I didn't see it in the text but Linda Gottfredson documented a case where government tampered with tests to make them show smaller gaps between races and in the effect make them nearly useless.
Racially gerrymandering the content of police tests to satisfy the U.S. Justice Department: A case study
'Bias is a major problem' is a major understatement. "In exhaustive detail Heather Mac Donald demonstrates that allegedely disadvantaged students (especially if black) have in fact long been the beneficiaries of a raft of racial preference policies whereby they gain admission to elite institutions with far lower entry qualifications than white or Asian students. In 2003 it was “disclosed that Berkeley had admitted 374 applicants in 2002 with SATs under 1000 – almost all of them students of colour – while rejecting 3,218 applicants with scores above 1400”. At Arizona State University in 2006, white and black students with the same academic credentials had respectively a 2% and a 96% chance of admission. She refrains from driving home the obvious corollary that the white and Asian applicants were the real victims." https://grahamcunningham.substack.com/p/how-diversity-narrows-the-mind
Your footnotes are more edifying than many people's full posts. Also the USMLE Step 1 pass/fail thing is a serious problem. Unless we change course, the downstream effects are both predictable and (very likely) bad.
Something will have to give at some point. It would be crazy to have a bunch of trainees taking their first real test once they're already in residency, or perhaps never really taking a serious test. I suppose one potential benefit of all this is that it will shorten the time until a critical mass realizes "ok med ed is broken and needs to be redone from the ground up". Something akin to the Flexner report of the early 1900's.
I'm glad you mentioned the Flexner Report. The main way I've seen it brought up in the past few years has been people attacking it for reducing the number of doctors. None of those attacks mentioned doctor quality!
About a decade ago when I was in med school I read a lot of history of that era, but it's all pretty foggy now. I suppose you could legitimately criticize early efforts to standardize things insofar as any form of standardization in a system may inadvertently weed out some creativity and dynamism, but then again having a lot more well trained physicians who know some science probably makes up for that. Perhaps there were some aspects of it that we now see as wrongheaded due to the Progressive Era influence of the time? But overall I'm convinced that Flexner or at least something like it was necessary given how many (almost-literal) snake oil salesmen and quacks were running around.
Merit is the only intelligent method for selection. Civilizations that are not based on merit will fail. Case study Western Civilization...it is in the final stages of failure.
> India and Bangladesh, where meritocracy has been superseded by a quasi-ethnic spoils system
The link is just for India, not Bangladesh. It's also my understanding that Bangladesh has much less population structure than India, so I'm surprised to hear that about it.
In Brazil we only had tests as an admission to universities, yes we have major disparities in Education, if you have cash your kid goes to private school, otherwise you are stuck at public schools.
Another disparity (and there was a valid point) was that public schools are 100% free, but to get in you had to pay a lot for a good quality education - so the government should invest more in the k-12 and make people pay for public schools (or at least partially), needless to say that left leaning governments and union based professors vetoed this idea.
One public school that was great was the military school, it would teach the contents of regular education plus all the military part. They are popular , but of course demand is higher than supply. Most students would go on and pass the hardest tests.
Instead of making public schools better, the left leaning governments adopted the American equity standard and now there are quotas for kids coming from public school, quotas by race (which became a huge point of controversy a lot of people are mixed race), etc -- the results are obvious and I have many friends (university professors) who tell me how the quality declined and there are less discussion and they had to adapt some of the content. So half of the class is separated for students coming from quotas, the other half come from students passing the tests.
Side note: before the quotas, Asian (Japanese) would typically dominate the rankings, and get into the top universiities.
I think you need to be careful intepreting the Card and Giuliano study on universal screening. The main reason that universal screening increased the fraction of economically disadvantaged students in the gifted program was that there was a significantly lower threshold applied to these students for entry into the program ("Plan B"), a key fact that was somehow left out of the abstract. Looking at figure 4 from the paper, I see no evidence that teachers were failing to identify gifted students at different rates for richer vs poorer students.
I agree that universal screening is good, but this study should not be taken as evidence that teachers are biased against noticing high ability in poor or minority students.
There were plenty who were identified above the traditional 130-point threshold for plans A and B, but also notice the difference in the y-axes for each plan.
Yes, I noticed the different axes. The teacheres failed to recommend many bright students in both plans, but I don't see much difference in the rates at which they failed to do so. For example in the 130-133 group it looks like teachers recommended around 60% of the kids from both groups.
No, the relevant finding is that students were missed regardless of threshold even if the Plan B students were overall less gifted. Split by race, the largest effect was for Hispanics, who should be the most neglected for reasons to do with language learning, consistent with the ELL and parental language results.
I didn't see it in the text but Linda Gottfredson documented a case where government tampered with tests to make them show smaller gaps between races and in the effect make them nearly useless.
Racially gerrymandering the content of police tests to satisfy the U.S. Justice Department: A case study
http://www.udel.edu/educ/gottfredson/reprints/1996gerrymandering.pdf
Classic!
