Not surprising. I recall reading about some inner city Los Angeles kids with high GPAs in high school being totally overwhelmed at UC Berkeley. They were not prepared for the intense competition there. And probably did not have the needed skills either. Too bad. They might have been successful at Cal State Los Angeles or the like.
There are many students at MIT who are overwhelmed by the competition. Remember, 50% of the freshmen class at MIT goes from being the top of their high school class to the bottom half of their college class.
The overwhelmed students mostly get over it and do fine. I had a high school friend who went to CalTech. He was outraged when he saw “Calculus” on his first semester schedule. He went to the professor assigned as his advisor and complained that he’d gotten a 5 on the AP calculus test. The response “this is CalTech. Everyone got a 5 on the AP calculus exam. One of your classmates did it when he was 11.” My friend was overwhelmed his first year, but graduated, and now has a nine figure net worth.
Cal Tech has 2k undergraduates. That is the same undergraduate size of a small liberal arts college.
As a joke, I always point out that the University of Central Florida has more undergraduates than the entire Ivy League. I heard the President of the University of Central Florida speak and his joke was that his university was not the same species at Cal Tech.
Yes, any freshman is likely to get surprised at how much more work college is compared to high school. In my life, I've only met one person who didn't feel pressured by that first year of college. (She became unhappy at how hard she'd been made to work in lower ed, then disgusted at how easy nursing school was.) For all the rest of us, the shock was how serious we had to be.
Yes, every year 50% of the freshmen class at MIT goes from being the top of their high school class to the bottom half of their college class. And that can't be changed by more tutoring or counseling or anything. It's math at Lake Woebegone Failure level
It appears that commenter Richard Bicker was calling for MIT to provide Lake Wobegon counseling services to those hapless freshmen, for the reason you describe.
I was strolling about Caltech's campus 20 years ago when I fell in with some high school seniors and their parents being given a tour by a sophomore. One dad asked their guide about Caltech's reputation for being hard.
"Freshman year was hard," the guide said, and then broke into tears. After about 10 seconds of sobbing, she recovered to say, "But this year is more reasonable."
I don't think MIT had a class rank. Do most colleges? The only ranking thing I can think of was for freshman physics did you do regular 8.01 or the more difficult 8.012. If you didn't wash out of 8.012 you could consider yourself a top 10% (maybe) brain at the 'tute. Usual caveat about my memory tending to hallucinate like chatGPT.
If one is taking quantum electrodynamics or tensor mechanics, the class will operate on some form of what most people call a curve. If one is making the C+ or B- that holds up the curve for everyone else, one is not going to enjoy being at MIT. That is where the idea of overmatched comes from. If one is in the bottom of the ability distribution, then other students are making higher grades with less effort.
I had a friend there who was course VI (EE/Computer Science) and he told me the exams were effectively IQ tests. No matter how he prepared he always got a solid C and he resigned to it. He enjoyed it as much as the higher scoring students I think. I don't think anyone in his first job would ask his GPA.
You're a jerk and your comment isn't responsive. At Berkeley, the students of color are getting overwhelmed in Econ 1, Comp. Lit., Spanish, and Ethnic Studoes, not Calculus, Physics and Chemistry.
UC-Berkeley is 2% black with about 50% of that number being immigrants or the children of immigrants. The six year graduation rate for black students is 83% as compared to 93% for non-Hispanic whites. And when the term overmatched was introduced, it was used to explain black students at Duke University dropping out of pre-med or engineering to an easier major. https://wol.iza.org/articles/what-is-the-nature-and-extent-of-student-university-mismatch
Great deflection. Neither Duke, blacks, the number of blacks at Berkeley nor their provenance have anything to do with the comment to which you responded above. The writer merely stated that some "inner-city" kids from L.A. were overwhelmed at Berkeley, and the fact is that those kids, whether black or latino, overwhelmingly take the soft courses, where they're nevertheless overwhelmed. Also, I've spent nearly three decades observing and interacting with black kids at Berkeley, and they're overwhelmingly non-immigrants. If you think Duke makes your point, then why not stick with it? Jeez, I'd lve to punch you.
at UC-Berkeley. 10% of the undergraduate are international students. Another 20% are immigrants who live in the U.S. Over 60% have one parents that was born outside the U.S. And one is going to argue that blacks are the one group that does not fit into that group.
Also, UC-Berkeley is 54% female, has a 94% graduation rate for women versus 91% for men and has an 86% graduation rate for Hispanics. There is no data to support the idea that UC-Berkeley is admitting unqualified students.
On the contrary, the extra 10% of blacks who don't graduate in six years is prima facie evidence that unqualified-for-success blacks are being admitted at a greater percentage than other groups. There could be other hypotheses -- e.g., maybe the black cohort is systematically taking harder courses so that they are having less success at completing their degrees than equally qualified Asians, etc. -- but that's not the way to bet.
Strange, in my state university no one cared where they ranked. Similar to most people’s belief about their driving skills, I think everyone just assumed they were in the top ten.
Given that most state university students change their major at least once and that public universities are rated on six year graduation rates, there is a lot of concerns at public universities. One of the interesting things that can happen at state universities is that student can be undermatched which is attending a state universities due to parents concerns on costs rather than attend the most selective university one can get admitted to.
I'm not sure. If you were the one 1600 SAT guy in an otherwise bad high school, chances are they didn't have many AP classes and you got As between naps. The kids who learned to buckle down when they were high school sophomores must have some advantage. The paper shows equivalent SAT from lesser high schools getting lower grades.
The inner-city LA kids referenced in the above post did not have 1600 SATs. They just had high, inflated GPAs. All the prep in the world can't overcome low intelligence.
I could be misreading the second set of graphs but it sure looks like they are showing at least one 1600 SAT from disadvantaged schools/underrepresented minorities
You are misreading this discussion. Mark Royer above mentioned some inner city LA kids he read about, most or all of whom certainly did not score 1600. That is the group he and I were commenting about. NOT the subjects of the NBER study that is the topic of Steve Sailer's article.
