down the stones or moving out of your glass house?

    P.S. I went back and double checked. From the Brookings article:

    (The College Board has separate categories for “Mexican or Mexican American” and “Other Hispanic, Latino, or Latin American.” We have combined them under the term Latino.)

    So it looks like they ignored the poorly performing (lowest mean of the Latino groups) Puerto Ricans (about 30k people, relative to 130k Mexicans). Funny definition of Latino they have that excludes Puerto Ricans.

    Also, anyone have any idea why ETS left “No Response” out of their percentile tables? That group accounts for about the same proportions as “Other” (i.e. a significant–about 5%–proportion of >750 scorers).

    • Replies: @dux.ie
    @res

    If you have deal with the ACT data you will learn to avoid assuming bell curve distributions at all time unless when there is no choice. "No child left behind"? Right, the left hand tail seems to have been compressed to shift the bottom scores upward, and reduces the ethnic score gap at the same time. The compression of the left tail seems to have broken the ACT maths distributions, they are distinctly bimodal. Try plotting out the ACT math histogram for any years. "The Asians are scoring too high in ACT math". Right, it seems that the Asian ACT math distribution right hand tail is compressed and the ceiling score lowered, resulting in the Asian ACT math distribution to be distinctly "trimodal", i.e. if you are lucky to grab from the ACT site to see the 2013 report which was mentioned here at unz.com before it disappeared. A scaled down by the Asian% of the effects of the third mode and lower ceiling can be seen in the national distribution right hand tail if you look hard enough.

    http://i63.tinypic.com/1z5lmj8.png

    Replies: @res, @Nathan

  138. @Barnard
    @Jack D

    They are willing to settle for 12%? My impression is there isn't a number that would satisfy them. Their percentage of the population has no influence on what they think they deserve.

    Replies: @ogunsiron

    The *minimum* acceptable percentage of blacks as members of a desirable group is more or less max( % black in neighborhood, % black in city, % black in state, % black in country, % black in the world). The min starts there.

    There is of course no max acceptable % of blacks in desirable categories.

    For undesirable categories there’s a formula for the max acceptable percentage. Not too difficult to figure out.

  139. @Nathan
    @Jack D

    That's just cheating on everyone's part.

    Replies: @Jack D

    Maybe so (though I don’t agree) but it helps to explain how HYP etc. can claim that their 75th %ile is 1600 when only a few hundred per year get 1600 AT ONE SITTING.

    • Replies: @res
    @Jack D


    Maybe so (though I don’t agree)
     
    Cheating or not, it makes the numbers more difficult to interpret (e.g. estimating IQs). Though they probably consider that a feature rather than bug. I guess it just wasn't good enough to pick the highest aggregate out of multiple attempts.

    Replies: @Jack D

  140. @Nathan
    @dux.ie

    Thank you! The 75th percentile of Cal Tech, Chicago, Harvard, and Yale works out to be 1,770 students for the class of 220. There are only about 500 perfect SAT scores last year. There were 504 in 2015:

    https://web.archive.org/web/20170106113421/https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

    Your table says explicitly that the 75th percentile is a *composite* 1600.

    Replies: @Jack D, @res

    Here are the corresponding pages for the 2015 subtests:
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-crit-reading-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-writing-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-mathematics-2015.pdf

    16668 Math 800s in 2015
    9906 Critical Reading 800s in 2015 (CR is equivalent to Verbal in 2 subtest years).

    1,698,521 total test takers if you like percentages. So Math 800 is about top 1%, Verbal 800 a little above 0.5%.

    So assuming those numbers are similar to 2018, your analysis showing 10,978 800 Math Scores accounts for about 2/3 of them. That actually sounds possible to me.

    • Replies: @Nathan
    @res

    Yes, my number crunching hasn't really panned out, especially considering that Harvard/Yale/Princeton will let you just *keep taking the test until it adds up to 1600*. No wonder they're 1,200 over what I assumed was possible. My suspicions were always directed more toward colleges that shoot up in the US News ranks for mysterious reasons,anyway. Not so much actual elite schools. Washington University in St. Louis... you know. Besides, It's not like schools haven't been caught doing it:

    www.insidehighered.com/admissions/article/2018/07/30/temple-admits-it-provided-false-rankings-data-six-more-programs

    www.washingtonpost.com/blogs/college-inc/post/claremont-mckenna-sat-scandal-more-at-stake-than-rankings/2012/02/07/gIQAHImVwQ_blog.html?utm_term=.d5a83db7f3ca

    www.insidehighered.com/news/2013/01/26/bucknell-misreported-test-scores-2006-2012

  141. @dux.ie
    @Nathan

    > The problem is when you add up all the other colleges that claim similar admissions stats.

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai's dataset with only SAT75. IQlike75 is separately calculated. MIT no. Princeton no. WUST no.

    Rank|SAT75|SATV75|SATM75|IQlike75|Inst
    1|1600|800|800|140.4|California Institute of Technology
    1|1600|800|800|140.4|Yale University
    1|1600|800|800|140.4|University of Chicago
    1|1600|800|800|140.4|Harvard University

    5|1590|790|800|139.7|Massachusetts Institute of Technology
    5|1590|790|800|139.7|Princeton University
    5|1590|790|800|139.7|Columbia University in the City of New York
    5|1590|790|800|139.7|Vanderbilt University
    9|1580|780|800|138.9|Harvey Mudd College
    9|1580|780|800|138.9|Stanford University
    9|1580|780|800|138.9|Franklin W Olin College of Engineering
    12|1570|770|800|138.2|Johns Hopkins University
    12|1570|770|800|138.2|Washington University in St Louis
    12|1570|770|800|138.2|Duke University
    12|1570|770|800|138.2|University of Pennsylvania
    12|1570|780|790|138.2|Brown University
    12|1570|770|800|138.2|Rice University
    18|1560|760|800|137.5|Northwestern University
    18|1560|780|780|137.5|Dartmouth College
    19|1555|775|780|137.1|Amherst College
    20|1550|750|800|136.8|Carnegie Mellon University
    21|1540|750|790|136.0|University of California-Berkeley
    51|1470|710|760|131.0|University of California-Los Angeles

    > If you aced the SAT, would you want to go to some no-name recognition school

    Top universities rejected more than accepted students with perfect scores (especially from some ethnic groups) to make room for the subpar applicants. There are self reported cases of those with perfect scores of 1600 in limbo wrt 7 tier1 universities and were on the waiting lists of another (he/she will be smart to start applying to some tier2 immediately), or with near perfect SAT scores on being immediately rejected wrt 7 tier1 universities (just below the 75th percentile for Berkeley) and settled for a tier2, e.g. from CollegeData.com dataset,

    ID|State|Class|Status|SAT|Inst
    AAA|MI|2022|Applied|1600|Brown
    AAA|MI|2022|Applied|1600|Cornell
    AAA|MI|2022|Applied|1600|Duke
    AAA|MI|2022|Applied|1600|Princeton
    AAA|MI|2022|Applied|1600|Stanford
    AAA|MI|2022|Applied|1600|UChicago
    AAA|MI|2022|Applied|1600|Yale
    AAA|MI|2022|Wait-Listed|1600|Caltech
    ---
    BBB|CA|2022|Denied|1520|Berkeley
    BBB|CA|2022|Denied|1520|Cornell
    BBB|CA|2022|Denied|1520|Duke
    BBB|CA|2022|Denied|1520|Harvard
    BBB|CA|2022|Denied|1520|Princeton
    BBB|CA|2022|Denied|1520|Stanford
    BBB|CA|2022|Denied|1520|UChicago
    BBB|CA|2022|Will Attend|1520|UCLA

    Replies: @Nathan, @res, @Calvin Hobbes

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai’s dataset with only SAT75.

    I wish you could find that link. It would be very helpful.

    This is the closest thing I found. It has the SAT 25-75 range for many universities.
    https://www.iecaonline.com/wp-content/uploads/2018/01/College-fall-2017-Reality-Check-SAT-25-75-college-Class-of-2020.pdf

    P.S. Some possibly relevant Wai papers: https://pdfs.semanticscholar.org/6fd7/de502146be112ef32d75f6814bea212d8455.pdf
    https://www.psychologytoday.com/files/attachments/56143/wai-americas-elite-2013.pdf
    https://www.journalofexpertise.org/articles/JoE_2018_1_1_Wai_Perina_Mar3.pdf

  142. @Jack D
    @Nathan

    Maybe so (though I don't agree) but it helps to explain how HYP etc. can claim that their 75th %ile is 1600 when only a few hundred per year get 1600 AT ONE SITTING.

    Replies: @res

    Maybe so (though I don’t agree)

    Cheating or not, it makes the numbers more difficult to interpret (e.g. estimating IQs). Though they probably consider that a feature rather than bug. I guess it just wasn’t good enough to pick the highest aggregate out of multiple attempts.

    • Replies: @Jack D
    @res

    As long as the same methodology is applied fairly to all applicants, it doesn't really make that much difference what the method is. If it makes everyone (kids feel that they are putting their best foot forward, schools get to claim 1600 as their 75th %ile, College Board gets extra revenue from people taking the test more than once) happy, I'm ok with it.

    Replies: @Nathan, @res

  143. @res
    @Jack D


    Maybe so (though I don’t agree)
     
    Cheating or not, it makes the numbers more difficult to interpret (e.g. estimating IQs). Though they probably consider that a feature rather than bug. I guess it just wasn't good enough to pick the highest aggregate out of multiple attempts.

    Replies: @Jack D

    As long as the same methodology is applied fairly to all applicants, it doesn’t really make that much difference what the method is. If it makes everyone (kids feel that they are putting their best foot forward, schools get to claim 1600 as their 75th %ile, College Board gets extra revenue from people taking the test more than once) happy, I’m ok with it.

    • Replies: @Nathan
    @Jack D

    Maybe we should let doctors and lawyers take board and bar exams until their scores add up to a pass.

    Replies: @Jack D

    , @res
    @Jack D

    Do the colleges use the same methodology for admissions as for the numbers they advertise? Would a rational admissions office consider taking the SAT twice and getting 750/800 and 800/750 (I think that's within the plausible variation range, even if not common) better than taking it once and getting 790/790? How about 700/700 followed by a 790/800 retest compared to either of those cases?

