Standardized admission tests have taken a reputational beating over the last few decades, particularly the high-stakes tests heralded by the No Child Left Behind Act. That's in part because white,...
Standardized admission tests have taken a reputational beating over the last few decades, particularly the high-stakes tests heralded by the No Child Left Behind Act. That's in part because white, wealthy test takers tend to score higher. This article argues with data showing correlations between scores and college performance across racial subgroups that the tests are only measuring an existing disparity. The author argues that the tests present a valuable opportunity for disadvantaged teens to prove their capability to colleges. Later in the article, the author discusses the need to challenge progressive ideas when data contradicts them
Moving into my own opinion and experiences:
I agree with this article that there should continue to be a place for standardized test scores in college admissions, if only because I have myself as an example (I had mediocre high school grades because of poor executive function, but got a great SAT score, which got me into the college program I wanted. I graduated and now work in my field.). I don't agree with all the author's conclusions (he makes a throwaway claim that covid school closures were a mistake, linking to another article... But that's another comment all together), but I think he does have a good point to be wary of dogma without evidence
Standardized tests, of course, will always have the problem of coaching and teaching to the test. However, free and high quality self-guided SAT prep is now widely available online. I've frequently seen claims that it's difficult for colleges to attract high quality, diverse students. I wonder if a program could be set up to identify disadvantaged students who are scoring especially well on that prep, to connect them to free personal tutoring and college admissions guidance?
Standardized tests in college admissions are also a bit of a self-fulfilling prophecy: High SAT scores (especially pre-2016 when there was more game strategy) were mostly an indicator of how well...
Standardized tests in college admissions are also a bit of a self-fulfilling prophecy:
High SAT scores (especially pre-2016 when there was more game strategy) were mostly an indicator of how well you take tests. People whom are good at taking tests will also earn high GPAs in college, especially when tests are a huge portion of the grades.
That does not mean college was not immensely valuable to that giant swath of college graduates that were not chasing the 3.6+ GPAs. Or even to the people who didn't end up graduating.
I'm wondering if, now that we have computers, we could develop some sort of system to prevent teaching to the test. I currently have no idea what that may look like, but I'd wager at a minimum it involves an exponentially larger question pool and randomized weighting of scoring.
I think that's too much of a simplification. I disagree that it's "mostly" an indicator. Rather, it is a significant but small component. It's not really possible to cram 11? years of schooling a...
High SAT scores (especially pre-2016 when there was more game strategy) were mostly an indicator of how well you take tests. People whom are good at taking tests will also earn high GPAs in college, especially when tests are a huge portion of the grades.
I think that's too much of a simplification. I disagree that it's "mostly" an indicator. Rather, it is a significant but small component.
It's not really possible to cram 11? years of schooling a few weeks or months before a test. Prep courses are very popular with affluent students, and they're all about teaching students test-taking skills. Various studies show that test prep improves SAT scores by roughly 15 to 30 points, according to Slate and NYT's Jay Kang. In admission for very competitive universities (Ivy League and Tier 1) where nearly everyone has a top decile score, it's an important edge; outside it, it's really not much. So the impact of one's test-taking skills may move one up or down by one decile or percentile, but they can't turn a 1100 into a 1550.
The downside is that it penalizes kids who grow up with parents who don't keep books in the house. I wonder what will happen to the vocabulary section. When I took the test I benefited greatly...
The downside is that it penalizes kids who grow up with parents who don't keep books in the house.
I wonder what will happen to the vocabulary section. When I took the test I benefited greatly from having read books at my grandparents house such as Ivanhoe and Last of the Mohicans. Modern published books don't include the range of vocabulary used in classics. Kids who are encouraged to read still aren't stretching with complex words and sentences the way we did for the most part
I think the penalty is necessarily 'fair' because the purpose of the test is to filter and assess college readiness. The SAT is pretty basic; it rarely has anything esoteric. If someone struggles...
I think the penalty is necessarily 'fair' because the purpose of the test is to filter and assess college readiness. The SAT is pretty basic; it rarely has anything esoteric. If someone struggles with the test, it's likely that they'll struggle in college where the expectations and pressure are much greater.
Ultimately, there are other perfectly viable and respectable life paths after high school that don't involve college.
Two thoughts I can easily express on a phone keyboard come to mind: Looks like the advantage of parents with large libraries will mostly equalize in the next generation or two. Rather, it’ll now...
Two thoughts I can easily express on a phone keyboard come to mind:
Looks like the advantage of parents with large libraries will mostly equalize in the next generation or two. Rather, it’ll now be a smaller subset who have the right books who retain the advantage.
Do you know what the functional difference between the SAT without vocabulary and a pure IQ test would be?
Tangential, but I’ve been told that my fanfic reads like someone ran a thesaurus through a mad lib because of my enjoyment of using new words I learn from Mark Danielewski novels (and also appending affixes to make up words)
If you’re teaching good material, but no longer teaching to the test, then the test needs to be modified because it’s a bad test. The whole point of tests is to accurately assess how well someone...
If you’re teaching good material, but no longer teaching to the test, then the test needs to be modified because it’s a bad test. The whole point of tests is to accurately assess how well someone has absorbed or retained knowledge.
The issue of testing the test being a poor curriculum isn’t an indictment against testing, it’s an indictment of a bad curriculum, including the test for that curriculum.
You're right that that's not an argument against tests. Let me make the argument against tests. But media are not neutral!—the medium is the message—that something is a test says something about...
You're right that that's not an argument against tests. Let me make the argument against tests.