The GPA versus SAT/ACT is why lots of public universities have firewalls between their hardest majors/departments/colleges and those admitted. Just because one is admitted to UT-Austin due to being a top ten percent student does not mean that one gets to major in engineering. Or as mentioned in the book "Paying for the Party" getting into the Indiana University College of Business is much harder than getting admitted to the university.
That is why parents experience and knowledge is important for most college students and college applicants.
Also, as shown at UT-Austin if one admits just top ten percent students, then many majors/departments will not survive. If one only admits students with the highest SAT scores, then all of the drama/music/performing arts/visual arts programs/liberal arts programs do not survive and the university is reduced to a trade school for engineers, pre-meds, business, and pre-law.
I knew a rich guy who showed up at Stanford and found out it was Leland Stanford Junior University, so he figured he'd have to find a different university to go to after 2 years.
Eventually, his physics professor told him one day, "Yeah, we've decided you'll never win us a physics Nobel, so you are no longer a physics major. But the electrical engineering department wants you, and they make more money."
Almost all PhD physicist who are experimentalist have the story about being told that they are not going to make it as a theoretical physicist and either need to get really good at doing experiments or started to specialize such as being a nuclear physicist at a National Lab or switching to medical physics and working at a teaching hospital.
And it is a job that a good physicist but not great physicist can do and one can work where one wants to and one does not have to go to an Ivy League school. The issue is there is not a high school counselor in the U.S. that knows to recommend that a good student pursue something like medical physics, nurse anesthetist, or petroleum engineer.
My experience with engineering school is that their is an out door for the student who cannot cut it but not an in-door for someone who decides to be an engineer after being an undeclared major. That would probably add a year of school for no good reason.
In listening to the interviews about "Paying for the Party" it was stated that it was almost impossible to change into a business major at Indiana University if one does not start out as a business major.
I think the SAT scoring algorithm has been adjusted over the years such that an 800 math score can be achieved despite failing to answer every question correctly. This probably doesn’t affect its’ predictive validity.
I also can “vaguely” recall my scores with three digit accuracy, as it was the lifetime apex of my brain’s capabilities. Soon afterwards I was introduced to the glories of beer.
I took the SAT in May 1994, the final sitting before the first recentering. I failed only one math question, which got me a 780. The next school year, my senior year in high school, my HS guidance counselor showed me a conversion chart, showing SAT/M and SAT/V raw scores from the old and new system given the same raw score. I would have gotten an 800 on the M under the new system in spite of missing one. On verbal, my 570 on the old would have become 640 on the new. He told me the college admissions boards have the same conversion charts, so they were wise to the score inflation. I asked him why do the recentering at all. He was at a loss for an answer, but I suspect that he knew the real reason, as did I, which was the self-esteem of NAMs.
Virtually everyone admitted to MIT or Cal Tech has an 800 of the SAT Math portion. The value of the test to highly selective universities is to just determine the minimum level of competence. And it is one's ranking versus everyone else. That is the point of the scoring to begin with.
I do not understand people who think they are being clever by refusing to understand how the SAT actually works.
"Virtually everyone admitted to MIT or Cal Tech has an 800 of the SAT Math portion. "
Precisely my point. Do they admit everyone who applies with an 800? With so many Asians, how do they tell them apart? Are differences in the Verbal now important to them? How does the 800 or near 800 student figure out where he'll fit in?
My former boss's nephew scored 1600 post centering. He went with a full-ride to U AZ (home state) largely because his parents were cheapskates. Maybe that was best, as he didn't go to grad school IIRC. Joined the EPA bureaucracy.
As far as I know, MIT does not admit every applicant that has an 800 math SAT score. However, the students need a strong resume. The students at MIT did summer mentorship programs, went to summer programs to learn to code, win science, engineering, robotics, and coding competitions.
I have judged, at science fairs, high school senior accepted to MIT and other high end science schools. Those students are very talented but some of them still end up in the bottom half of their college class.
As a response from a parent one time, a mother told me that her daughter majored in drama at Carnegie-Mellon. Mom pointed out that every student accepted to the program has been the star of their high school plays and drama programs. But everyone who was a star in high school does not get to be a star in college.
It’s not particularly scientific to have a test that causes you to lose a notch for missing one question. An adaptive test that gives you, say, 10 questions at the 800 level would be better, because if you get 9/10 right, the 1 wrong answer is probably a fluke, but if you get 4/10 right, you probably deserve that 780. Loser! I remember the year I took the LSAT, you could miss up to five questions and get a perfect score, which at the time was a 48.
I was talking about the LSAT. The one that let me miss questions and get a perfect score. I took the SAT in 1981. Long before recentering. I don’t remember the SAT that well, except I do remember my friends and I coming up with farcical and obscene analogies. On the LSAT, I remember just loving the logic games, which they apparently have eliminated.
I can remember my exact SAT scores (took twice ~25 years ago) and exact grad school scores (took twice ~18 years) ago and I don't have an especially great memory. Funny how the mind works.
I can't remember mine. Sort of remember percentiles. 99+ verbal, 98th Math. The latter wasn't good enough for my major. it turned out. But this was in the later 1960s so I have the excuse that it was a long time ago.
> I think the SAT scoring algorithm has been adjusted over the years such that an 800 math score can be achieved despite failing to answer every question correctly.
Only if the question you miss is contained within one of the experimental unscored sections. 800 math is not a particularly high score.
Notice that the SAT and GPA are not as strong an indicator of success for students who attend disadvantaged high schools. The idea that students can be overmatched lives on.
I would say the GPA that got me 10th in my class, but I'm worried some of you would reach through my device and start punching me. (Which a friend from another high school did!)