    I don't care for the arms race and lottery aspects of taking the SAT multiple times. I get that people can have a bad day (or realize they need to make an effort to prepare) and want to take the SAT again, but think both scores should be taken into account in some fashion. Especially when the later scores are lower. I am biased here because when I took the SAT I did well enough that I figured it was a fair assessment (fairly even chance of doing better or worse in a second attempt) so I chose not to take it again.

    I think picking the best subtest scores from multiple tests for admissions is ridiculous. As far as reporting the scores I am kind of with you. Marketers will be marketers--and anyone who takes them seriously deserves what they get.

    P.S. My sense is a big driver of the taking the SAT multiple times idea is giving poor scoring groups multiple lottery attempts. Someone is going to get lucky and get numbers that qualify. Another reason is probably to permit differing interpretations of the data to justify someone they want to admit for other reasons.

    P.P.S. A good reason to care about this is these shenanigans lower the effective ceiling of the test. Which is important at the level we are discussing. Especially post-1995 recentering.

    Replies: @Jack D

  144. @dux.ie
    @Nathan

    > The problem is when you add up all the other colleges that claim similar admissions stats.

    From a government source extracted from a very big dataset (forgot the link). 75th percentile SAT, SATverbal (or SATerw) and SATmath scores. Otherwise there is Wai's dataset with only SAT75. IQlike75 is separately calculated. MIT no. Princeton no. WUST no.

    Rank|SAT75|SATV75|SATM75|IQlike75|Inst
    1|1600|800|800|140.4|California Institute of Technology
    1|1600|800|800|140.4|Yale University
    1|1600|800|800|140.4|University of Chicago
    1|1600|800|800|140.4|Harvard University

    5|1590|790|800|139.7|Massachusetts Institute of Technology
    5|1590|790|800|139.7|Princeton University
    5|1590|790|800|139.7|Columbia University in the City of New York
    5|1590|790|800|139.7|Vanderbilt University
    9|1580|780|800|138.9|Harvey Mudd College
    9|1580|780|800|138.9|Stanford University
    9|1580|780|800|138.9|Franklin W Olin College of Engineering
    12|1570|770|800|138.2|Johns Hopkins University
    12|1570|770|800|138.2|Washington University in St Louis
    12|1570|770|800|138.2|Duke University
    12|1570|770|800|138.2|University of Pennsylvania
    12|1570|780|790|138.2|Brown University
    12|1570|770|800|138.2|Rice University
    18|1560|760|800|137.5|Northwestern University
    18|1560|780|780|137.5|Dartmouth College
    19|1555|775|780|137.1|Amherst College
    20|1550|750|800|136.8|Carnegie Mellon University
    21|1540|750|790|136.0|University of California-Berkeley
    51|1470|710|760|131.0|University of California-Los Angeles

    > If you aced the SAT, would you want to go to some no-name recognition school

    Top universities rejected more than accepted students with perfect scores (especially from some ethnic groups) to make room for the subpar applicants. There are self reported cases of those with perfect scores of 1600 in limbo wrt 7 tier1 universities and were on the waiting lists of another (he/she will be smart to start applying to some tier2 immediately), or with near perfect SAT scores on being immediately rejected wrt 7 tier1 universities (just below the 75th percentile for Berkeley) and settled for a tier2, e.g. from CollegeData.com dataset,

    ID|State|Class|Status|SAT|Inst
    AAA|MI|2022|Applied|1600|Brown
    AAA|MI|2022|Applied|1600|Cornell
    AAA|MI|2022|Applied|1600|Duke
    AAA|MI|2022|Applied|1600|Princeton
    AAA|MI|2022|Applied|1600|Stanford
    AAA|MI|2022|Applied|1600|UChicago
    AAA|MI|2022|Applied|1600|Yale
    AAA|MI|2022|Wait-Listed|1600|Caltech
    ---
    BBB|CA|2022|Denied|1520|Berkeley
    BBB|CA|2022|Denied|1520|Cornell
    BBB|CA|2022|Denied|1520|Duke
    BBB|CA|2022|Denied|1520|Harvard
    BBB|CA|2022|Denied|1520|Princeton
    BBB|CA|2022|Denied|1520|Stanford
    BBB|CA|2022|Denied|1520|UChicago
    BBB|CA|2022|Will Attend|1520|UCLA

    Replies: @Nathan, @res, @Calvin Hobbes

    I know of several applicants who were rejected by Harvard after getting IMO gold medals and who were subsequently Putnam fellows for MIT.

    Here’s a similar case:

    https://blog.tanyakhovanova.com/2009/07/what-does-it-take-to-get-accepted-by-harvard-or-princeton/

    • Replies: @Calvin Hobbes
    @Calvin Hobbes

    And one of those MIT undergrads rejected by Harvard, I think as an MIT sophomore undergrad, gave a Harvard grad student in Math the key idea for solving his thesis problem in algebraic geometry.
    I think their resulting joint paper is here:

    https://arxiv.org/pdf/1409.3802.pdf

    , @dux.ie
    @Calvin Hobbes

    Harvard seems to have problems with candidates with perfect SATeq score. The simple interim acceptance rate is just the ratio of Nacc/(Nacc+Nrej) excluding those applications in limbo. At the bottom of the self reported accepted list from CollegeData.com, the lowest was with SATeq=990 but with great Harvard personality (tm) !!!. Wondering how an IQ 96 student will be able to compete with IQ 140+ classmates.

    http://i68.tinypic.com/2hn7ytz.png

    Replies: @Jack D

  145. @res
    @Nathan

    Here are the corresponding pages for the 2015 subtests:
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-crit-reading-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-writing-2015.pdf
    https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-mathematics-2015.pdf

    16668 Math 800s in 2015
    9906 Critical Reading 800s in 2015 (CR is equivalent to Verbal in 2 subtest years).

    1,698,521 total test takers if you like percentages. So Math 800 is about top 1%, Verbal 800 a little above 0.5%.

    So assuming those numbers are similar to 2018, your analysis showing 10,978 800 Math Scores accounts for about 2/3 of them. That actually sounds possible to me.

    Replies: @Nathan

    Yes, my number crunching hasn’t really panned out, especially considering that Harvard/Yale/Princeton will let you just *keep taking the test until it adds up to 1600*. No wonder they’re 1,200 over what I assumed was possible. My suspicions were always directed more toward colleges that shoot up in the US News ranks for mysterious reasons,anyway. Not so much actual elite schools. Washington University in St. Louis… you know. Besides, It’s not like schools haven’t been caught doing it:

    http://www.insidehighered.com/admissions/article/2018/07/30/temple-admits-it-provided-false-rankings-data-six-more-programs

    http://www.washingtonpost.com/blogs/college-inc/post/claremont-mckenna-sat-scandal-more-at-stake-than-rankings/2012/02/07/gIQAHImVwQ_blog.html?utm_term=.d5a83db7f3ca

    http://www.insidehighered.com/news/2013/01/26/bucknell-misreported-test-scores-2006-2012

  146. @Jack D
    @res

    As long as the same methodology is applied fairly to all applicants, it doesn't really make that much difference what the method is. If it makes everyone (kids feel that they are putting their best foot forward, schools get to claim 1600 as their 75th %ile, College Board gets extra revenue from people taking the test more than once) happy, I'm ok with it.

    Replies: @Nathan, @res

    Maybe we should let doctors and lawyers take board and bar exams until their scores add up to a pass.

    • Replies: @Jack D
    @Nathan

    Not the same thing. College admissions tests are not professional qualification exams. They don't have to use SAT scores at all if they don't want to. They can take whomever they want based on whatever criteria they want to set (or none).

  147. @Calvin Hobbes
    @dux.ie

    I know of several applicants who were rejected by Harvard after getting IMO gold medals and who were subsequently Putnam fellows for MIT.

    Here’s a similar case:

    https://blog.tanyakhovanova.com/2009/07/what-does-it-take-to-get-accepted-by-harvard-or-princeton/

    Replies: @Calvin Hobbes, @dux.ie

    And one of those MIT undergrads rejected by Harvard, I think as an MIT sophomore undergrad, gave a Harvard grad student in Math the key idea for solving his thesis problem in algebraic geometry.
    I think their resulting joint paper is here:

    https://arxiv.org/pdf/1409.3802.pdf

  148. @Nathan
    @Jack D

    Maybe we should let doctors and lawyers take board and bar exams until their scores add up to a pass.

    Replies: @Jack D

    Not the same thing. College admissions tests are not professional qualification exams. They don’t have to use SAT scores at all if they don’t want to. They can take whomever they want based on whatever criteria they want to set (or none).

  149. @Jack D
    @res

    As long as the same methodology is applied fairly to all applicants, it doesn't really make that much difference what the method is. If it makes everyone (kids feel that they are putting their best foot forward, schools get to claim 1600 as their 75th %ile, College Board gets extra revenue from people taking the test more than once) happy, I'm ok with it.

    Replies: @Nathan, @res

    Do the colleges use the same methodology for admissions as for the numbers they advertise? Would a rational admissions office consider taking the SAT twice and getting 750/800 and 800/750 (I think that’s within the plausible variation range, even if not common) better than taking it once and getting 790/790? How about 700/700 followed by a 790/800 retest compared to either of those cases?

    I don’t care for the arms race and lottery aspects of taking the SAT multiple times. I get that people can have a bad day (or realize they need to make an effort to prepare) and want to take the SAT again, but think both scores should be taken into account in some fashion. Especially when the later scores are lower. I am biased here because when I took the SAT I did well enough that I figured it was a fair assessment (fairly even chance of doing better or worse in a second attempt) so I chose not to take it again.

    I think picking the best subtest scores from multiple tests for admissions is ridiculous. As far as reporting the scores I am kind of with you. Marketers will be marketers–and anyone who takes them seriously deserves what they get.

    P.S. My sense is a big driver of the taking the SAT multiple times idea is giving poor scoring groups multiple lottery attempts. Someone is going to get lucky and get numbers that qualify. Another reason is probably to permit differing interpretations of the data to justify someone they want to admit for other reasons.