The whole point of tests is to accurately assess how well someone has absorbed or retained knowledge.
But media are not neutral!—the medium is the message—that something is a test says something about it and it's incumbent upon us to consider if that is good or valuable or useful. Here are my observations about tests:
For teachers, tests provide a valuable quantitative metric of, in particular, which students are being 'left behind'. It's valuable because teachers have heavy workloads and are asked to teach large classes, and therefore lack the ability to have a close personal relationship with each student and a sense of where they are at and what they need. Tests may improve matters slightly, but the underlying problems are still there, and the addition of tests may serve to paper them over. (Chris Crawford discusses this a bit, in 1993—I don't necessarily agree with anything he says, but it is interesting.)
For students, tests create an adversarial relationship between themselves and the teachers (and the institutions they represent). It's not just the notion of external assessment—which is obviously valuable and need not be infantilising (see for example peer review—ok, obviously there are problems with that, but it is a lot more constructive)—it's the details, the medium, the structure around it. The distrust. I had a lovely biology teacher in high school who would offer to do oral make-up exams. Out of a somewhat fundamentalist desire for fairness, he would pretty much just note down what you said and grade it against the same rubric he used for written exams. But the shift in medium nevertheless caused a dramatic shift in tone and relationship, and people who struggled with or stressed over written exams often did much better in the oral makeups.
For students, tests, as other graded assignments, act as extrinsic motivation and pressure, destroying or at least dampening intrinsic motivation and interest. Unlike some other sorts of assignments, however, there's little that's creative or constructive about tests. At best, they can be tolerable, or funny. But they can't really have genuinely redeeming qualities.
Which is another problem: they have extremely limited scope. All they can test is what can be seen on a few sheets of paper given a few hours' turnaround time. If all you want to know is whether somebody really has no idea of what's going on, great, maybe; but, per above, that's not that interesting. I can poop out a half-baked essay comparing Aristotle to Sartre in a half-hour (it got full marks), but every really good idea I've had has taken a bit longer than that to work through.
Having such limited scope also means that tests' results are prone to be very volatile, because you're sampling only a small slice of a person. How much sleep did they get? Did they eat breakfast? Did their cat die last week?
I am not suggesting a solution here (though I'm also not proposing there isn't one). I am just saying that we must take the problem seriously.
I'm not actually convinced of this. I am someone who naturally does super well at these types of standardized tests, and I'm also someone who had a high GPA overall in college. But none of the...
High SAT scores (especially pre-2016 when there was more game strategy) were mostly an indicator of how well you take tests. People whom are good at taking tests will also earn high GPAs in college, especially when tests are a huge portion of the grades.
I'm not actually convinced of this. I am someone who naturally does super well at these types of standardized tests, and I'm also someone who had a high GPA overall in college. But none of the tests I took in college, even among the courses that had conventional tests at all, were particularly similar to standardized tests. Most were fundamentally different in their structure and requirements, and that's before we take into account how few of my courses even had tests like that at all. Final projects or papers were more common in my courses.
I'm not necessarily advocating for getting rid of SATs, but I have not found the skills that got me my high scores particularly applicable outside of standardized tests themselves. Some of the skills that benefitted my standardized test performance, such as willingness to quickly skip questions that seemed unintuitive to me, have probably negatively affected my performance in college and beyond.
Maybe part of the test is the ability or willingness to adapt to a different structure with unique requirements. Then the requirements would have to be unrelated to most real-life situations so no...
Maybe part of the test is the ability or willingness to adapt to a different structure with unique requirements. Then the requirements would have to be unrelated to most real-life situations so no student would have an advantage due to experience.
I'm not some huge fan of the SAT, my experience is similar to yours, it's just that your comment made me start wondering if the lack of real-life applicability is intentional.
I mean... I don't think so, because standardized tests like the SAT all have set structures (hence the standardized lol) and those structures don't really differ from each other. Performance...
Maybe part of the test is the ability or willingness to adapt to a different structure with unique requirements
I mean... I don't think so, because standardized tests like the SAT all have set structures (hence the standardized lol) and those structures don't really differ from each other. Performance improves the more familiar you are with that particular test, so these tests optimize for familiarity with standardized testing, not ability to adapt to a different structure. Kids my age have been exposed to standardized testing with a similar format from a young age. My parents had me take the full SAT a year early to prepare for the PSAT (and I was my school's only National Merit finalist, which is based on PSAT results, so I can't say it didn't work!)
Even outside of the at-least-yearly standardized testing, taking tests in school was a huge part of my real-life experience prior to college. The traits that made me good at standardized tests also made me good at most tests in middle and high school. I was a good test-taker. The problem is that once I actually moved into courses past very basic early GenEds in college, these skills were basically no longer applicable at all and were often detrimental. For your theory to hold water, you'd want the opposite -- for the tests to have little association with real-life prior to the tests but an association with real-life afterwards.
I mean personally I think college as it is really has the same problem. You could make college 4 years of learning to spit tobacco into different shaped buckets and those people would still go on...
I mean personally I think college as it is really has the same problem.
You could make college 4 years of learning to spit tobacco into different shaped buckets and those people would still go on to be more successful than their peers because of just how badly a college degree gates you in the job market.
Yes and no. At this point, a BA/BS may not even get you in the door anymore. It's becoming the new high school diploma. Since so many people are getting them, they're becoming devalued (which IMHO...
Yes and no. At this point, a BA/BS may not even get you in the door anymore. It's becoming the new high school diploma. Since so many people are getting them, they're becoming devalued (which IMHO is dumb).