We were overly and misguidedly interested in class rank since there were only 60 in my prep school class, all on the right slope of the Bell Curve, but I don't remember any discussion or even reporting to us of actual GPA scores and little about actual SAT scores, which seems weird now but it was the boys' school reticent culture. Our 3 Ivies were all legacies, which also checked their bragging rights a bit, and 10 went to UVA, one of whom, the quarterback, became a multi-millionaire hedge fund manager. There was some resentment expressed that BC Calculus got a 15% bonus to AB's 10%. AB grades were lower, which was blamed on more difficult material, not the students.
I went to a very good suburban public high school. Senior year was all AP classes and it was the same 10 or 15 guys, no dames as we didn't call them back then, every period. I take pride in the fact that those guys were ranked 1-15 and I was somewhere closer to 100.
Thanks. Not mortify. I was proud of it. All the other guys were grade grinds. I focused on what I found interesting and intellectually stimulating. I transferred in second year and none of my classes transferred as honors, so no way I was cracking the top twenty students anyway. I went to my first choice college (or Institute of Technology) and was happy. My parents didn't hound me about grades. Dad thought the point of college was education, not job training, and I agree.
I find it infuriating that these tests can't do a better job of separating kids out at the top end. A few years ago the PSAT was redesigned and basically made easier. I don't recall the exact numbers, but it was essentially as if they chopped off the hardest questions. So the test became more of a fluke.
(The PSAT is very useful for doing Sailer-like comparisons because the National Merit scores are separated out by state, so that you get a quick comparison of the cutoff scores for the top 2% all across the country; students can only qualify for National Merit once, so it's not as gamed as the ACT/SAT by kids taking it ten times; and the National Merit Semifinalist lists are published, so people such as our host can go through and analyze the names for ethnic information.)
At some point IQ probably doesn't tell you that much about who's going to be the best doctor or lawyer or politician, but it doesn't seem like a coincidence that the very, very best test takers seem to end up as renowned math or physics professors. If nothing else, you could reduce the amount of time test-takers have, but that would hurt kids' feelings, I guess. (Don't even get me started on the people who manipulate the system to get their kids made-up ADHD diagnoses for the Adderall and added time.)
I was a very lazy student in high school and never took a school book home. I had a fairly high GPA but that was in a Prince George's County MD public school. Football and nights-out with my friends was more important.
There is definitely quite a bit of grade inflation these days - aside from the bonus for honors/AP classes (I don't seem to recall that being in effect 30 years ago, you just got the benefit of that designation on your transcript), lots of schools allow partial credit for late assignments or extra credit assignments to prevent disorganized students from tanking their grades, neither of which were available at my high school. I assume there is still some discount or bonus to a transcript based on the admissions office's knowledge of how decent of a HS the student attended, which I totally believe was the case in my day based on my mediocre GPA from a selective high school and the schools I was admitted to.
It also highlights how soft affirmative actions initiatives like offering automatic admission to state schools for students in the top 10% of a HS class is a massive mistake. There is a gaping chasm in the quality of even a 50th percentile student at my former HS versus that at the local public urban achievers academy, and the latter would get killed at my state's flagship public university. I remember when Georgetown started offering admissions to DCPS students in the top decile over 20 years ago and the WaPo would periodically do a story on these kids and it was always the same - they were struggling mightily and realized that the coursework required for a 4.0 from Dunbar HS was a joke compared to virtually all of their non-DC peers former schools. This was prior to the era of equity so I assume most of them ended up withdrawing and going to Bowie State or something, whereas today they would be shepherded to graduation through some absurd degree program.
Bowie State was a good way fifty years ago for bright, but far from genius, blacks to get jobs in the lucrative Prince George's County public schools. I imagine it remains so.
I suspect that one of the reasons that Texas has become more conservative politically despite the changing demographics is the top 10% rule. Forcing every high school in Texas (public and private) to rank their students from 1 to the bottom lets students know where they stand in the pecking order. And all of the students and their parents can see the benefits of affirmative action when the child of two black white-collar professionals gets into an Ivy League/Ivy Like school when ranked below white and Asian-American classmates who have zero chance of getting into an Ivy League.
PSAT and SAT/ACT have always been IQ tests. In part, that's why they are hyped/standard: the Deep State wants to identify early those folks who could be useful assets or else troublemakers.
The PSAT in particular seems to the Deep State's game. Despite "not counting" towards your actual SAT/ACT score, they give prizes and recognition to those who score highly on them. Why would a practice SAT exam merit such distinction, when it "doesn't count"? Likely because there's a correlation to those naturally gifted folks who do well on the PSAT and raw intelligence --- i.e. good places to start if you're looking for a CIA or NSA asset.
Kids who are worried for college will study for the SAT/ACT, which skews their scores up. But kids who are just naturally smart will score best on the PSAT. For example, Hillary Clinton scored very highly on her PSAT, so much so that Vox Day believes she is the smartest person ever to run for president in the last 100 years.
I'm not saying they're not smart, I'm saying few politicians are trustworthy about themselves, and the Clintons aren't two of them. And Hillary smarter than Hoover, Nixon, and Romney? In 30+ years, I've seen little evidence for that--or much wisdom--but no way I'm looking for a light under her bushel.
What I really found interesting were the subtle differences of the results in figure 2: To recapitulate
1: students with same SAT performed worse at college if they went before to a bad high school
2: students with same SAT performed worse at college if they were underrepresented minority
But: bad high schools flattened the regression slope (from 0.11 to 0.10). On the other hand, being underrepresented minority appeared to depress the performance level while not affecting the regression slope. This difference - deviation in regression slope versus basic level should (and can be easily) tested with post-hoc statistics. One would only need to slightly modify the test regression formula to include effects of the high school (or minority parameter) not only on the basic level but also on slope. IE instead of Y = a + b + c*X, change to Y = a + b + c*X + d*X, where b and d are coefficients to model the effects of the high school (or minority parameter) on basic level and slope, respectively.
