“While you should always cite good science, art, literature, etc, consider the need to dig deeply into the literature to diversify your citations. Don’t just engage with the most highly cited articles; show how deeply you are engaging with the literature by exploring more of published research.”
Sounds good to me, while also bolstering the need for members of marginalized groups to gain greater recognition for their work when it has made a positive contribution to their field. Win win.
How are you supposed to know if any of the coauthors of a math paper relevant to your research are an URM? Google their profile pictures? Refer a yet-to-be centrally-assembled database of URM scientists?
Engage with sources deeply, sure. But what they are advocating is not good science. It’s madness.
American scientific supremacy is not a birthright. This race to the bottom has got to stop.
An author of a paper should cite another paper if and only if the latter has merit and is relevant. The names on the latter, their relationship with the author, their affiliations, countries, seniority, race, social identifies, etc. should not matter. Academia is already polluted with “citation games” – authors citing one another unnecessarily to artificially increase citation counts for everyone involved, making their work seem more important than they really are. Academia doesn’t need “citation justice.” Citations should happen organically.
Yield is primarily a function of selectivity, early admission policies, and for lack of better word uniqueness. It is not a good measure of preference. For example, suppose half of students prefer their less selective state flagship and half of students prefer the more selective HYPSM. The kids who prefer the less selective state flagship would not apply to more selective HYPSM as a backup in case the less selective flagship rejects them. That wouldn’t make any sense, so the kids who favor their flagship don’t hurt HYPSM yield. However, many of the kids who prefer HYPSM do apply to their less selective state flagship as a backup, so they do hurt the yield of their flagship. In this example, there was equal preference between HYPSM and flagships, yet yields are expected to be completely different.
The 10 colleges I have heard of, with the highest yield as listed in 2022 IPEDS are below. Note that some of the high yield more relates to uniqueness with no similar alternative, rather than extraordinarily high selectivity. Early admission policies also play a role, which I expect relates to why Chicago is slightly above HYPSM.
Haskell Indian Nations – 90%
Naval Academy – 85%
Chicago – 85%
MIT – 85%
Stanford – 84%
Harvard – 83%
BYU – 79%
Air Force Academy – 78%
West Point – 78%
Ozarks – 76%
Traditionally test optional/blind is associated with an increase in yield, both due to increased selectivity and increased uniqueness. However, post-COVID test optional is no longer especially unique. Instead test required may be more unique.
Indeed, holistic admission can also be used to increase inequities, such as favoring LDC without making it too obvious how much they are favored. Most hooked students can at least have plausible deniability of the belief that they only got in due to their hooks, due to the relative opaqueness of most holistic admission.
Actually, posts on this thread have opened me up to what the other side is saying. I still feel very strongly that SAT math should be required for STEM majors especially at average state schools like mine. I don’t have any data to back it up other than personal experience working with numerous students who are attracted to engineering because of the job/money prospect but are ill-prepared to do the work due to serious lack of very basic math skills.
I’m not concerned with issues like “should I retake SAT if my math score is only 790 so my MIT/Caltech apps won’t be derailed.” These are high-class problems. I’m concerned with those on the other end of the spectrum, who were accepted to engineering but probably shouldn’t. An additional data point in the form of SAT math scores on their apps would be helpful to know where they are, math-wise. It saves them time and money and also helps with their mental health.
Apart from “SAT math should be required for STEM majors,” I don’t have as strong a feeling as I used to on other situations after reading through the thread, seeing the back and forth. If you ask me how useful SAT EBRW scores are for STEM majors, I would say not very, as long as they’re not alarmingly low. As for majors that are very far from STEM, I can’t imagine SAT math scores to be super useful (other than to help rank applicants at selective schools) and I don’t know about these vastly different majors to have an opinion on the value of SAT EBRW scores.
I was wondering if it possible to discuss the SAT without bringing up Jon B. I suppose it’s a victory that we got to well past 500 posts without his name coming up.
Jon B frustrates me to no end. But before I get into criticizing him, let me say at least one nice thing about him.
Jon B knows how to make pretty graphs, like the following:
In this graph, he is showing that high ACT scores are predominantly achieved by people with higher incomes. And from pretty pictures like this, he writes that standardized tests are useless at anything except measuring the income of the test takers.
The problem is that Jon B keeps making the same statistical mistakes over and over, despite it being pointed out to him. He handwaves away the concept of looking at scores in context, saying that diamonds in the rough don’t exist, even though universities that have looked at this problem (and have far more statistical expertise than he does) have concluded the opposite, for their particular university.
It’s interesting that Jon B points to Saul Geiser’s rebuttal, discussed earlier. I pointed out that linear regression is very a useful tool that has limitations, because among other things, it assumes linearity. We can see from Jon B’s graph above that the relationship between test scores and income is clearly not linear, and that Saul Geiser’s assumption of linearity in his model was a mistake. It further justifies the UC Academic Senate’s approach of looking at outcomes visually for various demographic criteria, such as income and race.
Even his own data shows that ~10% of top scores are from low income folks. (just eyeballing the chart since there is no X-axis). And that 10% represents diamonds, IMO.
That is not his position though (see full blog post below), to quote him he says ‘that’s not entirely wrong.” It gets at something that I’ve asked several times on this thread…when scores are required, we know that FGLI/URM/dis-privileged students just don’t even apply. How do we address that if schools went back to requiring test scores? Jon says it this way: “It’s great that one student gets a chance. But what about the 500 who don’t? ” His position is that the “juice isn’t worth the squeeze.”