I've seen so many jobs posted recently (I've been helping spouse and several friends job hunt) that are barely above entry level requiring a master's degree and X years of experience. Will those postings get them? likely not, but the fact that they're asking for a master's now instead of a bachelor's is insane to earn $23/hr in the Bay Area, CA.
I'm seeing that shift in higher education, one of the biggest culprits of the masters required entry level position. The "Great Resignation" hit us pretty strongly and we both raised salaries and...
I'm seeing that shift in higher education, one of the biggest culprits of the masters required entry level position.
The "Great Resignation" hit us pretty strongly and we both raised salaries and lowered our required qualifications. To a good outcome IMO.
There are so many jobs that require higher education that don't need it. Many jobs that I've done personally that require not only a bachelor's degree, but a high GPA. There is no reason for a...
There are so many jobs that require higher education that don't need it. Many jobs that I've done personally that require not only a bachelor's degree, but a high GPA. There is no reason for a receptionist to need more than a high school education, not with the tools we have now.
Requirement inflation is almost purely a ploy to cut the applicant pile into something manageable. This is the negative side of online job sites allowing people to spam every opportunity across...
Requirement inflation is almost purely a ploy to cut the applicant pile into something manageable. This is the negative side of online job sites allowing people to spam every opportunity across the nation.
This op-ed falls into a common trap, failing to question the premises of its argument. The SAT is designed to predict how well a student will score in their college courses. When College Board...
This op-ed falls into a common trap, failing to question the premises of its argument. The SAT is designed to predict how well a student will score in their college courses. When College Board evaluates a question for inclusion on the test, their main criterion is whether it accurately predicts what a student's grades will be. The article spends the whole time defending the SAT on the grounds that it does accurately predict what a student's grades will be, but it does not question why you would want to predict that.
Here are a few other reliable predictors of college student grades:
Students get better grades if their parents went to college.
Students get better grades if their parents are wealthy.
Students get better grades if they were raised in a wealthy zip code.
Students get better grades if they attended an expensive private school.
Students get better grades if they are perceived as white, because of structural racism in the college environment.
Students get better grades in engineering courses if they are men, because of structural sexism.
Students get better grades if they don't have a major disability.
I could go on. I think it's obvious why it would be wrong to use any of those as criteria for admissions. Part of the goal of a university is to provide opportunity to the disadvantaged. It would be wrong to adopt an admissions practice which reinforces existing social inequalities, even if it does result in a student body which gets better grades.
College Board goes to great lengths to control for effects of race, gender, and class in their tests. They are entirely aware of this criticism. Personally, I think they are fighting a losing battle. I don't think it is possible to create a test of "scholastic aptitude" which does not simply reinforce existing social inequality. In psychometry jargon, I don't think the construct is well-defined. Instead of trying to sort students into the categories of will-do-well and will-not-do-well, I believe that we should direct our energy to making colleges work for all students.
Everything in the college admissions process will reinforce existing inequities though. Extracurriculars do that, GPA does that, projects/volunteering do that. The SAT does too, but it’s...
Everything in the college admissions process will reinforce existing inequities though. Extracurriculars do that, GPA does that, projects/volunteering do that. The SAT does too, but it’s relatively easier for poorer students to play it in their favor.
The reality is that college inequities are only a symptom of broader issues in the educational system, and fixing that is required to fix college issues. Unfortunately, we’ve been trying to fix it for decades, spending inordinate amounts of money in the process and in some cases making it worse.
The SAT/ACT can be studied for, and it doesn't have to be expensive. SAT prep books you can usually get for $30, there's no lack of online resources, and many schools, community colleges, etc....
The SAT/ACT can be studied for, and it doesn't have to be expensive. SAT prep books you can usually get for $30, there's no lack of online resources, and many schools, community colleges, etc. offer free SAT prep classes before or after school.
I went to a high school that is usually in the top 5 in our state and our median household income was 137k so I know a lot of the tricks.
What are the other factors? GPA. Many magnet schools or just schools in expensive areas would have connections with college recruiters - they "knew" what our "gpa" meant, and would weight it more heavily.
Our GPA was out of 100, but AP classes counted as 110, and the school covered the exam fee, so we would all spam the shit out of AP classes. By sophomore year, at least half of your schedule would be AP classes, and it would be 100% AP classes by junior year. I ended up with a 107 GPA - that's what was being sent to universities.
Classes could be rigorous, but there would always be an opportunity to "make" up your grade at the end, no matter the trials and tribulations you went throughout the class.
I can guarantee that our GPAs were considerably more inflated (I mean, they were above 100 in scale that out of 100 lol) a school with less resources.
Extracurriculars - it basically was a running joke that everyone had their own NGO. Parents would be in groupchats where they talked about the best kind of NGO to make for admissions, or how to actually have the NGO "do something" while actually doing the minimum amount of work. Of course, everyone played at least one instrument.
Our robotics teams was an incorporated LLC that not only received significant funding from alumni but also had corporate sponsorships.
We have a club that made those low ride solar car things. I'm pretty sure each one was at least $10k in materials.
Compared to all that, prepping for SAT is a much closer gap to close financially.
Not to mention, prepping for the SAT has a much lower impact on overall score than other test preparations. It's not easily gamed by cramming, unlike other tests.
Yes. I think @stu2b50 covered a lot of points I wanted to make, but I’ll still outline my thoughts below: Let’s take a hypothetical poor but bright student in a bad neighborhood of a generic Rust...