I found the apparent differences (slope vs basic level) could suggest two things: students visiting a bad high school don't learn how to thrive in a rich academic environment as well as those from a good school. And also - SAT possibly doesn't pick all of the deficit in cognitive thrust in underrepresented minorities.
Underlyer for college GPA performance. I admit I partly wanted to use a different phrase so I invented it out of whole cloth. But there is more to it. Usually, all parts of intelligence correlate strongly with each other - that's why the psychometrics results over the last 100 years suggest something general - usually abbreviated as g. But here in figure 2B, the deficit in College GPAs stayed more or less the same across the whole spectrum of different SAT results. In other words - the general deficit did NOT correlate with the SAT results.
Or is it just admissions mismatch--the brainier disadvantaged students were DEI-ed into schools over their head? The absolute numbers probably aren't that high, so the top schools rob much of the 2nd tier's, and all the way down.
It was not an admissions mismatch because the comparison is between minorities and non-minorities with the SAME SAT score. Based on that SAT score, underrepresented minorities do worse with college grades.
It would have been a terrible methodical error to simply allow GPAs from one college to be counted in absolute terms together with GPAs from another college one on one. Because some universities are tough, others suffer horribly from grade inflation. One could of course correct this by balancing GPAs and SATS center of gravities and standard deviations across different colleges.
But yes, if one didn't do such a correction, that methodical error would be a good explanation.
Another analysis would show that if one is the first in the family to attend college, the SAT/ACT has less predictive power no matter the race. A school like UC-Berkeley depends upon the college educated parents to do the real counselling and college prep.
Yep - that's in the appendix figures 2B&C. Did you read the paper as far as that? Anyway, there, the effects of coming from a poor family (Appendix fig 2B) or from a non-academic family (Appendix fig 2C) both reduced the slope between college GPA and SAT.
For some imbecillic reason, authors didn't do statistics into the figures, but the effects were eye-catchingly profound. So the effects were like that of disadvantaged schools, only a factor of 2x or 3x stronger. Like with bad schools the effect was probably due to their defective social environment unable to really challenge the exceptionally gifted.
I think "Matilda" 1996 film by Danny DeVito deals with this issue too.
IMHO it is less about counselling or college prep. Growing with academics and going to academically good schools enhances human flourishing because the gifted person learns not to rest on his laurels.
> Moon Unit Zappa’s grandpa seems likely to have been of use
Her dad was certainly an odd ball and had an odd career in that he was a famous popular musician who never had a hit song. Marc Maron idolized him to the point where he adopted his look and dated his daughter
Zappa hosted SNL in 1978 and was so obnoxious that he was never asked back; he committed the cardinal sin of not playing along with the fiction that the players memorized their lines and would make it blatantly obvious he was reading from cue cards
As a teacher I can tell you that grades are mostly bullshit; if I gave the grades my students earned then my life would be much more difficult, perhaps to the point of losing my job, so I give inflated grades. They are only accurate in the sense that I don't jump one student over another; a list of my students ranked by grades would be the same under a bullshit scale and an actual scale
One of the golden rules of education is one can either have high standards with high failure rates or low standards with low failure rates. Politicians, school administrators, etc. have found that life is much easier with low standards and low failure rates.
When I was attending a DoDEA high school, I was a cheerleader and one of our cheers to razz the opposing teams was: "Cornbread! Chicken! Rice! Peas! We got higher SATs!"
I wonder how many, if any, civilian high schools would taunt their football rivals that way.
Most child of military members do not attend a DODEA school. They are a legacy of the segregated school systems in the South along with some international locations.
And according to wiki, there are only two DODEA high schools in the U.S.
And the DODEA schools are a smaller school system that the public school systems of North Dakota or Wyoming.
As a child in an FDNF family, I received an outstanding education while growing up almost entirely in the Pacific and Europe thanks to the Department of Defense. And, yes, I was inculcated in Americanism and civic nationalism.
"My vague recollection is that I was seventh out of 181 with a 3.81 GPA."
Ha ha, that's really "vague"... very cheeky Steve... :-)
Not surprising. I recall reading about some inner city Los Angeles kids with high GPAs in high school being totally overwhelmed at UC Berkeley. They were not prepared for the intense competition there. And probably did not have the needed skills either. Too bad. They might have been successful at Cal State Los Angeles or the like.
There are many students at MIT who are overwhelmed by the competition. Remember, 50% of the freshmen class at MIT goes from being the top of their high school class to the bottom half of their college class.
Every year? They should provide more tutoring or counseling or something.
They do not care. That is why MIT has a higher suicide rate than most universities.
Like how high? I always thought Cornell was number one due to the gorge.
😆
LOL, literally.
The overwhelmed students mostly get over it and do fine. I had a high school friend who went to CalTech. He was outraged when he saw “Calculus” on his first semester schedule. He went to the professor assigned as his advisor and complained that he’d gotten a 5 on the AP calculus test. The response “this is CalTech. Everyone got a 5 on the AP calculus exam. One of your classmates did it when he was 11.” My friend was overwhelmed his first year, but graduated, and now has a nine figure net worth.
Interesting story. Just getting through the grind is the great accomplishment. Resilience is important.
Cal Tech has 2k undergraduates. That is the same undergraduate size of a small liberal arts college.
As a joke, I always point out that the University of Central Florida has more undergraduates than the entire Ivy League. I heard the President of the University of Central Florida speak and his joke was that his university was not the same species at Cal Tech.
Yes, any freshman is likely to get surprised at how much more work college is compared to high school. In my life, I've only met one person who didn't feel pressured by that first year of college. (She became unhappy at how hard she'd been made to work in lower ed, then disgusted at how easy nursing school was.) For all the rest of us, the shock was how serious we had to be.
Yes, every year 50% of the freshmen class at MIT goes from being the top of their high school class to the bottom half of their college class. And that can't be changed by more tutoring or counseling or anything. It's math at Lake Woebegone Failure level
It appears that commenter Richard Bicker was calling for MIT to provide Lake Wobegon counseling services to those hapless freshmen, for the reason you describe.