Big picture, I just don’t understand why people care how a given school chooses its students. I also don’t understand why some think they know better than the board/admin/enrollment leaders from a given school.
Anyway, here’s Jon addressing the diamond in the rough theory. From his blog post:
Point 2: The “Diamond in the Rough” (DITR) theory. This heart-warming approach posits that there are low-income and/or first-generation, and/or students of color out there who will be disadvantaged if tests go away, because the tests help them get identified as bright. And of course, this is not completely wrong, but it’s only true because a) some college admissions officers don’t know much about tests, and b) they do know that people ask them about average test scores all the time, again (here comes the absurdity) even though they vary so strongly with wealth. If you get the impression of a dog chasing its tail, you wouldn’t be wrong.
Not coincidentally, The DITR aphorism is also designed to appeal to people who don’t think critically.
Some propositional logic might be in order. It was perhaps the best class I ever took in college, but I don’t expect you to dive deeply into it, unless it’s raining and you have nothing better to do. Let me outline the challenge here.
These high-scoring students are bright; the tests measure something, but again whether that something is valuable is another question.
It’s a sad state of affairs in US higher ed that students from less advantaged background need the equivalent of a lottery to get a chance, but that’s exactly how things shake out. And the corollary to this is that there are loads of smart, capable students with lots of maturity and abilities no test can measure at the other end of the standardized testing scale who get overlooked because of the tests. That’s the cost no one ever considers when they spout the “Diamond in the Rough” argument: Human potential that goes unrecognized because of a test, developed by private companies whose only obligation is to themselves. It’s great that one student gets a chance. But what about the 500 who don’t?
Back to logic. When you say, “high-testers are smart” many people jump to an inappropriate conclusion that says, “non-high-testers are not smart.” That’s the problem: All S is P in no way implies that All Non-S are Non-P. (All poodles are dogs, but all non-poodles are not non-dogs, in case you’re having some trouble deciphering this.)
Tests like this have low rates false-positives. But they also have high rates of false negatives. That’s a problem, but it’s also the reason (along with the income correlation) super-selective institutions like them: With few slots to give out, they want some assurance of “it,” whatever “it” is. And the flotsam and jetsam of low-income, first-generation, students of color left behind are just the cost of doing business.
The perception that high SAT averages in your freshman class means your students are “smart,” when it can also mean you’ve built a gate to keep poor people out, is a double whammy of brand goodness for the nation’s “elite” institutions. As I wrote once, “Perhaps ‘elite” really means ‘uncluttered by poor people’.”
At the same conference (maybe not the same year, though) where I heard Harry Brighouse, I heard the Dean of Admissions at an Ivy League institution say–in public, in a presentation–that they spent “all their time” looking for low-income students, which they defined as “family incomes of under $60,000.” (Median family income at the time was about $55,000.) Unfortunately, in the words of this dean, and despite these Herculean efforts, they “just could not find enough low-income students who could do the work at the university.”
Two things struck me about this statement: One was the fact that this university was notorious for not enrolling many low-income students, so how could the dean know that, but more concerning was the ease with which it was uttered, as though it had been floated out in meetings so many times at this university without being challenged that it seemed as natural as the leaves returning to the trees in the spring.
and his hyperbole it over the top, which brings into question his theory.
He works at an excellent Uni which has an 85% acceptance rate, essentially open enrollment. So, suggesting that 500 don’t’ get an ‘opportunity’ defies common sense. (Perhaps facts were not in his propositional logic class.).
More importantly, where is the research that 500 don’t apply? And, don’t apply to where? Sure, maybe some Oregon kids won’t apply to MIT, but will still receive a fine engineering education at Oregon or Oregon State.
We must be talking about different students, I am talking about low income, underprivileged students. Oregon State is not affordable for low income students, maybe for some in-state but certainly not for any from OOS. Many low income students can not afford their state flagship (if they can get admitted). The point is that a four year residential college experience is not possible for low income students unless they hit the lottery by being accepted to a meet full need school.
Not sure the preeminent studies in that area, I am sure a google search will turn them up.
perhaps true, but such meets-full-need school are also a lottery for those that score a 1400+.
Another way to look at it is that instead of lobbying his state for more funding so that his Uni can meet full need of its own residents, Jon complains about a barrier (in his mind) that other schools use which inhibit Oregon residents to obtaining full need at at OOS live-away school.
And even if they do apply to MIT, in the absence of extraordinary extracurricular accomplishments expected of White and Asian kids on top of their stellar stats, what exactly is admissions committee supposed to consider in order to take the chance on a kid from an unknown high school without established high standards of academic rigor, under a TO regime?
Fascinating (and fitting) how some people on CC instantly try to discredit Beockenstedt by attacking his qualifications, despite his having been involved in admissions for decades. Apparently, the opinions of AOs at a certain universities trump all the other universities that agree with Beockenstedt, even when those universities make it crystal clear their analysis only applies to their own situation. Do I have that about right?
And it looks like Beockenstedt favorably refers to Vikdor’s critique, which you admit is mathemaltically sound . . .
So you two agree on that, at least. Yet then you go on to blatantly misrepresent his position, as evidenced by @Mwfan1921’s post.
But even setting Beockenstedt aside, if Vigdor is correct then you and others are blatantly over-exaggerating the value of these tests.