Yes. I think @stu2b50 covered a lot of points I wanted to make, but I’ll still outline my thoughts below:
Let’s take a hypothetical poor but bright student in a bad neighborhood of a generic Rust Belt city. They live with their mom who works as a CSR, in a two bedroom rental apartment. The student works after school and on some Saturdays at McBurger. They are the top performer at one of the worst high schools in the city, and their guidance counselor has said they should go to college. What options do they have for impressing admissions officers?
GPA: the school system recently got in trouble for massively inflating grades. While our student is genuinely a high B/low A student, the admissions officer will be wary of folks from this district.
Coursework: The school only offers up to trigonometry (no calculus) in math, no computer science, and few science classes.
Essays: In a vacuum, a personal essay reflecting on their challenges would be a good option to highlight their readiness and capability. Unfortunately, they are competing with essays that have similar themes, written by other students with professional writing coaches, and which will only be skimmed by admissions officers.
Extracurriculars: our student works 5 days a week, and spends the rest of their time doing homework or decompressing. The options that exist are expensive and time-consuming. Even if they put something down, one or two items versus the four or five isn’t likely to be competitive.
The SAT: they just need to take off one day from work to take the test. They can do test prep questions on their breaks at work or on the bus ride to school. They don’t need to spend a dime (borrowing a test prep book from the library, or looking at posts on Reddit or videos on YT/TikTok.
I don't know the data but test prep is one of the least resource intensive options. It can be costly but it doesn't have to be. And even poorer schools will have study guides available. It...
I don't know the data but test prep is one of the least resource intensive options. It can be costly but it doesn't have to be. And even poorer schools will have study guides available. It wouldn't surprise me if the differential is less.
I would go a step further and argue that it is not possible to create a university admissions process which does not simply reinforce existing social inequality. At the most fundamental level,...
I don't think it is possible to create a test of "scholastic aptitude" which does not simply reinforce existing social inequality.
I would go a step further and argue that it is not possible to create a university admissions process which does not simply reinforce existing social inequality. At the most fundamental level, university admissions is about deciding who is not going to be invited to attend a university. With the current US education system, even an admissions lottery would be reinforcing societal inequalities because of the cost of attending university.
Education is an investment into society. Highly educated people are good for making societal progress, improving the economy, and are well-suited to be the leaders of a society. Inequality in university admissions would be vastly decreased by improvements in the k12 public education system and a stronger public education system for free adult education (like community college). This is where I think we should be focusing our energy.
rant on scholastic aptitude
I think assessing "scholastic aptitude" is important before attending college, but we should ask ourselves what "scholastic aptitude" actually means. If you asked someone today, they might say it means "being good at taking tests" because that is how you succeed in university today–I don't think it's fair to students to admit them to a test-heavy university program if they're bad at taking tests, especially in the current system where students have to pay per class (and can even be subject to fees of ~5x standard tuition for taking too many classes in Texas).
Put another way: the primary filter for university admissions should reflect the skills necessary to succeed in that university's program.
We already see this with music schools: an audition is typically a part of your admissions package. Art schools will typically ask for a portfolio. It gets trickier when you start to look at programs like computer science, because the expectation is that you will be taught all of the necessary CS skills in the university program.
Going back to "scholastic aptitude," I think it's worth revisiting course exams as a whole. In many cases, if you fail all the tests in a class, you'll fail the class–even if you already have mastery of the course content and are just a bad test taker.
Projects are an excellent way to demonstrate mastery of a concept, but have challenges associated with them (bias in subjective grading, potential to pass off someone else's work as your own) and some programs are better suited to projects than others (computer science is a good program for projects, liberal arts perhaps less so).
Another thing to consider is what else universities can look at to determine admissions. A few I'm aware of colleges looking at are:
Extracurricular activities
These can be challenging for low-income families to participate in due to cost or time. If students need to work an after-school job to help support their family, they likely will have little time to spare for a school club in addition to their homework and home responsibilities.
It can also be challenging for families that have poor transit access and don't have a spare car for the kid to take to and from school.
High school grades
Depends entirely on the school, basically doesn't work at all for comparing against other people. Not to mention these can also reflect the exact same biases as SAT scores.
Strength of curriculum
Some schools do not offer advanced courses due to poor funding. Disadvantaged students may not have the time or resources to commit to an AP/IB/Honors course due to the additional time required to study.
Essays
I don't generally see essays called out as problematic like exams are, but I feel like they face rather similar issues (just to a lesser degree). Absolutely still prone to bias based on writing style, probably to a higher degree than multiple choice exams.
While I don't think standardized tests are particularly good measures of who should be admitted to college, I think they're one of the least bad options we have. There's plenty of free online practice (KhanAcademy even has one that adapts to how you've performed on the PSAT and their own practice exams) and plenty of low-cost SAT prep books that can also be bought used.
Bias and inequality are just so deeply entrenched in our society that essentially any system implemented at a societal scale is going to reinforce these problems. The best solution to inequality in college admissions is to address the inequality in society.
This has been long and windy and probably doesn't make any sense, but that's just my brain dump of thoughts on the SAT and university.
I have a massive headache, and I want to reread your comment tomorrow and give a real response, but I just really needed to address this part tonight Anyone who goes into a CS program assuming...
I have a massive headache, and I want to reread your comment tomorrow and give a real response, but I just really needed to address this part tonight
because the expectation is that you will be taught all of the necessary CS skills in the university program
Anyone who goes into a CS program assuming that all skills they need for a job will be taught is in deep shit. Most classes glance over early concepts, and later courses often assume skills (like, say, navigating a Linux command line) that are only ever barely touched on in class.