I LOL'd, literally.
I was strolling about Caltech's campus 20 years ago when I fell in with some high school seniors and their parents being given a tour by a sophomore. One dad asked their guide about Caltech's reputation for being hard.
"Freshman year was hard," the guide said, and then broke into tears. After about 10 seconds of sobbing, she recovered to say, "But this year is more reasonable."
I don't think MIT had a class rank. Do most colleges? The only ranking thing I can think of was for freshman physics did you do regular 8.01 or the more difficult 8.012. If you didn't wash out of 8.012 you could consider yourself a top 10% (maybe) brain at the 'tute. Usual caveat about my memory tending to hallucinate like chatGPT.
If one is taking quantum electrodynamics or tensor mechanics, the class will operate on some form of what most people call a curve. If one is making the C+ or B- that holds up the curve for everyone else, one is not going to enjoy being at MIT. That is where the idea of overmatched comes from. If one is in the bottom of the ability distribution, then other students are making higher grades with less effort.
I had a friend there who was course VI (EE/Computer Science) and he told me the exams were effectively IQ tests. No matter how he prepared he always got a solid C and he resigned to it. He enjoyed it as much as the higher scoring students I think. I don't think anyone in his first job would ask his GPA.
You're a jerk and your comment isn't responsive. At Berkeley, the students of color are getting overwhelmed in Econ 1, Comp. Lit., Spanish, and Ethnic Studoes, not Calculus, Physics and Chemistry.
UC-Berkeley is 2% black with about 50% of that number being immigrants or the children of immigrants. The six year graduation rate for black students is 83% as compared to 93% for non-Hispanic whites. And when the term overmatched was introduced, it was used to explain black students at Duke University dropping out of pre-med or engineering to an easier major. https://wol.iza.org/articles/what-is-the-nature-and-extent-of-student-university-mismatch
Great deflection. Neither Duke, blacks, the number of blacks at Berkeley nor their provenance have anything to do with the comment to which you responded above. The writer merely stated that some "inner-city" kids from L.A. were overwhelmed at Berkeley, and the fact is that those kids, whether black or latino, overwhelmingly take the soft courses, where they're nevertheless overwhelmed. Also, I've spent nearly three decades observing and interacting with black kids at Berkeley, and they're overwhelmingly non-immigrants. If you think Duke makes your point, then why not stick with it? Jeez, I'd lve to punch you.
at UC-Berkeley. 10% of the undergraduate are international students. Another 20% are immigrants who live in the U.S. Over 60% have one parents that was born outside the U.S. And one is going to argue that blacks are the one group that does not fit into that group.
Also, UC-Berkeley is 54% female, has a 94% graduation rate for women versus 91% for men and has an 86% graduation rate for Hispanics. There is no data to support the idea that UC-Berkeley is admitting unqualified students.
On the contrary, the extra 10% of blacks who don't graduate in six years is prima facie evidence that unqualified-for-success blacks are being admitted at a greater percentage than other groups. There could be other hypotheses -- e.g., maybe the black cohort is systematically taking harder courses so that they are having less success at completing their degrees than equally qualified Asians, etc. -- but that's not the way to bet.
I don't think name calling belongs in the level of discussion we are going for here.
Simply by mathematic necessity. Math is so unfair! Zilch social justice.
Strange, in my state university no one cared where they ranked. Similar to most people’s belief about their driving skills, I think everyone just assumed they were in the top ten.
Given that most state university students change their major at least once and that public universities are rated on six year graduation rates, there is a lot of concerns at public universities. One of the interesting things that can happen at state universities is that student can be undermatched which is attending a state universities due to parents concerns on costs rather than attend the most selective university one can get admitted to.
The problem was not lack of "preparation" for "competition".
The problem was they lacked the intelligence to grasp the subject matter.
I'm not sure. If you were the one 1600 SAT guy in an otherwise bad high school, chances are they didn't have many AP classes and you got As between naps. The kids who learned to buckle down when they were high school sophomores must have some advantage. The paper shows equivalent SAT from lesser high schools getting lower grades.
The inner-city LA kids referenced in the above post did not have 1600 SATs. They just had high, inflated GPAs. All the prep in the world can't overcome low intelligence.
I could be misreading the second set of graphs but it sure looks like they are showing at least one 1600 SAT from disadvantaged schools/underrepresented minorities
You are misreading this discussion. Mark Royer above mentioned some inner city LA kids he read about, most or all of whom certainly did not score 1600. That is the group he and I were commenting about. NOT the subjects of the NBER study that is the topic of Steve Sailer's article.
The GPA versus SAT/ACT is why lots of public universities have firewalls between their hardest majors/departments/colleges and those admitted. Just because one is admitted to UT-Austin due to being a top ten percent student does not mean that one gets to major in engineering. Or as mentioned in the book "Paying for the Party" getting into the Indiana University College of Business is much harder than getting admitted to the university.
That is why parents experience and knowledge is important for most college students and college applicants.
Also, as shown at UT-Austin if one admits just top ten percent students, then many majors/departments will not survive. If one only admits students with the highest SAT scores, then all of the drama/music/performing arts/visual arts programs/liberal arts programs do not survive and the university is reduced to a trade school for engineers, pre-meds, business, and pre-law.
I knew a rich guy who showed up at Stanford and found out it was Leland Stanford Junior University, so he figured he'd have to find a different university to go to after 2 years.
Eventually, his physics professor told him one day, "Yeah, we've decided you'll never win us a physics Nobel, so you are no longer a physics major. But the electrical engineering department wants you, and they make more money."
Almost all PhD physicist who are experimentalist have the story about being told that they are not going to make it as a theoretical physicist and either need to get really good at doing experiments or started to specialize such as being a nuclear physicist at a National Lab or switching to medical physics and working at a teaching hospital.
medical physics turned into a great job in the early 2000s because of computer advances in radiation oncology.