    P.P.S. A good reason to care about this is these shenanigans lower the effective ceiling of the test. Which is important at the level we are discussing. Especially post-1995 recentering.

    • Replies: @Jack D
    @res

    https://blog.collegevine.com/which-colleges-superscore-the-sat/

    The main reason I suspect is not to help the students but to allow the colleges to make higher SAT claims. Since (almost) every student benefit from superscoring, no one benefits (on average).

    Replies: @res

  150. @res
    @dux.ie


    Your calculations assumed that the distributions are perfect bell curve which might not be the case.
     
    Right. That was why I explicitly said: "assuming a normal distribution." And later qualified "I suspect the distribution is fat tailed." And then backed off to the JBHE number rather than my own estimate (roughly a factor of two difference).

    How often do you discuss your assumptions so clearly?

    The College Board’s data gave the integer percentile for the given scores thus they can be calculated empirically and directly without assuming any distribution shapes but subjected to rounding off errors.
     
    Right. I linked that document above. The problem is that for blacks (the case under consideration), 650 was 98th percentile, 700 was 99th percentile, and 750 (the case of interest) was 99+. Rendering the rounding off errors much less acceptable than the normal assumption. I included the other cases to give perspective on how accurate my methodology seemed to be.

    FWIW, my guess is their percentiles are integer rounded (so 98.5 - 99.49 is 99, and > 99.5 is 99+). If anyone knows otherwise, please give a reference. That would turn your 2190 blacks scoring over 750 into 1095 which is much closer to reality (but what's a factor of 2 error between friends? not to mention the additional factor of 4 which is likely, so you are probably only a factor of 8 off for the black case. nice work. perhaps that explains why I chose to do things the other way?). Thinking about this, that's probably how Brookings got "not more than 1,000." Not some elaborate calculation involving "Other" and "No Response" like I speculated.

    The results are close to that from Brookings, especially for large Pct75, e.g. “””16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.””” compares to the re-calculated 16005, 29573, 2190 and 1627. The last 2 are with Pct75 of 1 hence large rounding off errors.
     
    I find some of your comments interesting (they would be even better if you included your data sources), but that you post that black number (2190) without apparently realizing those "large rounding off errors" make the analysis unusable for blacks make me question the validity of anything you write. I suspect the reason your numbers agree so well with Brookings is that they did the same calculations except for your error with 99+ with blacks and Brookings ignoring the other groups in their aggregate percentage.

    In addition, you only looked at one of the three Latino groups--Mexican and Puerto Rican being the other two. As far as I can tell Brookings aggregated them given the percentages they quoted (I also noted this in my comment, did you miss that?). That seems like another rather substantial error (leaving out half of all Latinos, both groups having 750 at the 99+ percentile). Perhaps you should consider either putting down the stones or moving out of your glass house?

    P.S. I went back and double checked. From the Brookings article:

    (The College Board has separate categories for “Mexican or Mexican American” and “Other Hispanic, Latino, or Latin American.” We have combined them under the term Latino.)
     
    So it looks like they ignored the poorly performing (lowest mean of the Latino groups) Puerto Ricans (about 30k people, relative to 130k Mexicans). Funny definition of Latino they have that excludes Puerto Ricans.

    Also, anyone have any idea why ETS left "No Response" out of their percentile tables? That group accounts for about the same proportions as "Other" (i.e. a significant--about 5%--proportion of >750 scorers).

    Replies: @dux.ie

    If you have deal with the ACT data you will learn to avoid assuming bell curve distributions at all time unless when there is no choice. “No child left behind”? Right, the left hand tail seems to have been compressed to shift the bottom scores upward, and reduces the ethnic score gap at the same time. The compression of the left tail seems to have broken the ACT maths distributions, they are distinctly bimodal. Try plotting out the ACT math histogram for any years. “The Asians are scoring too high in ACT math”. Right, it seems that the Asian ACT math distribution right hand tail is compressed and the ceiling score lowered, resulting in the Asian ACT math distribution to be distinctly “trimodal”, i.e. if you are lucky to grab from the ACT site to see the 2013 report which was mentioned here at unz.com before it disappeared. A scaled down by the Asian% of the effects of the third mode and lower ceiling can be seen in the national distribution right hand tail if you look hard enough.

    • Replies: @res
    @dux.ie

    That's pretty wild. Perhaps you could calculate how close the normal distribution comes to predicting the right tail? Which is what I was using it for with the SAT data.

    Hopefully my comment made clear why it's a really bad idea to base calculations on percentile data in the 99+% range. And even worse to assume 99+ = 99.

    P.S. Why do you consistently refuse to link your sources?
    Is this the 2013 ACT report you mean?
    https://www.act.org/content/dam/act/unsecured/documents/Natl-Scores-2013-National2013.pdf
    2016 version: https://www.act.org/content/dam/act/unsecured/documents/P_99_999999_N_S_N00_ACT-GCPR_National.pdf

    Though not directly relevant, this is worth a look: https://www.unz.com/isteve/synthesizing-average-sat-and-act-scores-by-state/

    , @Nathan
    @dux.ie

    That distribition is a mess.

    And doesn't everyone acknowledge that the ACT is less IQ like than the SAT?

    Replies: @dux.ie

  151. @res
    @Jack D

    Do the colleges use the same methodology for admissions as for the numbers they advertise? Would a rational admissions office consider taking the SAT twice and getting 750/800 and 800/750 (I think that's within the plausible variation range, even if not common) better than taking it once and getting 790/790? How about 700/700 followed by a 790/800 retest compared to either of those cases?

    I don't care for the arms race and lottery aspects of taking the SAT multiple times. I get that people can have a bad day (or realize they need to make an effort to prepare) and want to take the SAT again, but think both scores should be taken into account in some fashion. Especially when the later scores are lower. I am biased here because when I took the SAT I did well enough that I figured it was a fair assessment (fairly even chance of doing better or worse in a second attempt) so I chose not to take it again.

    I think picking the best subtest scores from multiple tests for admissions is ridiculous. As far as reporting the scores I am kind of with you. Marketers will be marketers--and anyone who takes them seriously deserves what they get.

    P.S. My sense is a big driver of the taking the SAT multiple times idea is giving poor scoring groups multiple lottery attempts. Someone is going to get lucky and get numbers that qualify. Another reason is probably to permit differing interpretations of the data to justify someone they want to admit for other reasons.

    P.P.S. A good reason to care about this is these shenanigans lower the effective ceiling of the test. Which is important at the level we are discussing. Especially post-1995 recentering.

    Replies: @Jack D

    https://blog.collegevine.com/which-colleges-superscore-the-sat/

    The main reason I suspect is not to help the students but to allow the colleges to make higher SAT claims. Since (almost) every student benefit from superscoring, no one benefits (on average).

    • Replies: @res
    @Jack D

    Thanks for the link. I had not realized how prevalent "superscoring" was.

    I think my earlier comment captured my issues with this. Some additions though.


    Since (almost) every student benefit from superscoring, no one benefits (on average).
     
    Two more things.

    1. I think this benefits the lucky over the skilled. Since so many take the SAT adding a significant lottery component is a problem IMHO.

    2. This especially disadvantages those who can straight up achieve a score near (or at!) the ceiling since their luck can not carry them over 800. As I said before, this lowers the effective ceiling of the test. For places like MIT and Caltech or the Ivies this makes the test even less useful for discriminating at the high end. And is a real disadvantage for those not lucky enough to be at a secondary school with high end math opportunities which do allow discrimination at higher levels.

    Perhaps I am overly cynical, but the only reason I can see for making the test less capable of discrimination at the top end is to make it easier to admit by criteria other than ability.

    P.S. Does anyone have data about typical distribution of scores over multiple administrations of the test to the same person? That would help make this discussion more concrete and clarify how much of an issue truly exists.

    Replies: @Jack D

  152. @Jack D
    @res

    https://blog.collegevine.com/which-colleges-superscore-the-sat/

    The main reason I suspect is not to help the students but to allow the colleges to make higher SAT claims. Since (almost) every student benefit from superscoring, no one benefits (on average).

    Replies: @res

    Thanks for the link. I had not realized how prevalent “superscoring” was.

    I think my earlier comment captured my issues with this. Some additions though.

    Since (almost) every student benefit from superscoring, no one benefits (on average).

    Two more things.

    1. I think this benefits the lucky over the skilled. Since so many take the SAT adding a significant lottery component is a problem IMHO.

    2. This especially disadvantages those who can straight up achieve a score near (or at!) the ceiling since their luck can not carry them over 800. As I said before, this lowers the effective ceiling of the test. For places like MIT and Caltech or the Ivies this makes the test even less useful for discriminating at the high end. And is a real disadvantage for those not lucky enough to be at a secondary school with high end math opportunities which do allow discrimination at higher levels.

    Perhaps I am overly cynical, but the only reason I can see for making the test less capable of discrimination at the top end is to make it easier to admit by criteria other than ability.

    P.S. Does anyone have data about typical distribution of scores over multiple administrations of the test to the same person? That would help make this discussion more concrete and clarify how much of an issue truly exists.

    • Agree: Nathan
    • Replies: @Jack D
    @res

    From what I understand of MIT's philosophy, given the narrow score range that they deal in (where 770M is the 75th %ile) they don't view the SAT as a useful tool for separating the wheat from the chaff - it's merely a minimum gatekeeping device to give them some idea that you will be able to keep up with the class work. Once you are over the threshold they don't see a 770 as being "better" than an 800, because frankly it isn't - it's more or less a matter of luck if you have 0 or 1 wrong on the math section. Once you are "qualified" on the basis of your SATs they look at all the other stuff to make their final decision. They will reject boring Asian nerd 800s all day every day in favor of 770s that they like better for some reason. They are getting 15 applications for each one they accept and can afford to be picky (Stanford gets 23).