Many CS programs are also more theoretical, and focus very little on the sort of programming you do in the workplace -- my first internship expected me to work with REST apis, and parsing HTML, which never even got mentioned in 4 years outside of an elective web programming course that I withdrew from.
Perhaps something like nursing, or accounting might be a better example?
But yeah, sorry about that, the "someone is wrong on the internet" hurt a little
Oh believe me I agree but that was a very common opinion in my program. Playing off of what I mentioned elsewhere, everyone should have the opportunity to learn the fundamentals in a free public...
Oh believe me I agree but that was a very common opinion in my program.
Playing off of what I mentioned elsewhere, everyone should have the opportunity to learn the fundamentals in a free public education setting (both in high school and community college). But that’s just not an option for all students unless they self-teach.
I wound up self-teaching a lot of the practical concepts before and during high school and was fortunate to attend a high school with some absolutely astounding computer science teachers- they were better than nearly every CS professor I had.
I'll toss out a partial counter to that: Universities tend to also have massive amounts of financial aid available. On top of government grants and aid. A lottery process would eliminate a lot of...
With the current US education system, even an admissions lottery would be reinforcing societal inequalities because of the cost of attending university.
I'll toss out a partial counter to that: Universities tend to also have massive amounts of financial aid available. On top of government grants and aid.
A lottery process would eliminate a lot of those systemic problems. Mandating higher percentages of aid going to lower-income students and banning higher-income students from recieving performance-based aid would free up a ton.
You're not wrong, but once a lottery system is in place smoothing out the budget/loan gap is a bit easier.
How are you defining "higher income" here? No one who otherwise could have gotten a full academic scholarship should have to take out loans merely because their parents made too much money.
banning higher-income students from recieving performance-based aid
How are you defining "higher income" here? No one who otherwise could have gotten a full academic scholarship should have to take out loans merely because their parents made too much money.
Wait, there is something very off about this statement. Asian Americans have the highest mean college GPAs (the source tracks from 1990 to 2009, but we can assume the trends track similarly with...
Students get better grades if they are perceived as white, because of structural racism in the college environment.
Wait, there is something very off about this statement. Asian Americans have the highest mean college GPAs (the source tracks from 1990 to 2009, but we can assume the trends track similarly with recent high school GPA data), and they're definitely not perceived as white.
Latino Americans are significantly culturally and genetically white-adjacent yet have the second lowest mean GPA. I'm getting a little in the weeds here, but a 2015 paper published in the American Journal of Human Genetics estimates that Latinos in the US on average carry 65.1% European ancestry. Many Latinos are able to pass with some degree of whiteness — it doesn't seem to help them much.
If there were structural racism in the college environment, then its effects must be very weak and easily surmountable.
Latino Americans are not generally perceived as white. https://www.pewresearch.org/hispanic/2021/11/04/measuring-the-racial-identity-of-latinos/ I don’t know what genetics have to do with...
Latino Americans are not generally perceived as white.
As a fourth measure of racial identity, the survey asked Latinos how they would describe their race or origin in their own words. The most common responses for Latinos regarding their race in this open-end format were the pan-ethnic terms Hispanic, Latino or Latinx (28%) or responses that linked their racial origin to the country or region of their ancestors (28%). A smaller share also chose to identify their race or origin as American, either as a single answer or in combination with another response (11%), 9% identified their race as White, and 9% mentioned another racial group such as Asian, Black or Indigenous.
That's true, but they are perceived as whiter than Asians. Some are perceived as somewhat white; and some present as unambiguously white. Genetics determine appearance. Whether or not race has any...
Latino Americans are not generally perceived as white.
That's true, but they are perceived as whiter than Asians. Some are perceived as somewhat white; and some present as unambiguously white.
Genetics determine appearance. Whether or not race has any biological meaning, people racially classify others by their appearance.
An average 65.1% European ancestry across the Latino American population means that a significant number of Latino American individuals present as partially or very European/white.
At the very end of the "very" spectrum is my best friend from college. He's Latino American: Colombian. His heritage is mixed indigenous and European. He looks and passes as totally white. He identifies as a Colombian American. I doubt that anyone has visually recognized and treated him as Latino in a discriminating manner.
In college, I met a number of Latino Americans like him who I wouldn't have clocked as such until they told me about their heritage and how they identified as such.
Anyway, my original thought is that the statement that students get better grades when they're perceived as white stands at odds with the reality. If it were true, we'd see white students with the highest average GPA; Latinos, who are not generally perceived as white but some of whom are perceived as white and others are perceived at least whiter than Asians, with the second highest; and Asians in third.
I think that’s more complicated than simply being perceived as “whiter” and I don’t know if I’d feel comfortable characterizing it in those terms at all. This is an unwarranted extrapolation from...
That's true, but they are perceived as whiter than Asians.
I think that’s more complicated than simply being perceived as “whiter” and I don’t know if I’d feel comfortable characterizing it in those terms at all.
Anyway, my original thought is that the statement that students get better grades when they're perceived as white stands at odds with the reality. If it were true, we'd see white students with the highest average GPA; Latinos, who are not generally perceived as white but some of whom are perceived as white and others are perceived at least whiter than Asians, with the second highest; and Asians in third.