Indeed, but that was twenty-eight years ago. Imagine it now; it must be a hundred times worse.
'Bias is a major problem' is a major understatement. "In exhaustive detail Heather Mac Donald demonstrates that allegedely disadvantaged students (especially if black) have in fact long been the beneficiaries of a raft of racial preference policies whereby they gain admission to elite institutions with far lower entry qualifications than white or Asian students. In 2003 it was “disclosed that Berkeley had admitted 374 applicants in 2002 with SATs under 1000 – almost all of them students of colour – while rejecting 3,218 applicants with scores above 1400”. At Arizona State University in 2006, white and black students with the same academic credentials had respectively a 2% and a 96% chance of admission. She refrains from driving home the obvious corollary that the white and Asian applicants were the real victims." https://grahamcunningham.substack.com/p/how-diversity-narrows-the-mind
Your footnotes are more edifying than many people's full posts. Also the USMLE Step 1 pass/fail thing is a serious problem. Unless we change course, the downstream effects are both predictable and (very likely) bad.
Agreed. Unfortunately, I think we might be in for more examinations switching to pass/fail.
Something will have to give at some point. It would be crazy to have a bunch of trainees taking their first real test once they're already in residency, or perhaps never really taking a serious test. I suppose one potential benefit of all this is that it will shorten the time until a critical mass realizes "ok med ed is broken and needs to be redone from the ground up". Something akin to the Flexner report of the early 1900's.
I'm glad you mentioned the Flexner Report. The main way I've seen it brought up in the past few years has been people attacking it for reducing the number of doctors. None of those attacks mentioned doctor quality!
About a decade ago when I was in med school I read a lot of history of that era, but it's all pretty foggy now. I suppose you could legitimately criticize early efforts to standardize things insofar as any form of standardization in a system may inadvertently weed out some creativity and dynamism, but then again having a lot more well trained physicians who know some science probably makes up for that. Perhaps there were some aspects of it that we now see as wrongheaded due to the Progressive Era influence of the time? But overall I'm convinced that Flexner or at least something like it was necessary given how many (almost-literal) snake oil salesmen and quacks were running around.
Merit is the only intelligent method for selection. Civilizations that are not based on merit will fail. Case study Western Civilization...it is in the final stages of failure.
> India and Bangladesh, where meritocracy has been superseded by a quasi-ethnic spoils system
The link is just for India, not Bangladesh. It's also my understanding that Bangladesh has much less population structure than India, so I'm surprised to hear that about it.
In Brazil we only had tests as an admission to universities, yes we have major disparities in Education, if you have cash your kid goes to private school, otherwise you are stuck at public schools.
Another disparity (and there was a valid point) was that public schools are 100% free, but to get in you had to pay a lot for a good quality education - so the government should invest more in the k-12 and make people pay for public schools (or at least partially), needless to say that left leaning governments and union based professors vetoed this idea.
One public school that was great was the military school, it would teach the contents of regular education plus all the military part. They are popular , but of course demand is higher than supply. Most students would go on and pass the hardest tests.
Instead of making public schools better, the left leaning governments adopted the American equity standard and now there are quotas for kids coming from public school, quotas by race (which became a huge point of controversy a lot of people are mixed race), etc -- the results are obvious and I have many friends (university professors) who tell me how the quality declined and there are less discussion and they had to adapt some of the content. So half of the class is separated for students coming from quotas, the other half come from students passing the tests.
Side note: before the quotas, Asian (Japanese) would typically dominate the rankings, and get into the top universiities.
It is crazy for me to see what's happening at major universities, specially the UC system. I feel that they forgot what made them great.
Side note: I recently watched Oppenheimer and one can't stop to draw a parallel and think how this project would be in today's standards.
1) Each scientist having to write a DEI statement
2) A quota for woman, race or whatever
In other words, the US would have lost that war.
I think you need to be careful intepreting the Card and Giuliano study on universal screening. The main reason that universal screening increased the fraction of economically disadvantaged students in the gifted program was that there was a significantly lower threshold applied to these students for entry into the program ("Plan B"), a key fact that was somehow left out of the abstract. Looking at figure 4 from the paper, I see no evidence that teachers were failing to identify gifted students at different rates for richer vs poorer students.
I agree that universal screening is good, but this study should not be taken as evidence that teachers are biased against noticing high ability in poor or minority students.
There were plenty who were identified above the traditional 130-point threshold for plans A and B, but also notice the difference in the y-axes for each plan.
Yes, I noticed the different axes. The teacheres failed to recommend many bright students in both plans, but I don't see much difference in the rates at which they failed to do so. For example in the 130-133 group it looks like teachers recommended around 60% of the kids from both groups.
Is there something I am missing?
No, the relevant finding is that students were missed regardless of threshold even if the Plan B students were overall less gifted. Split by race, the largest effect was for Hispanics, who should be the most neglected for reasons to do with language learning, consistent with the ELL and parental language results.