  153. @dux.ie
    @res

    If you have deal with the ACT data you will learn to avoid assuming bell curve distributions at all time unless when there is no choice. "No child left behind"? Right, the left hand tail seems to have been compressed to shift the bottom scores upward, and reduces the ethnic score gap at the same time. The compression of the left tail seems to have broken the ACT maths distributions, they are distinctly bimodal. Try plotting out the ACT math histogram for any years. "The Asians are scoring too high in ACT math". Right, it seems that the Asian ACT math distribution right hand tail is compressed and the ceiling score lowered, resulting in the Asian ACT math distribution to be distinctly "trimodal", i.e. if you are lucky to grab from the ACT site to see the 2013 report which was mentioned here at unz.com before it disappeared. A scaled down by the Asian% of the effects of the third mode and lower ceiling can be seen in the national distribution right hand tail if you look hard enough.

    http://i63.tinypic.com/1z5lmj8.png

    Replies: @res, @Nathan

    That’s pretty wild. Perhaps you could calculate how close the normal distribution comes to predicting the right tail? Which is what I was using it for with the SAT data.

    Hopefully my comment made clear why it’s a really bad idea to base calculations on percentile data in the 99+% range. And even worse to assume 99+ = 99.

    P.S. Why do you consistently refuse to link your sources?
    Is this the 2013 ACT report you mean?
    https://www.act.org/content/dam/act/unsecured/documents/Natl-Scores-2013-National2013.pdf
    2016 version: https://www.act.org/content/dam/act/unsecured/documents/P_99_999999_N_S_N00_ACT-GCPR_National.pdf

    Though not directly relevant, this is worth a look: https://www.unz.com/isteve/synthesizing-average-sat-and-act-scores-by-state/

  154. @res
    @Jack D

    Thanks for the link. I had not realized how prevalent "superscoring" was.

    I think my earlier comment captured my issues with this. Some additions though.


    Since (almost) every student benefit from superscoring, no one benefits (on average).
     
    Two more things.

    1. I think this benefits the lucky over the skilled. Since so many take the SAT adding a significant lottery component is a problem IMHO.

    2. This especially disadvantages those who can straight up achieve a score near (or at!) the ceiling since their luck can not carry them over 800. As I said before, this lowers the effective ceiling of the test. For places like MIT and Caltech or the Ivies this makes the test even less useful for discriminating at the high end. And is a real disadvantage for those not lucky enough to be at a secondary school with high end math opportunities which do allow discrimination at higher levels.

    Perhaps I am overly cynical, but the only reason I can see for making the test less capable of discrimination at the top end is to make it easier to admit by criteria other than ability.

    P.S. Does anyone have data about typical distribution of scores over multiple administrations of the test to the same person? That would help make this discussion more concrete and clarify how much of an issue truly exists.

    Replies: @Jack D

    From what I understand of MIT’s philosophy, given the narrow score range that they deal in (where 770M is the 75th %ile) they don’t view the SAT as a useful tool for separating the wheat from the chaff – it’s merely a minimum gatekeeping device to give them some idea that you will be able to keep up with the class work. Once you are over the threshold they don’t see a 770 as being “better” than an 800, because frankly it isn’t – it’s more or less a matter of luck if you have 0 or 1 wrong on the math section. Once you are “qualified” on the basis of your SATs they look at all the other stuff to make their final decision. They will reject boring Asian nerd 800s all day every day in favor of 770s that they like better for some reason. They are getting 15 applications for each one they accept and can afford to be picky (Stanford gets 23).

  155. @Calvin Hobbes
    @dux.ie

    I know of several applicants who were rejected by Harvard after getting IMO gold medals and who were subsequently Putnam fellows for MIT.

    Here’s a similar case:

    https://blog.tanyakhovanova.com/2009/07/what-does-it-take-to-get-accepted-by-harvard-or-princeton/

    Replies: @Calvin Hobbes, @dux.ie

    Harvard seems to have problems with candidates with perfect SATeq score. The simple interim acceptance rate is just the ratio of Nacc/(Nacc+Nrej) excluding those applications in limbo. At the bottom of the self reported accepted list from CollegeData.com, the lowest was with SATeq=990 but with great Harvard personality ™ !!!. Wondering how an IQ 96 student will be able to compete with IQ 140+ classmates.

    • Replies: @Jack D
    @dux.ie

    Self-reported data is pretty worthless. #1 you don't have a random sample and #2 people lie.

    Replies: @dux.ie

  156. @Edward
    @JimB


    Since the renorming of the SAT, 750 corresponds to 99%tile. It’s an understatement to say math PhDs are in the top 1% (2.5 sigma). I’m guessing from personal experience that productive math grads are more like NFL players, from the 5 sigma level of talent. At that level, you are selecting from tiny sub populations of Europeans, East Asians, and Indians, only.
     
    5-sigma is too high. The Fields Medalist Richard Borcherds has a performance IQ (a measure of nonverbal ability) of 147 (top 0.09%) and a full-scale IQ of 137 (top 0.7%).

    https://www.tandfonline.com/doi/abs/10.1080/13554799908402743

    Performance IQ is only a proxy for mathematical ability, but if his nonverbal IQ is a bit over 3 SD above the mean, his mathematical ability is probably only going to be ~4SD above the mean, at most.

    Borcherds is probably on the low end of the ability distribution of Fields Medalists (Terence Tao reached the ceiling on the Stanford-Binet, with an IQ of 175). Even so, you're still right: it shouldn't be surprising that the only minority groups we see with good representation in math departments are Jewish-Americans (avg IQ 110-115, with their mathematical abilities probably even higher than this), Indian-Americans (avg IQ ~110) and East Asian-Americans (avg IQ ~106, with mathematical abilities probably higher than this).

    Replies: @Peter Johnson

    One important correction to your comments — as David Reich has shown, Indians (from India) are best not viewed as a single ethnic group but rather as a collection of quite distinct ethnic groups. Some of these ethnic groups have high average IQ and others do not. The country is unusual in that respect.

    • Replies: @Edward
    @Peter Johnson

    Quite right. I was referring to the average of Indian-Americans, however, not Indians in India.

  157. @Peter Johnson
    @Edward

    One important correction to your comments -- as David Reich has shown, Indians (from India) are best not viewed as a single ethnic group but rather as a collection of quite distinct ethnic groups. Some of these ethnic groups have high average IQ and others do not. The country is unusual in that respect.

    Replies: @Edward

    Quite right. I was referring to the average of Indian-Americans, however, not Indians in India.

  158. @dux.ie
    @res

    If you have deal with the ACT data you will learn to avoid assuming bell curve distributions at all time unless when there is no choice. "No child left behind"? Right, the left hand tail seems to have been compressed to shift the bottom scores upward, and reduces the ethnic score gap at the same time. The compression of the left tail seems to have broken the ACT maths distributions, they are distinctly bimodal. Try plotting out the ACT math histogram for any years. "The Asians are scoring too high in ACT math". Right, it seems that the Asian ACT math distribution right hand tail is compressed and the ceiling score lowered, resulting in the Asian ACT math distribution to be distinctly "trimodal", i.e. if you are lucky to grab from the ACT site to see the 2013 report which was mentioned here at unz.com before it disappeared. A scaled down by the Asian% of the effects of the third mode and lower ceiling can be seen in the national distribution right hand tail if you look hard enough.

    http://i63.tinypic.com/1z5lmj8.png

    Replies: @res, @Nathan

    That distribition is a mess.

    And doesn’t everyone acknowledge that the ACT is less IQ like than the SAT?

    • Replies: @dux.ie
    @Nathan

    The collection of social economic data is like making sausages, you do not want to know the details. What is nice and perfect on the outside has hidden digital iterative surgery procedure performed on them out of sights. A revealing report from the Bank of International Settlement BIS (the central bank of national central banks, one level up from the NewYork Federal Reserve) showed the messy details of slicing the humps and filling the dips,

    https://www.bis.org/publ/work184.pdf

    The real US income distribution is bimodal, it might be better correlated to the bimodal ACT Maths distribution than the seemingly nice artificial single modal bell shape SAT distribution.

    SAT and ACT scores are not absolute metrics and they are regularly re-calibrated with the national cognitive percentile. The SAT and ACT results are differently bias because of the demographics of the test takers. SAT's HQ is in NewYork and it tended to attract more city students. ACT's HQ is in Iowa and it tended to attract more rural and suburban students.

    https://www.edweek.org/ew/articles/2017/05/24/in-race-for-test-takers-act-outscores-sat--for.html
    """the SAT is strongest on the coasts and the ACT in the Midwest and the South."""

    Some universities have hidden preferences for ACT takers where the unstated SAT equivalent of the advertised nominal minimum ACT scores to be less than the explicitly advertised SAT scores to give preference to the rural or other student groups.

    The smart Asian American students tended to be in cities and some have perceived that to have tougher competitions in Maths test as the SAT Math score is dependent on the test taker's percentile among the SAT Math test takers. The NAEP has shown that (for Grade8?) on average the rural and suburban students are smarter than those from the cities. Thus taking only the SAT results represents a one-sided bias view over-emphasizing the percentage advantage of the Asian American results, e.g. the Brookings' SAT 750+ ratio for White vs Asian of 33/60.

    If you believe the authenticity of the accidentally released ACT reports from the ACT site detailing the ethnic groups performances whose urls were mentioned here previously on unz.com but had since disappeared, the ACT 34+ (eqivalent to about SAT 755+) for White+/Asian ratio is about 73/22, White+ is imputed to be including the much smaller "other" group as explictly White data are not available but substracting the national data from the various given ethnic groups (i.e. Black, Hispanic, Asian) data gives the "White+" group. Strong arming the SAT 750+ and ACT 34+ from two close (2 years diff) but different testing years gives the rough ratio of 47/42.

  159. @dux.ie
    @Calvin Hobbes

    Harvard seems to have problems with candidates with perfect SATeq score. The simple interim acceptance rate is just the ratio of Nacc/(Nacc+Nrej) excluding those applications in limbo. At the bottom of the self reported accepted list from CollegeData.com, the lowest was with SATeq=990 but with great Harvard personality (tm) !!!. Wondering how an IQ 96 student will be able to compete with IQ 140+ classmates.

    http://i68.tinypic.com/2hn7ytz.png

    Replies: @Jack D

    Self-reported data is pretty worthless. #1 you don’t have a random sample and #2 people lie.