This is an unwarranted extrapolation from what @alden actually said. They brought up 7 separate factors, but you have created an argument around how one single factor is not totally determinative when we shouldn’t expect it to be in the first place, particularly when four of the other points are clearly correlated with the results:
Students get better grades if their parents went to college.
https://nces.ed.gov/programs/raceindicators/indicator_red.asp#
The 150 percent graduation rate was highest for Asian students (36 percent), followed by Pacific Islander students (34 percent), White students (32 percent), Hispanic students (30 percent), American Indian/Alaska Native students (27 percent), students of Two or more races (25 percent), and Black students (23 percent).
Students get better grades if their parents are wealthy.
Standardized admission tests have taken a reputational beating over the last few decades, particularly the high-stakes tests heralded by the No Child Left Behind Act. That's in part because white, wealthy test takers tend to score higher. This article argues with data showing correlations between scores and college performance across racial subgroups that the tests are only measuring an existing disparity. The author argues that the tests present a valuable opportunity for disadvantaged teens to prove their capability to colleges. Later in the article, the author discusses the need to challenge progressive ideas when data contradicts them
Moving into my own opinion and experiences:
I agree with this article that there should continue to be a place for standardized test scores in college admissions, if only because I have myself as an example (I had mediocre high school grades because of poor executive function, but got a great SAT score, which got me into the college program I wanted. I graduated and now work in my field.). I don't agree with all the author's conclusions (he makes a throwaway claim that covid school closures were a mistake, linking to another article... But that's another comment all together), but I think he does have a good point to be wary of dogma without evidence
Standardized tests, of course, will always have the problem of coaching and teaching to the test. However, free and high quality self-guided SAT prep is now widely available online. I've frequently seen claims that it's difficult for colleges to attract high quality, diverse students. I wonder if a program could be set up to identify disadvantaged students who are scoring especially well on that prep, to connect them to free personal tutoring and college admissions guidance?
Standardized tests in college admissions are also a bit of a self-fulfilling prophecy:
High SAT scores (especially pre-2016 when there was more game strategy) were mostly an indicator of how well you take tests. People whom are good at taking tests will also earn high GPAs in college, especially when tests are a huge portion of the grades.
That does not mean college was not immensely valuable to that giant swath of college graduates that were not chasing the 3.6+ GPAs. Or even to the people who didn't end up graduating.
I'm wondering if, now that we have computers, we could develop some sort of system to prevent teaching to the test. I currently have no idea what that may look like, but I'd wager at a minimum it involves an exponentially larger question pool and randomized weighting of scoring.
I think that's too much of a simplification. I disagree that it's "mostly" an indicator. Rather, it is a significant but small component.
It's not really possible to cram 11? years of schooling a few weeks or months before a test. Prep courses are very popular with affluent students, and they're all about teaching students test-taking skills. Various studies show that test prep improves SAT scores by roughly 15 to 30 points, according to Slate and NYT's Jay Kang. In admission for very competitive universities (Ivy League and Tier 1) where nearly everyone has a top decile score, it's an important edge; outside it, it's really not much. So the impact of one's test-taking skills may move one up or down by one decile or percentile, but they can't turn a 1100 into a 1550.
That, IMO, is one of the deepest merits of the SAT: it can't be studied for. Test prep can only help pick up marginal points, not cram content.
The downside is that it penalizes kids who grow up with parents who don't keep books in the house.
I wonder what will happen to the vocabulary section. When I took the test I benefited greatly from having read books at my grandparents house such as Ivanhoe and Last of the Mohicans. Modern published books don't include the range of vocabulary used in classics. Kids who are encouraged to read still aren't stretching with complex words and sentences the way we did for the most part
I think the penalty is necessarily 'fair' because the purpose of the test is to filter and assess college readiness. The SAT is pretty basic; it rarely has anything esoteric. If someone struggles with the test, it's likely that they'll struggle in college where the expectations and pressure are much greater.
Ultimately, there are other perfectly viable and respectable life paths after high school that don't involve college.
Two thoughts I can easily express on a phone keyboard come to mind:
Tangential, but I’ve been told that my fanfic reads like someone ran a thesaurus through a mad lib because of my enjoyment of using new words I learn from Mark Danielewski novels (and also appending affixes to make up words)
If you’re teaching good material, but no longer teaching to the test, then the test needs to be modified because it’s a bad test. The whole point of tests is to accurately assess how well someone has absorbed or retained knowledge.
The issue of testing the test being a poor curriculum isn’t an indictment against testing, it’s an indictment of a bad curriculum, including the test for that curriculum.
You're right that that's not an argument against tests. Let me make the argument against tests.
But media are not neutral!—the medium is the message—that something is a test says something about it and it's incumbent upon us to consider if that is good or valuable or useful. Here are my observations about tests:
For teachers, tests provide a valuable quantitative metric of, in particular, which students are being 'left behind'. It's valuable because teachers have heavy workloads and are asked to teach large classes, and therefore lack the ability to have a close personal relationship with each student and a sense of where they are at and what they need. Tests may improve matters slightly, but the underlying problems are still there, and the addition of tests may serve to paper them over. (Chris Crawford discusses this a bit, in 1993—I don't necessarily agree with anything he says, but it is interesting.)
For students, tests create an adversarial relationship between themselves and the teachers (and the institutions they represent). It's not just the notion of external assessment—which is obviously valuable and need not be infantilising (see for example peer review—ok, obviously there are problems with that, but it is a lot more constructive)—it's the details, the medium, the structure around it. The distrust. I had a lovely biology teacher in high school who would offer to do oral make-up exams. Out of a somewhat fundamentalist desire for fairness, he would pretty much just note down what you said and grade it against the same rubric he used for written exams. But the shift in medium nevertheless caused a dramatic shift in tone and relationship, and people who struggled with or stressed over written exams often did much better in the oral makeups.