And it is a job that a good physicist but not great physicist can do and one can work where one wants to and one does not have to go to an Ivy League school. The issue is there is not a high school counselor in the U.S. that knows to recommend that a good student pursue something like medical physics, nurse anesthetist, or petroleum engineer.
There might be a few high school counselors in Texas who know about petroleum engineering.
"the university is reduced to a trade school for engineers, pre-meds, business, and pre-law."
Back to where most of them started.
There were 84 grads in my grandfather's UNC-CH class of 1912. Not many majors on offer.
So if I get into one of those schools, can I transfer to the more demanding major if I do well? Well, not me. I’m too old to learn to be an engineer.
My experience with engineering school is that their is an out door for the student who cannot cut it but not an in-door for someone who decides to be an engineer after being an undeclared major. That would probably add a year of school for no good reason.
In listening to the interviews about "Paying for the Party" it was stated that it was almost impossible to change into a business major at Indiana University if one does not start out as a business major.
"the university is reduced to a trade school for engineers, pre-meds,..."
You say that as if it's a bad thing.
I think the SAT scoring algorithm has been adjusted over the years such that an 800 math score can be achieved despite failing to answer every question correctly. This probably doesn’t affect its’ predictive validity.
I also can “vaguely” recall my scores with three digit accuracy, as it was the lifetime apex of my brain’s capabilities. Soon afterwards I was introduced to the glories of beer.
I took the SAT in May 1994, the final sitting before the first recentering. I failed only one math question, which got me a 780. The next school year, my senior year in high school, my HS guidance counselor showed me a conversion chart, showing SAT/M and SAT/V raw scores from the old and new system given the same raw score. I would have gotten an 800 on the M under the new system in spite of missing one. On verbal, my 570 on the old would have become 640 on the new. He told me the college admissions boards have the same conversion charts, so they were wise to the score inflation. I asked him why do the recentering at all. He was at a loss for an answer, but I suspect that he knew the real reason, as did I, which was the self-esteem of NAMs.
It had to do with more students taking the test and wanting to keep the mean math/verbal combined around 1000.
Thereby making the test less useful to the most selective schools. They should have added points on the right tail at the same time.
Virtually everyone admitted to MIT or Cal Tech has an 800 of the SAT Math portion. The value of the test to highly selective universities is to just determine the minimum level of competence. And it is one's ranking versus everyone else. That is the point of the scoring to begin with.
I do not understand people who think they are being clever by refusing to understand how the SAT actually works.
"Virtually everyone admitted to MIT or Cal Tech has an 800 of the SAT Math portion. "
Precisely my point. Do they admit everyone who applies with an 800? With so many Asians, how do they tell them apart? Are differences in the Verbal now important to them? How does the 800 or near 800 student figure out where he'll fit in?
My former boss's nephew scored 1600 post centering. He went with a full-ride to U AZ (home state) largely because his parents were cheapskates. Maybe that was best, as he didn't go to grad school IIRC. Joined the EPA bureaucracy.
As far as I know, MIT does not admit every applicant that has an 800 math SAT score. However, the students need a strong resume. The students at MIT did summer mentorship programs, went to summer programs to learn to code, win science, engineering, robotics, and coding competitions.
I have judged, at science fairs, high school senior accepted to MIT and other high end science schools. Those students are very talented but some of them still end up in the bottom half of their college class.
As a response from a parent one time, a mother told me that her daughter majored in drama at Carnegie-Mellon. Mom pointed out that every student accepted to the program has been the star of their high school plays and drama programs. But everyone who was a star in high school does not get to be a star in college.
It’s not particularly scientific to have a test that causes you to lose a notch for missing one question. An adaptive test that gives you, say, 10 questions at the 800 level would be better, because if you get 9/10 right, the 1 wrong answer is probably a fluke, but if you get 4/10 right, you probably deserve that 780. Loser! I remember the year I took the LSAT, you could miss up to five questions and get a perfect score, which at the time was a 48.
The funny thing is that your class was the first one that took the new SAT, the one with open-ended math questions, no antonyms, and no TSWE
I was talking about the LSAT. The one that let me miss questions and get a perfect score. I took the SAT in 1981. Long before recentering. I don’t remember the SAT that well, except I do remember my friends and I coming up with farcical and obscene analogies. On the LSAT, I remember just loving the logic games, which they apparently have eliminated.
I wasn't responding to you; why are you responding to me as if I did?
It was threaded as a response to me. Or at least that was how it seemed on my device. No hostility intended.
I can remember my exact SAT scores (took twice ~25 years ago) and exact grad school scores (took twice ~18 years) ago and I don't have an especially great memory. Funny how the mind works.
I can't remember mine. Sort of remember percentiles. 99+ verbal, 98th Math. The latter wasn't good enough for my major. it turned out. But this was in the later 1960s so I have the excuse that it was a long time ago.
> I think the SAT scoring algorithm has been adjusted over the years such that an 800 math score can be achieved despite failing to answer every question correctly.
Only if the question you miss is contained within one of the experimental unscored sections. 800 math is not a particularly high score.
Are the data for these graphs/correlations controlled by race?
Not the first graph, but the URM graph is.
Notice that the SAT and GPA are not as strong an indicator of success for students who attend disadvantaged high schools. The idea that students can be overmatched lives on.
They also skated through high school and only learned good-enough study habits.
I skated through high school and never learned study habits until I went to college. That was a rude awakening.
lol. "disadvantaged"= black.
Or Hispanic.
Puerto Rican.
I would say the GPA that got me 10th in my class, but I'm worried some of you would reach through my device and start punching me. (Which a friend from another high school did!)