    • Replies: @dux.ie
    @Jack D

    I doubt people will fake results that they had been rejected by any universities. The ID is anonymised and there is no advantage for boasting. The cross-check and validation are to compare the self reported accepted average SAT scores against Wai's average SAT dataset and the ratio ranges from 0.95 to 1.13 . Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher. Nevertheless the sample SATavg for tier1 universities like Caltech, MIT, UChicago, Columbia, Yale, Harvard are actually less than that in Wai's data. It does not look like the self reported applicants are boasting about their SAT scores or being accepted into tier1 universities.

    http://i67.tinypic.com/35k4ju8.png

    Unfounded criticism is pretty worthless.

    Replies: @dux.ie, @res

  160. Has anyone ever profiled NYT journalists, taking into account the section of the paper that they work in, and the length and prominence of their stories? For example, how many African-American journalists write headline stories in the business section?

  161. @Jack D
    @dux.ie

    Self-reported data is pretty worthless. #1 you don't have a random sample and #2 people lie.

    Replies: @dux.ie

    I doubt people will fake results that they had been rejected by any universities. The ID is anonymised and there is no advantage for boasting. The cross-check and validation are to compare the self reported accepted average SAT scores against Wai’s average SAT dataset and the ratio ranges from 0.95 to 1.13 . Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher. Nevertheless the sample SATavg for tier1 universities like Caltech, MIT, UChicago, Columbia, Yale, Harvard are actually less than that in Wai’s data. It does not look like the self reported applicants are boasting about their SAT scores or being accepted into tier1 universities.

    Unfounded criticism is pretty worthless.

    • Replies: @dux.ie
    @dux.ie

    The 3 institutions with ratio above 1.10 are CPObispo, UTAustin and Rutgers. I dont know much about CalPoly. In Texas the state guarantees the top 10% performers of high schools irrespective of their standard test scores places in the local universities, the universities are only freely allowed to select small fractions of the incoming students thus the competitions for those places should be intense hence the higher SATavg in the sample. From the sample Rutgers seemed to accept ALL applicants with perfect SATeq scores. Thus Rutgers tended to be a favourite PlanB universities for those with near perfect SATeq scores and hence the higher SATavg.

    Replies: @res

    , @res
    @dux.ie


    Unfounded criticism is pretty worthless.
     
    Jack D's criticism was neither unfounded nor worthless.

    That said, I disagree with his conclusion (your data is pretty worthless). The data does have some use (especially with the cross-checks you do) given that it is the best available.

    I would be worried about the selection effects. Which you at least acknowledge: "Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher."

    I think it would make more sense to express the difference with Wai's data as a Z score (SD units) rather than a ratio.

    P.S. I would really like to know more about the person who was accepted to Berkeley with an 830 SAT! Or UCLA with a 690, or Harvard with a 990. Or MIT with a 800, or Caltech with a 1000.
    Assuming those aren't lies.

    Replies: @dux.ie

  162. @dux.ie
    @Jack D

    I doubt people will fake results that they had been rejected by any universities. The ID is anonymised and there is no advantage for boasting. The cross-check and validation are to compare the self reported accepted average SAT scores against Wai's average SAT dataset and the ratio ranges from 0.95 to 1.13 . Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher. Nevertheless the sample SATavg for tier1 universities like Caltech, MIT, UChicago, Columbia, Yale, Harvard are actually less than that in Wai's data. It does not look like the self reported applicants are boasting about their SAT scores or being accepted into tier1 universities.

    http://i67.tinypic.com/35k4ju8.png

    Unfounded criticism is pretty worthless.

    Replies: @dux.ie, @res

    The 3 institutions with ratio above 1.10 are CPObispo, UTAustin and Rutgers. I dont know much about CalPoly. In Texas the state guarantees the top 10% performers of high schools irrespective of their standard test scores places in the local universities, the universities are only freely allowed to select small fractions of the incoming students thus the competitions for those places should be intense hence the higher SATavg in the sample. From the sample Rutgers seemed to accept ALL applicants with perfect SATeq scores. Thus Rutgers tended to be a favourite PlanB universities for those with near perfect SATeq scores and hence the higher SATavg.

    • Replies: @res
    @dux.ie

    Is your data "Accepted" as it says rather than "Matriculated" (which is what Wai's data captures, right?)? Like you say for Rutgers, I think it is easy to posit all of those three as common safety schools which would explain why the "Accepted" number is higher.

    Replies: @res, @dux.ie

  163. @Nathan
    @dux.ie

    That distribition is a mess.

    And doesn't everyone acknowledge that the ACT is less IQ like than the SAT?

    Replies: @dux.ie

    The collection of social economic data is like making sausages, you do not want to know the details. What is nice and perfect on the outside has hidden digital iterative surgery procedure performed on them out of sights. A revealing report from the Bank of International Settlement BIS (the central bank of national central banks, one level up from the NewYork Federal Reserve) showed the messy details of slicing the humps and filling the dips,

    https://www.bis.org/publ/work184.pdf

    The real US income distribution is bimodal, it might be better correlated to the bimodal ACT Maths distribution than the seemingly nice artificial single modal bell shape SAT distribution.

    SAT and ACT scores are not absolute metrics and they are regularly re-calibrated with the national cognitive percentile. The SAT and ACT results are differently bias because of the demographics of the test takers. SAT’s HQ is in NewYork and it tended to attract more city students. ACT’s HQ is in Iowa and it tended to attract more rural and suburban students.

    https://www.edweek.org/ew/articles/2017/05/24/in-race-for-test-takers-act-outscores-sat–for.html
    “””the SAT is strongest on the coasts and the ACT in the Midwest and the South.”””

    Some universities have hidden preferences for ACT takers where the unstated SAT equivalent of the advertised nominal minimum ACT scores to be less than the explicitly advertised SAT scores to give preference to the rural or other student groups.

    The smart Asian American students tended to be in cities and some have perceived that to have tougher competitions in Maths test as the SAT Math score is dependent on the test taker’s percentile among the SAT Math test takers. The NAEP has shown that (for Grade8?) on average the rural and suburban students are smarter than those from the cities. Thus taking only the SAT results represents a one-sided bias view over-emphasizing the percentage advantage of the Asian American results, e.g. the Brookings’ SAT 750+ ratio for White vs Asian of 33/60.

    If you believe the authenticity of the accidentally released ACT reports from the ACT site detailing the ethnic groups performances whose urls were mentioned here previously on unz.com but had since disappeared, the ACT 34+ (eqivalent to about SAT 755+) for White+/Asian ratio is about 73/22, White+ is imputed to be including the much smaller “other” group as explictly White data are not available but substracting the national data from the various given ethnic groups (i.e. Black, Hispanic, Asian) data gives the “White+” group. Strong arming the SAT 750+ and ACT 34+ from two close (2 years diff) but different testing years gives the rough ratio of 47/42.

  164. @dux.ie
    @Jack D

    I doubt people will fake results that they had been rejected by any universities. The ID is anonymised and there is no advantage for boasting. The cross-check and validation are to compare the self reported accepted average SAT scores against Wai's average SAT dataset and the ratio ranges from 0.95 to 1.13 . Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher. Nevertheless the sample SATavg for tier1 universities like Caltech, MIT, UChicago, Columbia, Yale, Harvard are actually less than that in Wai's data. It does not look like the self reported applicants are boasting about their SAT scores or being accepted into tier1 universities.

    http://i67.tinypic.com/35k4ju8.png

    Unfounded criticism is pretty worthless.

    Replies: @dux.ie, @res

    Unfounded criticism is pretty worthless.

    Jack D’s criticism was neither unfounded nor worthless.

    That said, I disagree with his conclusion (your data is pretty worthless). The data does have some use (especially with the cross-checks you do) given that it is the best available.

    I would be worried about the selection effects. Which you at least acknowledge: “Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher.”

    I think it would make more sense to express the difference with Wai’s data as a Z score (SD units) rather than a ratio.

    P.S. I would really like to know more about the person who was accepted to Berkeley with an 830 SAT! Or UCLA with a 690, or Harvard with a 990. Or MIT with a 800, or Caltech with a 1000.
    Assuming those aren’t lies.

    • Replies: @dux.ie
    @res

    > Assuming those aren’t lies.

    One late admittance case for a high school shooting survivor and student leader that received much publicity, the SAT score 1270 involved was not that low but well below the Harvard SAT 25th percentile at 1470 (SATwai25=1410), and the Harvard special invite for White rural students at 1310, below the average score for Pol Sci he wanted to major in, and was previously denied to UCLA, UCSD and UCIrvine,

    https://www.businessinsider.com.au/david-hogg-attend-harvard-university-1270-sat-score-reactions-2018-12

    http://i68.tinypic.com/35hr1uw.png

    The one allegely accepted to Harvard with SATeq 990 the sample data that I have showed that with that SAT score from NY she ONLY applied to Harvard which accepted and Princeton which denied. She must be full of self-confidence, no SUNY, no CUNY, no PlanB. She must be one of those special invitee, from the Harvard court case SFFA document, Harvard special invitees lowest cut-off is SAT 1100.

    Replies: @Steve Sailer, @res

  165. @dux.ie
    @dux.ie

    The 3 institutions with ratio above 1.10 are CPObispo, UTAustin and Rutgers. I dont know much about CalPoly. In Texas the state guarantees the top 10% performers of high schools irrespective of their standard test scores places in the local universities, the universities are only freely allowed to select small fractions of the incoming students thus the competitions for those places should be intense hence the higher SATavg in the sample. From the sample Rutgers seemed to accept ALL applicants with perfect SATeq scores. Thus Rutgers tended to be a favourite PlanB universities for those with near perfect SATeq scores and hence the higher SATavg.

    Replies: @res

    Is your data “Accepted” as it says rather than “Matriculated” (which is what Wai’s data captures, right?)? Like you say for Rutgers, I think it is easy to posit all of those three as common safety schools which would explain why the “Accepted” number is higher.

    • Replies: @res
    @res

    Given this safety school idea, if you have access to the yield rate for colleges that might be useful data to incorporate in your analysis.