For students, tests, as other graded assignments, act as extrinsic motivation and pressure, destroying or at least dampening intrinsic motivation and interest. Unlike some other sorts of assignments, however, there's little that's creative or constructive about tests. At best, they can be tolerable, or funny. But they can't really have genuinely redeeming qualities.
Which is another problem: they have extremely limited scope. All they can test is what can be seen on a few sheets of paper given a few hours' turnaround time. If all you want to know is whether somebody really has no idea of what's going on, great, maybe; but, per above, that's not that interesting. I can poop out a half-baked essay comparing Aristotle to Sartre in a half-hour (it got full marks), but every really good idea I've had has taken a bit longer than that to work through.
Having such limited scope also means that tests' results are prone to be very volatile, because you're sampling only a small slice of a person. How much sleep did they get? Did they eat breakfast? Did their cat die last week?
I am not suggesting a solution here (though I'm also not proposing there isn't one). I am just saying that we must take the problem seriously.
I'm not actually convinced of this. I am someone who naturally does super well at these types of standardized tests, and I'm also someone who had a high GPA overall in college. But none of the tests I took in college, even among the courses that had conventional tests at all, were particularly similar to standardized tests. Most were fundamentally different in their structure and requirements, and that's before we take into account how few of my courses even had tests like that at all. Final projects or papers were more common in my courses.
I'm not necessarily advocating for getting rid of SATs, but I have not found the skills that got me my high scores particularly applicable outside of standardized tests themselves. Some of the skills that benefitted my standardized test performance, such as willingness to quickly skip questions that seemed unintuitive to me, have probably negatively affected my performance in college and beyond.
Maybe part of the test is the ability or willingness to adapt to a different structure with unique requirements. Then the requirements would have to be unrelated to most real-life situations so no student would have an advantage due to experience.
I'm not some huge fan of the SAT, my experience is similar to yours, it's just that your comment made me start wondering if the lack of real-life applicability is intentional.
I mean... I don't think so, because standardized tests like the SAT all have set structures (hence the standardized lol) and those structures don't really differ from each other. Performance improves the more familiar you are with that particular test, so these tests optimize for familiarity with standardized testing, not ability to adapt to a different structure. Kids my age have been exposed to standardized testing with a similar format from a young age. My parents had me take the full SAT a year early to prepare for the PSAT (and I was my school's only National Merit finalist, which is based on PSAT results, so I can't say it didn't work!)
Even outside of the at-least-yearly standardized testing, taking tests in school was a huge part of my real-life experience prior to college. The traits that made me good at standardized tests also made me good at most tests in middle and high school. I was a good test-taker. The problem is that once I actually moved into courses past very basic early GenEds in college, these skills were basically no longer applicable at all and were often detrimental. For your theory to hold water, you'd want the opposite -- for the tests to have little association with real-life prior to the tests but an association with real-life afterwards.
I mean personally I think college as it is really has the same problem.
You could make college 4 years of learning to spit tobacco into different shaped buckets and those people would still go on to be more successful than their peers because of just how badly a college degree gates you in the job market.
Yes and no. At this point, a BA/BS may not even get you in the door anymore. It's becoming the new high school diploma. Since so many people are getting them, they're becoming devalued (which IMHO is dumb).
I've seen so many jobs posted recently (I've been helping spouse and several friends job hunt) that are barely above entry level requiring a master's degree and X years of experience. Will those postings get them? likely not, but the fact that they're asking for a master's now instead of a bachelor's is insane to earn $23/hr in the Bay Area, CA.
I'm seeing that shift in higher education, one of the biggest culprits of the masters required entry level position.
The "Great Resignation" hit us pretty strongly and we both raised salaries and lowered our required qualifications. To a good outcome IMO.
There are so many jobs that require higher education that don't need it. Many jobs that I've done personally that require not only a bachelor's degree, but a high GPA. There is no reason for a receptionist to need more than a high school education, not with the tools we have now.
Requirement inflation is almost purely a ploy to cut the applicant pile into something manageable. This is the negative side of online job sites allowing people to spam every opportunity across the nation.
This op-ed falls into a common trap, failing to question the premises of its argument. The SAT is designed to predict how well a student will score in their college courses. When College Board evaluates a question for inclusion on the test, their main criterion is whether it accurately predicts what a student's grades will be. The article spends the whole time defending the SAT on the grounds that it does accurately predict what a student's grades will be, but it does not question why you would want to predict that.
Here are a few other reliable predictors of college student grades:
I could go on. I think it's obvious why it would be wrong to use any of those as criteria for admissions. Part of the goal of a university is to provide opportunity to the disadvantaged. It would be wrong to adopt an admissions practice which reinforces existing social inequalities, even if it does result in a student body which gets better grades.
College Board goes to great lengths to control for effects of race, gender, and class in their tests. They are entirely aware of this criticism. Personally, I think they are fighting a losing battle. I don't think it is possible to create a test of "scholastic aptitude" which does not simply reinforce existing social inequality. In psychometry jargon, I don't think the construct is well-defined. Instead of trying to sort students into the categories of will-do-well and will-not-do-well, I believe that we should direct our energy to making colleges work for all students.
Everything in the college admissions process will reinforce existing inequities though. Extracurriculars do that, GPA does that, projects/volunteering do that. The SAT does too, but it’s relatively easier for poorer students to play it in their favor.
The reality is that college inequities are only a symptom of broader issues in the educational system, and fixing that is required to fix college issues. Unfortunately, we’ve been trying to fix it for decades, spending inordinate amounts of money in the process and in some cases making it worse.