We were overly and misguidedly interested in class rank since there were only 60 in my prep school class, all on the right slope of the Bell Curve, but I don't remember any discussion or even reporting to us of actual GPA scores and little about actual SAT scores, which seems weird now but it was the boys' school reticent culture. Our 3 Ivies were all legacies, which also checked their bragging rights a bit, and 10 went to UVA, one of whom, the quarterback, became a multi-millionaire hedge fund manager. There was some resentment expressed that BC Calculus got a 15% bonus to AB's 10%. AB grades were lower, which was blamed on more difficult material, not the students.
I went to a very good suburban public high school. Senior year was all AP classes and it was the same 10 or 15 guys, no dames as we didn't call them back then, every period. I take pride in the fact that those guys were ranked 1-15 and I was somewhere closer to 100.
In the 21st century, the advanced math and science classes have more females than male students. Richard Reeves has written about this.
"I was somewhere closer to 100"
If it was ever mortifying to be at the bottom of your senior year classes, I hope your ambition was rewarded in college.
Two lovely girls from our sister school drove several miles for BC Calculus, AP Physics, and a lot of male attention. They wisely sat in the back row.
Thanks. Not mortify. I was proud of it. All the other guys were grade grinds. I focused on what I found interesting and intellectually stimulating. I transferred in second year and none of my classes transferred as honors, so no way I was cracking the top twenty students anyway. I went to my first choice college (or Institute of Technology) and was happy. My parents didn't hound me about grades. Dad thought the point of college was education, not job training, and I agree.
I find it infuriating that these tests can't do a better job of separating kids out at the top end. A few years ago the PSAT was redesigned and basically made easier. I don't recall the exact numbers, but it was essentially as if they chopped off the hardest questions. So the test became more of a fluke.
(The PSAT is very useful for doing Sailer-like comparisons because the National Merit scores are separated out by state, so that you get a quick comparison of the cutoff scores for the top 2% all across the country; students can only qualify for National Merit once, so it's not as gamed as the ACT/SAT by kids taking it ten times; and the National Merit Semifinalist lists are published, so people such as our host can go through and analyze the names for ethnic information.)
At some point IQ probably doesn't tell you that much about who's going to be the best doctor or lawyer or politician, but it doesn't seem like a coincidence that the very, very best test takers seem to end up as renowned math or physics professors. If nothing else, you could reduce the amount of time test-takers have, but that would hurt kids' feelings, I guess. (Don't even get me started on the people who manipulate the system to get their kids made-up ADHD diagnoses for the Adderall and added time.)
This follows Kingsley Amis’s Lucky Jim rule: “There was no end to the ways in which nice things are nicer than nasty ones.”
I recently read "Lucky Jim." I came away thinking the titular character was more nasty than nice.
I was a very lazy student in high school and never took a school book home. I had a fairly high GPA but that was in a Prince George's County MD public school. Football and nights-out with my friends was more important.
And there it is.
For many careers, football and friends were good uses of your time. Making yourself agreeable is more important than subject-verb agreement.
I had a good time and wouldn't trade it. I'm still friends with several of these boyhood friends over fifty years later.
> Making yourself agreeable is more important than subject-verb agreement.
I think you're underestimating how much it will bother everyone you interact with if you mess up on subject-verb agreement.
However, nothing you do in school is going to affect your subject-verb agreement in any way.
Nonsense. Practice make perfect.
There is definitely quite a bit of grade inflation these days - aside from the bonus for honors/AP classes (I don't seem to recall that being in effect 30 years ago, you just got the benefit of that designation on your transcript), lots of schools allow partial credit for late assignments or extra credit assignments to prevent disorganized students from tanking their grades, neither of which were available at my high school. I assume there is still some discount or bonus to a transcript based on the admissions office's knowledge of how decent of a HS the student attended, which I totally believe was the case in my day based on my mediocre GPA from a selective high school and the schools I was admitted to.
It also highlights how soft affirmative actions initiatives like offering automatic admission to state schools for students in the top 10% of a HS class is a massive mistake. There is a gaping chasm in the quality of even a 50th percentile student at my former HS versus that at the local public urban achievers academy, and the latter would get killed at my state's flagship public university. I remember when Georgetown started offering admissions to DCPS students in the top decile over 20 years ago and the WaPo would periodically do a story on these kids and it was always the same - they were struggling mightily and realized that the coursework required for a 4.0 from Dunbar HS was a joke compared to virtually all of their non-DC peers former schools. This was prior to the era of equity so I assume most of them ended up withdrawing and going to Bowie State or something, whereas today they would be shepherded to graduation through some absurd degree program.
Bowie State was a good way fifty years ago for bright, but far from genius, blacks to get jobs in the lucrative Prince George's County public schools. I imagine it remains so.
I suspect that one of the reasons that Texas has become more conservative politically despite the changing demographics is the top 10% rule. Forcing every high school in Texas (public and private) to rank their students from 1 to the bottom lets students know where they stand in the pecking order. And all of the students and their parents can see the benefits of affirmative action when the child of two black white-collar professionals gets into an Ivy League/Ivy Like school when ranked below white and Asian-American classmates who have zero chance of getting into an Ivy League.
PSAT and SAT/ACT have always been IQ tests. In part, that's why they are hyped/standard: the Deep State wants to identify early those folks who could be useful assets or else troublemakers.
The PSAT in particular seems to the Deep State's game. Despite "not counting" towards your actual SAT/ACT score, they give prizes and recognition to those who score highly on them. Why would a practice SAT exam merit such distinction, when it "doesn't count"? Likely because there's a correlation to those naturally gifted folks who do well on the PSAT and raw intelligence --- i.e. good places to start if you're looking for a CIA or NSA asset.
Kids who are worried for college will study for the SAT/ACT, which skews their scores up. But kids who are just naturally smart will score best on the PSAT. For example, Hillary Clinton scored very highly on her PSAT, so much so that Vox Day believes she is the smartest person ever to run for president in the last 100 years.
We'll need to see receipts for anything about the brainpower of Hillary or Bill or Trump or....
Don't sleep on Bill or Hillary's brains. You don't get to the top of the game like they have --- without convictions ---without smarts.