    , @dux.ie
    @res

    "Accepted" is the term they used, it is more like the university made an offer. These applicants are like the "swinging voters" shopping for universities. The applicants submitted their data to judge their chances of being accepted. Once the applicants received offers from the universities they wanted I wonder if they would bother to update the status further. I am only interested in the Nacc/(Nacc+Nrej) ratio. Other metrics like yield, etc, are not available, few bother to report their "Will Attend" status.

    The years I covered were also "eventful" with dominant pertrubation which might have large effects for a few years. Unless there were other events that I am not aware of, year 2014 (class 2018) was the time where the publicity culminating in a student with perfect test score sued Harvard for rejecting him. The case did not seem to impact much the interests (number of self reported applications) to Harvard, but the numbers for Berkeley and UCLA almost doubled and slowly declined over the years and those were not from California but mostly from mid-west and east coast. Because of the distance normally those from east-coast are not interested in Berkeley and UCLA. Dispite the alleged large number of Asian Americans in NY Stuyvesant HS, I have two years data of their university destinations, there was only ONE that went to Berkeley and from his name he most probably has Russian ancestry. And high tide lifts all boats, there were significant but lesser surge in some other university applications. These I would speculate as the PlanB applications since the number to Harvard did not drop. %incr is the % increase from year 2013 to 2014, %AsianAm is the demographic pop %, impact is the normalized effect = %incr/%AsianAm,

    http://i65.tinypic.com/2ewd5og.png

    Berkeley
    State %incr %AsianAm Impact
    Total 68.6 5.6 12.26
    CA 71.1 14.9 4.77
    NonCA 63.2 5.6 11.28

    UCLA
    Total 74.1 5.6 13.23
    CA 67.8 14.9 4.55
    NonCA 95.5 5.6 17.05

    Harvard
    Total 13.0 5.6 2.33
    CA 0.0 14.9 0.00
    NonCA 18.8 5.6 3.35

    For Berkeley and UCLA the impact from California was small compared to that of outside CA. For Harvard the court cases did not seem to have stirred up anythin in CA. The Chinese Californians seemed to have used to the Harvard practice and submitted their applications anyway together with many other planB universities like UPenn and Cornell whose number surged up a lot.

    Replies: @res

  166. @res
    @dux.ie

    Is your data "Accepted" as it says rather than "Matriculated" (which is what Wai's data captures, right?)? Like you say for Rutgers, I think it is easy to posit all of those three as common safety schools which would explain why the "Accepted" number is higher.

    Replies: @res, @dux.ie

    Given this safety school idea, if you have access to the yield rate for colleges that might be useful data to incorporate in your analysis.

  167. @res
    @dux.ie

    Is your data "Accepted" as it says rather than "Matriculated" (which is what Wai's data captures, right?)? Like you say for Rutgers, I think it is easy to posit all of those three as common safety schools which would explain why the "Accepted" number is higher.

    Replies: @res, @dux.ie

    “Accepted” is the term they used, it is more like the university made an offer. These applicants are like the “swinging voters” shopping for universities. The applicants submitted their data to judge their chances of being accepted. Once the applicants received offers from the universities they wanted I wonder if they would bother to update the status further. I am only interested in the Nacc/(Nacc+Nrej) ratio. Other metrics like yield, etc, are not available, few bother to report their “Will Attend” status.

    The years I covered were also “eventful” with dominant pertrubation which might have large effects for a few years. Unless there were other events that I am not aware of, year 2014 (class 2018) was the time where the publicity culminating in a student with perfect test score sued Harvard for rejecting him. The case did not seem to impact much the interests (number of self reported applications) to Harvard, but the numbers for Berkeley and UCLA almost doubled and slowly declined over the years and those were not from California but mostly from mid-west and east coast. Because of the distance normally those from east-coast are not interested in Berkeley and UCLA. Dispite the alleged large number of Asian Americans in NY Stuyvesant HS, I have two years data of their university destinations, there was only ONE that went to Berkeley and from his name he most probably has Russian ancestry. And high tide lifts all boats, there were significant but lesser surge in some other university applications. These I would speculate as the PlanB applications since the number to Harvard did not drop. %incr is the % increase from year 2013 to 2014, %AsianAm is the demographic pop %, impact is the normalized effect = %incr/%AsianAm,

    Berkeley
    State %incr %AsianAm Impact
    Total 68.6 5.6 12.26
    CA 71.1 14.9 4.77
    NonCA 63.2 5.6 11.28

    UCLA
    Total 74.1 5.6 13.23
    CA 67.8 14.9 4.55
    NonCA 95.5 5.6 17.05

    Harvard
    Total 13.0 5.6 2.33
    CA 0.0 14.9 0.00
    NonCA 18.8 5.6 3.35

    For Berkeley and UCLA the impact from California was small compared to that of outside CA. For Harvard the court cases did not seem to have stirred up anythin in CA. The Chinese Californians seemed to have used to the Harvard practice and submitted their applications anyway together with many other planB universities like UPenn and Cornell whose number surged up a lot.

    • Replies: @res
    @dux.ie

    That's "Accepted" all right. There are three numbers that matter: matriculated (enrolled), accepted, rejected. The number who applied should be (accepted + rejected) though there are some complexities like wait listing. Yield rate is matriculated / accepted. Acceptance rate is accepted / (accepted + rejected).

    This site gives what they call a selectivity index: https://www.collegeraptor.com/college-rankings/details/Selectivity


    The selectivity index is a rating system that looks at the percentage of accepted students who enroll (yield rate) divided by the overall acceptance rate. It’s a measurement of both the selectivity and desirability of the college or university.
     
    They also give acceptance rate: https://www.collegeraptor.com/college-rankings/details/AcceptanceRate

    So in theory you can calculate yield rate from their data.

    The also have median est. SAT https://www.collegeraptor.com/college-rankings/details/MedianSAT

    Here is how they describe their data output:

    School data from 2016 - 2017 (top 300 results shown) for schools with at least 100 undergraduate students
     
    To work around the top 300 display limit you can choose smaller groupings (e.g. conference, state, region). So in theory you should be able to extract all their data from the public web interface (though it would be much easier if they would give out data directly in CSV or equivalent).

    If you are interested in a dry run with this data, maybe try the Ivy Group Conference. Here is the selectivity index page for that group: https://www.collegeraptor.com/college-rankings/details/Selectivity/Conference/Ivy%20Group

    and acceptance rate: https://www.collegeraptor.com/college-rankings/details/AcceptanceRate/Conference/Ivy%20Group

    Using that data (sorting by school name to make data entry easier) I get the following Ivy Group yield rates. I think these yield rates (just multiply the other two) are plausible (notice Harvard is the highest).

    Name AcceptRate SelectInd YieldRate
    Barnard 15.4% 3.3 50.8%
    Brown 8.5% 6.9 58.7%
    Columbia 6.6% 9.3 61.4%
    Cornell 12.7% 4.4 55.9%
    Dartmouth 10.5% 5.6 58.8%
    Harvard 5.2% 16.1 83.7%
    Princeton 6.4% 10.2 65.3%
    Penn 9.3% 7 65.1%
    Yale 6.9% 10 69.0%

    For comparison, here is CalPolySLO which I think validates my safety school hypothesis:
    CalPolySLO 34.60% 0.9 31.1%

    If you have the patience to do the rest of the data entry I think these numbers would be a good addition to your analysis.
  168. @res
    @dux.ie


    Unfounded criticism is pretty worthless.
     
    Jack D's criticism was neither unfounded nor worthless.

    That said, I disagree with his conclusion (your data is pretty worthless). The data does have some use (especially with the cross-checks you do) given that it is the best available.

    I would be worried about the selection effects. Which you at least acknowledge: "Given that the applicants interested in such data tended to be better qualified it is expected the results might be slightly higher."

    I think it would make more sense to express the difference with Wai's data as a Z score (SD units) rather than a ratio.

    P.S. I would really like to know more about the person who was accepted to Berkeley with an 830 SAT! Or UCLA with a 690, or Harvard with a 990. Or MIT with a 800, or Caltech with a 1000.
    Assuming those aren't lies.

    Replies: @dux.ie

    > Assuming those aren’t lies.

    One late admittance case for a high school shooting survivor and student leader that received much publicity, the SAT score 1270 involved was not that low but well below the Harvard SAT 25th percentile at 1470 (SATwai25=1410), and the Harvard special invite for White rural students at 1310, below the average score for Pol Sci he wanted to major in, and was previously denied to UCLA, UCSD and UCIrvine,

    https://www.businessinsider.com.au/david-hogg-attend-harvard-university-1270-sat-score-reactions-2018-12

    The one allegely accepted to Harvard with SATeq 990 the sample data that I have showed that with that SAT score from NY she ONLY applied to Harvard which accepted and Princeton which denied. She must be full of self-confidence, no SUNY, no CUNY, no PlanB. She must be one of those special invitee, from the Harvard court case SFFA document, Harvard special invitees lowest cut-off is SAT 1100.

    • Replies: @Steve Sailer
    @dux.ie

    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?

    Replies: @res, @dux.ie

    , @res
    @dux.ie

    That Harvard graphic is very interesting. I think it also provides a good test case for the validity of your data. My guess would be the self reports are biased towards people who got a high SAT score and were denied admission somewhere then found that site. Not sure how big an effect that would be, but it might matter.

    That said, I think the acceptance rate non-monotonicity is better explained by applicant self selection. Presumably people scoring 1400 (or lower!) and applying to Harvard are more likely to have an in (legacy, special talent, race, ...) and will be admitted at a higher frequency than the average person with that score. On the other hand there will be many 1500 scorers who think they have a chance because of that alone (or with semi-competitive rest of the application). They will be rejected at a higher rate.

    So if that hypothesis is correct then your data is correctly reflecting reality (roughly anyway), but one should NOT conclude that the acceptance rates at a given SAT are reflective of what a random person with that score would experience.

  169. @dux.ie
    @res

    > Assuming those aren’t lies.