As for Trump, being a self-made billionaire who won 3 presidential election after becoming a hit TV star speaks for itself.
I'm not saying they're not smart, I'm saying few politicians are trustworthy about themselves, and the Clintons aren't two of them. And Hillary smarter than Hoover, Nixon, and Romney? In 30+ years, I've seen little evidence for that--or much wisdom--but no way I'm looking for a light under her bushel.
What I really found interesting were the subtle differences of the results in figure 2: To recapitulate
1: students with same SAT performed worse at college if they went before to a bad high school
2: students with same SAT performed worse at college if they were underrepresented minority
But: bad high schools flattened the regression slope (from 0.11 to 0.10). On the other hand, being underrepresented minority appeared to depress the performance level while not affecting the regression slope. This difference - deviation in regression slope versus basic level should (and can be easily) tested with post-hoc statistics. One would only need to slightly modify the test regression formula to include effects of the high school (or minority parameter) not only on the basic level but also on slope. IE instead of Y = a + b + c*X, change to Y = a + b + c*X + d*X, where b and d are coefficients to model the effects of the high school (or minority parameter) on basic level and slope, respectively.
I found the apparent differences (slope vs basic level) could suggest two things: students visiting a bad high school don't learn how to thrive in a rich academic environment as well as those from a good school. And also - SAT possibly doesn't pick all of the deficit in cognitive thrust in underrepresented minorities.
What is cognitive thrust? I can't wrap my brain around the term.
Underlyer for college GPA performance. I admit I partly wanted to use a different phrase so I invented it out of whole cloth. But there is more to it. Usually, all parts of intelligence correlate strongly with each other - that's why the psychometrics results over the last 100 years suggest something general - usually abbreviated as g. But here in figure 2B, the deficit in College GPAs stayed more or less the same across the whole spectrum of different SAT results. In other words - the general deficit did NOT correlate with the SAT results.
Or is it just admissions mismatch--the brainier disadvantaged students were DEI-ed into schools over their head? The absolute numbers probably aren't that high, so the top schools rob much of the 2nd tier's, and all the way down.
It was not an admissions mismatch because the comparison is between minorities and non-minorities with the SAME SAT score. Based on that SAT score, underrepresented minorities do worse with college grades.
it could be an admissions mis match if urm's with score x get into a better university than non-urms with score x.
It would have been a terrible methodical error to simply allow GPAs from one college to be counted in absolute terms together with GPAs from another college one on one. Because some universities are tough, others suffer horribly from grade inflation. One could of course correct this by balancing GPAs and SATS center of gravities and standard deviations across different colleges.
But yes, if one didn't do such a correction, that methodical error would be a good explanation.
Another analysis would show that if one is the first in the family to attend college, the SAT/ACT has less predictive power no matter the race. A school like UC-Berkeley depends upon the college educated parents to do the real counselling and college prep.
Yep - that's in the appendix figures 2B&C. Did you read the paper as far as that? Anyway, there, the effects of coming from a poor family (Appendix fig 2B) or from a non-academic family (Appendix fig 2C) both reduced the slope between college GPA and SAT.
For some imbecillic reason, authors didn't do statistics into the figures, but the effects were eye-catchingly profound. So the effects were like that of disadvantaged schools, only a factor of 2x or 3x stronger. Like with bad schools the effect was probably due to their defective social environment unable to really challenge the exceptionally gifted.
I think "Matilda" 1996 film by Danny DeVito deals with this issue too.
IMHO it is less about counselling or college prep. Growing with academics and going to academically good schools enhances human flourishing because the gifted person learns not to rest on his laurels.
> Moon Unit Zappa’s grandpa seems likely to have been of use
Her dad was certainly an odd ball and had an odd career in that he was a famous popular musician who never had a hit song. Marc Maron idolized him to the point where he adopted his look and dated his daughter
Zappa was a real Leader of Men type for other musicians. They'd even stay off drugs just to be in his band.
Zappa hosted SNL in 1978 and was so obnoxious that he was never asked back; he committed the cardinal sin of not playing along with the fiction that the players memorized their lines and would make it blatantly obvious he was reading from cue cards
As a teacher I can tell you that grades are mostly bullshit; if I gave the grades my students earned then my life would be much more difficult, perhaps to the point of losing my job, so I give inflated grades. They are only accurate in the sense that I don't jump one student over another; a list of my students ranked by grades would be the same under a bullshit scale and an actual scale
That's why statisticians prefer non-parametric (rank) statistics over parametric ones.
One of the golden rules of education is one can either have high standards with high failure rates or low standards with low failure rates. Politicians, school administrators, etc. have found that life is much easier with low standards and low failure rates.
When I was attending a DoDEA high school, I was a cheerleader and one of our cheers to razz the opposing teams was: "Cornbread! Chicken! Rice! Peas! We got higher SATs!"
I wonder how many, if any, civilian high schools would taunt their football rivals that way.
DoDEA school usually outrank any state in NAEP scores by race (except for whites in D.C.)
Most child of military members do not attend a DODEA school. They are a legacy of the segregated school systems in the South along with some international locations.
And according to wiki, there are only two DODEA high schools in the U.S.
And the DODEA schools are a smaller school system that the public school systems of North Dakota or Wyoming.
For any actual human beings who may be interested --
DoDEA Fact Sheet:
https://dodea.widen.net/s/hwlwrrdfc5/dodea-by-the-numbers-placemat-sy-22-23
DoDEA sports program:
https://www.dodea.edu/europe/europe-student-athletics
As a child in an FDNF family, I received an outstanding education while growing up almost entirely in the Pacific and Europe thanks to the Department of Defense. And, yes, I was inculcated in Americanism and civic nationalism.
Oh, the horror! _The horror!_
What's FDNF?
Two minutes:
https://youtu.be/6KPs70is7Eo?si=wQvbhyUl4QJOUmqn
:) Very pleased.