    One late admittance case for a high school shooting survivor and student leader that received much publicity, the SAT score 1270 involved was not that low but well below the Harvard SAT 25th percentile at 1470 (SATwai25=1410), and the Harvard special invite for White rural students at 1310, below the average score for Pol Sci he wanted to major in, and was previously denied to UCLA, UCSD and UCIrvine,

    https://www.businessinsider.com.au/david-hogg-attend-harvard-university-1270-sat-score-reactions-2018-12

    http://i68.tinypic.com/35hr1uw.png

    The one allegely accepted to Harvard with SATeq 990 the sample data that I have showed that with that SAT score from NY she ONLY applied to Harvard which accepted and Princeton which denied. She must be full of self-confidence, no SUNY, no CUNY, no PlanB. She must be one of those special invitee, from the Harvard court case SFFA document, Harvard special invitees lowest cut-off is SAT 1100.

    Replies: @Steve Sailer, @res

    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?

    • Replies: @res
    @Steve Sailer


    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?
     
    That's an interesting question. I would add "The child of a donor of over $X?"

    One relevant point is the biggest boost is hard to detect unless the starting point is low. I am defining boost as difference between necessary SAT absent/present the special characteristic holding all other data constant.

    My thought would be Obama daughter >> non-donor billionaire. Not sure how much a big donation could move that. What do you think? Does it depend on the billionaire? For example, crime-thinking Peter Thiel?
    , @dux.ie
    @Steve Sailer

    We will never know what sort of arm twisting happened behind closed doors. I am not implying anything but I noted that the new Harvard President was special advisor in Obama's special committee on Historically Black Colleges and Universities HBCU. Harvard, Horvard, Howard ... May be JKR's Hogwarts was modelled from that.

    When there was news that nannies were paid $100K pa many considered that as excessive. I raised the case of K whom high school counsellor allegely said that his academic performance was so so and he had no chance of getting into Harvard. It was alleged that K's father had to donate $200K pa for 10 years to get him into Harvard. If K's father is as smart as K who hired a special tiger nanny for his daughter, lets say for 10 years, he might have saved $1 million from that. As the collateral gain, K's daughter had already pulled off a spectacular priceless diplomatic coup with a foreign VIP. Look at the way she conduct herself now,

    https://www.youtube.com/watch?v=RoSFgo8_6Ag

    She might save K with future headache as she wanted to go to Penn State, the people's university appropriate for her constituents. I wonder why not UPenn like grand-daddy.

  170. @dux.ie
    @res

    "Accepted" is the term they used, it is more like the university made an offer. These applicants are like the "swinging voters" shopping for universities. The applicants submitted their data to judge their chances of being accepted. Once the applicants received offers from the universities they wanted I wonder if they would bother to update the status further. I am only interested in the Nacc/(Nacc+Nrej) ratio. Other metrics like yield, etc, are not available, few bother to report their "Will Attend" status.

    The years I covered were also "eventful" with dominant pertrubation which might have large effects for a few years. Unless there were other events that I am not aware of, year 2014 (class 2018) was the time where the publicity culminating in a student with perfect test score sued Harvard for rejecting him. The case did not seem to impact much the interests (number of self reported applications) to Harvard, but the numbers for Berkeley and UCLA almost doubled and slowly declined over the years and those were not from California but mostly from mid-west and east coast. Because of the distance normally those from east-coast are not interested in Berkeley and UCLA. Dispite the alleged large number of Asian Americans in NY Stuyvesant HS, I have two years data of their university destinations, there was only ONE that went to Berkeley and from his name he most probably has Russian ancestry. And high tide lifts all boats, there were significant but lesser surge in some other university applications. These I would speculate as the PlanB applications since the number to Harvard did not drop. %incr is the % increase from year 2013 to 2014, %AsianAm is the demographic pop %, impact is the normalized effect = %incr/%AsianAm,

    http://i65.tinypic.com/2ewd5og.png

    Berkeley
    State %incr %AsianAm Impact
    Total 68.6 5.6 12.26
    CA 71.1 14.9 4.77
    NonCA 63.2 5.6 11.28

    UCLA
    Total 74.1 5.6 13.23
    CA 67.8 14.9 4.55
    NonCA 95.5 5.6 17.05

    Harvard
    Total 13.0 5.6 2.33
    CA 0.0 14.9 0.00
    NonCA 18.8 5.6 3.35

    For Berkeley and UCLA the impact from California was small compared to that of outside CA. For Harvard the court cases did not seem to have stirred up anythin in CA. The Chinese Californians seemed to have used to the Harvard practice and submitted their applications anyway together with many other planB universities like UPenn and Cornell whose number surged up a lot.

    Replies: @res

    That’s “Accepted” all right. There are three numbers that matter: matriculated (enrolled), accepted, rejected. The number who applied should be (accepted + rejected) though there are some complexities like wait listing. Yield rate is matriculated / accepted. Acceptance rate is accepted / (accepted + rejected).

    This site gives what they call a selectivity index: https://www.collegeraptor.com/college-rankings/details/Selectivity

    The selectivity index is a rating system that looks at the percentage of accepted students who enroll (yield rate) divided by the overall acceptance rate. It’s a measurement of both the selectivity and desirability of the college or university.

    They also give acceptance rate: https://www.collegeraptor.com/college-rankings/details/AcceptanceRate

    So in theory you can calculate yield rate from their data.

    The also have median est. SAT https://www.collegeraptor.com/college-rankings/details/MedianSAT

    Here is how they describe their data output:

    School data from 2016 – 2017 (top 300 results shown) for schools with at least 100 undergraduate students

    To work around the top 300 display limit you can choose smaller groupings (e.g. conference, state, region). So in theory you should be able to extract all their data from the public web interface (though it would be much easier if they would give out data directly in CSV or equivalent).

    If you are interested in a dry run with this data, maybe try the Ivy Group Conference. Here is the selectivity index page for that group: https://www.collegeraptor.com/college-rankings/details/Selectivity/Conference/Ivy%20Group

    and acceptance rate: https://www.collegeraptor.com/college-rankings/details/AcceptanceRate/Conference/Ivy%20Group

    Using that data (sorting by school name to make data entry easier) I get the following Ivy Group yield rates. I think these yield rates (just multiply the other two) are plausible (notice Harvard is the highest).

    Name AcceptRate SelectInd YieldRate
    Barnard 15.4% 3.3 50.8%
    Brown 8.5% 6.9 58.7%
    Columbia 6.6% 9.3 61.4%
    Cornell 12.7% 4.4 55.9%
    Dartmouth 10.5% 5.6 58.8%
    Harvard 5.2% 16.1 83.7%
    Princeton 6.4% 10.2 65.3%
    Penn 9.3% 7 65.1%
    Yale 6.9% 10 69.0%

    For comparison, here is CalPolySLO which I think validates my safety school hypothesis:
    CalPolySLO 34.60% 0.9 31.1%

    If you have the patience to do the rest of the data entry I think these numbers would be a good addition to your analysis.

  171. @Steve Sailer
    @dux.ie

    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?

    Replies: @res, @dux.ie

    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?

    That’s an interesting question. I would add “The child of a donor of over $X?”

    One relevant point is the biggest boost is hard to detect unless the starting point is low. I am defining boost as difference between necessary SAT absent/present the special characteristic holding all other data constant.

    My thought would be Obama daughter >> non-donor billionaire. Not sure how much a big donation could move that. What do you think? Does it depend on the billionaire? For example, crime-thinking Peter Thiel?

  172. @dux.ie
    @res

    > Assuming those aren’t lies.

    One late admittance case for a high school shooting survivor and student leader that received much publicity, the SAT score 1270 involved was not that low but well below the Harvard SAT 25th percentile at 1470 (SATwai25=1410), and the Harvard special invite for White rural students at 1310, below the average score for Pol Sci he wanted to major in, and was previously denied to UCLA, UCSD and UCIrvine,

    https://www.businessinsider.com.au/david-hogg-attend-harvard-university-1270-sat-score-reactions-2018-12

    http://i68.tinypic.com/35hr1uw.png

    The one allegely accepted to Harvard with SATeq 990 the sample data that I have showed that with that SAT score from NY she ONLY applied to Harvard which accepted and Princeton which denied. She must be full of self-confidence, no SUNY, no CUNY, no PlanB. She must be one of those special invitee, from the Harvard court case SFFA document, Harvard special invitees lowest cut-off is SAT 1100.

    Replies: @Steve Sailer, @res

    That Harvard graphic is very interesting. I think it also provides a good test case for the validity of your data. My guess would be the self reports are biased towards people who got a high SAT score and were denied admission somewhere then found that site. Not sure how big an effect that would be, but it might matter.

    That said, I think the acceptance rate non-monotonicity is better explained by applicant self selection. Presumably people scoring 1400 (or lower!) and applying to Harvard are more likely to have an in (legacy, special talent, race, …) and will be admitted at a higher frequency than the average person with that score. On the other hand there will be many 1500 scorers who think they have a chance because of that alone (or with semi-competitive rest of the application). They will be rejected at a higher rate.

    So if that hypothesis is correct then your data is correctly reflecting reality (roughly anyway), but one should NOT conclude that the acceptance rates at a given SAT are reflective of what a random person with that score would experience.

  173. @Steve Sailer
    @dux.ie

    Who would be likely to get the biggest boost on admission of anybody in the country? An Obama daughter? The child of a billionaire?

    Replies: @res, @dux.ie

    We will never know what sort of arm twisting happened behind closed doors. I am not implying anything but I noted that the new Harvard President was special advisor in Obama’s special committee on Historically Black Colleges and Universities HBCU. Harvard, Horvard, Howard … May be JKR’s Hogwarts was modelled from that.

    When there was news that nannies were paid $100K pa many considered that as excessive. I raised the case of K whom high school counsellor allegely said that his academic performance was so so and he had no chance of getting into Harvard. It was alleged that K’s father had to donate $200K pa for 10 years to get him into Harvard. If K’s father is as smart as K who hired a special tiger nanny for his daughter, lets say for 10 years, he might have saved $1 million from that. As the collateral gain, K’s daughter had already pulled off a spectacular priceless diplomatic coup with a foreign VIP. Look at the way she conduct herself now,

    She might save K with future headache as she wanted to go to Penn State, the people’s university appropriate for her constituents. I wonder why not UPenn like grand-daddy.

  174. @PhysicistDave
    @Thulean Friend

    Thulean Friend wrote:


    UMass Amherst is also having math PhDs. 750 mean in maths at SAT would probably be enough for a 2nd or 3rd tier University PhD.
     
    No, I don't think so. As others have said, for those of us who are reasonably good at math, an 800 on the old SAT Math, before they made the scoring easier, was not very difficult.

    When I was at Caltech back in the '70s, we sort of looked down our noses at the students who were somehow admitted with less than an 800 on math: we figured they might be okay in biology.

    Yes, Caltech is harder than anywhere else except MIT, but I am talking undergrads, most of whom were not math majors. I would bet that a Ph.D. in math, even at UMass Amherst, requires being as good at math as the average BS chemical engineer from Caltech.

    Replies: @Thulean Friend

    I would bet that a Ph.D. in math, even at UMass Amherst, requires being as good at math as the average BS chemical engineer from Caltech.

    UMass Amherst is still more like a 2nd tier school, remember that there are 3rd and 4th tier ones putting out math PhDs too. I’d highly doubt that having a perfect math SAT score at those colleges would be required for a math PhD, especially given the affirmative action factor. You’re too blinded by the fact that you went to what is arguably the US’ most competitive university.

  175. Thulean Friend wrote to me:

    You’re too blinded by the fact that you went to what is arguably the US’ most competitive university.

    I think I’ll take that as a backhanded compliment!

    Perhaps you’re right. I’d like to think that, with the backstop of the axiomatic method, higher academic math is not completely corrupt. (The area of physics in which I did my Ph.D., high-energy elementary-particle physics is quite corrupt, but I like to attribute that to lack of new experimental data: comparison of theory with experiment is our backstop.)

    But maybe academic math is worse than I expect.

  176. Anonymous[219] • Disclaimer says:
    @Steve Sailer
    @Anon

    I suspect it's pretty likely that parents' jobs have a sizable influence on what jobs kids go into. A lot of military officers are the children of a military officer, a lot of college professors are the children of a college professor. Some of that is genetic, but another part is simply that you can imagine yourself doing that job because you saw somebody in the family do it.

    Replies: @Anon, @Anonymous

    I have a good sample of elite math grad students. A lot of the (American) boys were the children of profs, often in other fields. But the girls! A full half had fathers who were math/physics profs.

    • Replies: @Steve Sailer
    @Anonymous

    My guess is that a very high rate of girls going into male-dominated fields (math professor, military officer, athlete, etc.) is where daughters take after their beloved fathers. Next generation female accomplishment tends to depend upon having a devoted father in the house.

  177. @Anonymous
    @Steve Sailer

    I have a good sample of elite math grad students. A lot of the (American) boys were the children of profs, often in other fields. But the girls! A full half had fathers who were math/physics profs.

    Replies: @Steve Sailer

    My guess is that a very high rate of girls going into male-dominated fields (math professor, military officer, athlete, etc.) is where daughters take after their beloved fathers. Next generation female accomplishment tends to depend upon having a devoted father in the house.

  178. dux.ie [AKA "BlackPajama"] says:

    Harmon does not seem to be dumb. Being a graduate of University of Michigan the worst case projection is that she might be on the UMich 25th percentiles on SATV=640 and SATM=670 which are reasoanbly good, with an estimated IQlkie of 120. Thus her problems might be other than intellect.

    It might seem that she is an arm-chair critics living in a bubble with the “why cant those starving people eat cake” mentality. She might have lost touch with the reality. In real life almost none of the Blacks want to be maths professor. A single anecdotal data from Molyneux’s interview with an intelligent Black,

    “The Black Man’s Burden”

    an alleged Black with the old SAT score of 2120 that should correspond to IQlike 127 (96.8045 percentile) but Molyneux interpreted that as IQ 150 (99.949 percentile). Anyway even IQlkie 127 is fairly high, the caller would have IQ higher than that for Harmon. The caller was offered ivy league uni places without SAT/ACT and was pressured to become a POLITICIAN to serve the people, not medical doctor that he aspired to be, NOT Maths Professor.

    Another example might be Obama. When he was young in high school he seemed to be interested in mathematical physics and he was pondering if time is really a fourth physical dimension. Later in university he flirted with the idea to become an architect.

    https://www.curbed.com/2011/6/3/10463548/who-knew-president-obama-once-wanted-to-be-an-architect

    And he was only dating white girl-friends then. Somehow somebody must had discovered him and had persuaded him to study law at Harvard. While at Harvard Obama did help a law professor to published a paper on blending constitution law and general theory of relativity.

    https://www.jstor.org/stable/pdf/1341407.pdf “The Curvature of Constitutional Space: What Lawyers Can Learn from Modern Physics … Tyler Professor of Constitutional Law, Harvard Law School … grateful to … Barack Obama for their analytic and research assistance.”

    Obama later married a black woman from Chicago, setting the path to become the first US black president.

    Harmon’s ignorant quest to find more black maths professors will most probably be futile. Yes there are intelligent blacks but most of them desire or under pressure to become lawyers and POLITICIANs.

    • Replies: @Steve Sailer
    @dux.ie

    The U.S. has lots and lots of affirmative action for blacks who could become math professors. E.g., economist Roland Fryer, a bright guy, is the highest salaried professor at Harvard College. I think he was being paid over $600k. Presumably, he would have outside part time opportunities to add to that, until his recent #MeToo difficulties (which seem overblown to me). There are a number of other extremely bright black economists. I would imagine a 145 IQ black can make a lot more money as an econ professor or a business school professor than as a math professor. I am in favor of extremely smart black guys socking away a fair amount of money, if they want to, so that there kids can have trust funds and choose more unworldly careers.

    Replies: @Benjaminl

  179. @dux.ie
    Harmon does not seem to be dumb. Being a graduate of University of Michigan the worst case projection is that she might be on the UMich 25th percentiles on SATV=640 and SATM=670 which are reasoanbly good, with an estimated IQlkie of 120. Thus her problems might be other than intellect.

    It might seem that she is an arm-chair critics living in a bubble with the "why cant those starving people eat cake" mentality. She might have lost touch with the reality. In real life almost none of the Blacks want to be maths professor. A single anecdotal data from Molyneux's interview with an intelligent Black,

    "The Black Man's Burden"
    https://www.youtube.com/watch?v=mZExVmr4gqI

    an alleged Black with the old SAT score of 2120 that should correspond to IQlike 127 (96.8045 percentile) but Molyneux interpreted that as IQ 150 (99.949 percentile). Anyway even IQlkie 127 is fairly high, the caller would have IQ higher than that for Harmon. The caller was offered ivy league uni places without SAT/ACT and was pressured to become a POLITICIAN to serve the people, not medical doctor that he aspired to be, NOT Maths Professor.

    Another example might be Obama. When he was young in high school he seemed to be interested in mathematical physics and he was pondering if time is really a fourth physical dimension. Later in university he flirted with the idea to become an architect.

    https://www.curbed.com/2011/6/3/10463548/who-knew-president-obama-once-wanted-to-be-an-architect

    And he was only dating white girl-friends then. Somehow somebody must had discovered him and had persuaded him to study law at Harvard. While at Harvard Obama did help a law professor to published a paper on blending constitution law and general theory of relativity.

    https://www.jstor.org/stable/pdf/1341407.pdf "The Curvature of Constitutional Space: What Lawyers Can Learn from Modern Physics ... Tyler Professor of Constitutional Law, Harvard Law School ... grateful to ... Barack Obama for their analytic and research assistance."

    Obama later married a black woman from Chicago, setting the path to become the first US black president.

    Harmon's ignorant quest to find more black maths professors will most probably be futile. Yes there are intelligent blacks but most of them desire or under pressure to become lawyers and POLITICIANs.

    Replies: @Steve Sailer

    The U.S. has lots and lots of affirmative action for blacks who could become math professors. E.g., economist Roland Fryer, a bright guy, is the highest salaried professor at Harvard College. I think he was being paid over $600k. Presumably, he would have outside part time opportunities to add to that, until his recent #MeToo difficulties (which seem overblown to me). There are a number of other extremely bright black economists. I would imagine a 145 IQ black can make a lot more money as an econ professor or a business school professor than as a math professor. I am in favor of extremely smart black guys socking away a fair amount of money, if they want to, so that there kids can have trust funds and choose more unworldly careers.

    • Replies: @Benjaminl
    @Steve Sailer

    Stuart Taylor has published a report that makes the #metoo case against Fryer look pretty shaky:

    https://www.realclearinvestigations.com/articles/2019/01/27/harvard_the_new_york_times_and_the_metoo_takedown_of_a_black_academic_star.html


    Despite being a Respectable Establishment type, Taylor has seemed pretty skeptical of the Wisdom of the Current Year at least since his research demolishing the Duke Lacrosse Fiasco.

    On a quick Google, I misread Taylor's title as Nonresident Senior Fellow in Grievance Studies at the Brookings Institution.

  180. @Steve Sailer
    @dux.ie

    The U.S. has lots and lots of affirmative action for blacks who could become math professors. E.g., economist Roland Fryer, a bright guy, is the highest salaried professor at Harvard College. I think he was being paid over $600k. Presumably, he would have outside part time opportunities to add to that, until his recent #MeToo difficulties (which seem overblown to me). There are a number of other extremely bright black economists. I would imagine a 145 IQ black can make a lot more money as an econ professor or a business school professor than as a math professor. I am in favor of extremely smart black guys socking away a fair amount of money, if they want to, so that there kids can have trust funds and choose more unworldly careers.

    Replies: @Benjaminl

    Stuart Taylor has published a report that makes the #metoo case against Fryer look pretty shaky:

    https://www.realclearinvestigations.com/articles/2019/01/27/harvard_the_new_york_times_and_the_metoo_takedown_of_a_black_academic_star.html

    Despite being a Respectable Establishment type, Taylor has seemed pretty skeptical of the Wisdom of the Current Year at least since his research demolishing the Duke Lacrosse Fiasco.

    On a quick Google, I misread Taylor’s title as Nonresident Senior Fellow in Grievance Studies at the Brookings Institution.

Comments are closed.

Subscribe to All Steve Sailer Comments via RSS