Do you think so? It's not obvious to me why that would be the case.
The SAT/ACT can be studied for, and it doesn't have to be expensive. SAT prep books you can usually get for $30, there's no lack of online resources, and many schools, community colleges, etc. offer free SAT prep classes before or after school.
I went to a high school that is usually in the top 5 in our state and our median household income was 137k so I know a lot of the tricks.
What are the other factors? GPA. Many magnet schools or just schools in expensive areas would have connections with college recruiters - they "knew" what our "gpa" meant, and would weight it more heavily.
Our GPA was out of 100, but AP classes counted as 110, and the school covered the exam fee, so we would all spam the shit out of AP classes. By sophomore year, at least half of your schedule would be AP classes, and it would be 100% AP classes by junior year. I ended up with a 107 GPA - that's what was being sent to universities.
Classes could be rigorous, but there would always be an opportunity to "make" up your grade at the end, no matter the trials and tribulations you went throughout the class.
I can guarantee that our GPAs were considerably more inflated (I mean, they were above 100 in scale that out of 100 lol) a school with less resources.
Extracurriculars - it basically was a running joke that everyone had their own NGO. Parents would be in groupchats where they talked about the best kind of NGO to make for admissions, or how to actually have the NGO "do something" while actually doing the minimum amount of work. Of course, everyone played at least one instrument.
Our robotics teams was an incorporated LLC that not only received significant funding from alumni but also had corporate sponsorships.
We have a club that made those low ride solar car things. I'm pretty sure each one was at least $10k in materials.
Compared to all that, prepping for SAT is a much closer gap to close financially.
Not to mention, prepping for the SAT has a much lower impact on overall score than other test preparations. It's not easily gamed by cramming, unlike other tests.
Yes. I think @stu2b50 covered a lot of points I wanted to make, but I’ll still outline my thoughts below:
Let’s take a hypothetical poor but bright student in a bad neighborhood of a generic Rust Belt city. They live with their mom who works as a CSR, in a two bedroom rental apartment. The student works after school and on some Saturdays at McBurger. They are the top performer at one of the worst high schools in the city, and their guidance counselor has said they should go to college. What options do they have for impressing admissions officers?
I don't know the data but test prep is one of the least resource intensive options. It can be costly but it doesn't have to be. And even poorer schools will have study guides available. It wouldn't surprise me if the differential is less.
I would go a step further and argue that it is not possible to create a university admissions process which does not simply reinforce existing social inequality. At the most fundamental level, university admissions is about deciding who is not going to be invited to attend a university. With the current US education system, even an admissions lottery would be reinforcing societal inequalities because of the cost of attending university.
Education is an investment into society. Highly educated people are good for making societal progress, improving the economy, and are well-suited to be the leaders of a society. Inequality in university admissions would be vastly decreased by improvements in the k12 public education system and a stronger public education system for free adult education (like community college). This is where I think we should be focusing our energy.
rant on scholastic aptitude
I think assessing "scholastic aptitude" is important before attending college, but we should ask ourselves what "scholastic aptitude" actually means. If you asked someone today, they might say it means "being good at taking tests" because that is how you succeed in university today–I don't think it's fair to students to admit them to a test-heavy university program if they're bad at taking tests, especially in the current system where students have to pay per class (and can even be subject to fees of ~5x standard tuition for taking too many classes in Texas).
Put another way: the primary filter for university admissions should reflect the skills necessary to succeed in that university's program.
We already see this with music schools: an audition is typically a part of your admissions package. Art schools will typically ask for a portfolio. It gets trickier when you start to look at programs like computer science, because the expectation is that you will be taught all of the necessary CS skills in the university program.
Going back to "scholastic aptitude," I think it's worth revisiting course exams as a whole. In many cases, if you fail all the tests in a class, you'll fail the class–even if you already have mastery of the course content and are just a bad test taker.
Projects are an excellent way to demonstrate mastery of a concept, but have challenges associated with them (bias in subjective grading, potential to pass off someone else's work as your own) and some programs are better suited to projects than others (computer science is a good program for projects, liberal arts perhaps less so).
Another thing to consider is what else universities can look at to determine admissions. A few I'm aware of colleges looking at are:
While I don't think standardized tests are particularly good measures of who should be admitted to college, I think they're one of the least bad options we have. There's plenty of free online practice (KhanAcademy even has one that adapts to how you've performed on the PSAT and their own practice exams) and plenty of low-cost SAT prep books that can also be bought used.
Bias and inequality are just so deeply entrenched in our society that essentially any system implemented at a societal scale is going to reinforce these problems. The best solution to inequality in college admissions is to address the inequality in society.
This has been long and windy and probably doesn't make any sense, but that's just my brain dump of thoughts on the SAT and university.
I have a massive headache, and I want to reread your comment tomorrow and give a real response, but I just really needed to address this part tonight
Anyone who goes into a CS program assuming that all skills they need for a job will be taught is in deep shit. Most classes glance over early concepts, and later courses often assume skills (like, say, navigating a Linux command line) that are only ever barely touched on in class.
Many CS programs are also more theoretical, and focus very little on the sort of programming you do in the workplace -- my first internship expected me to work with REST apis, and parsing HTML, which never even got mentioned in 4 years outside of an elective web programming course that I withdrew from.
Perhaps something like nursing, or accounting might be a better example?
But yeah, sorry about that, the "someone is wrong on the internet" hurt a little
Oh believe me I agree but that was a very common opinion in my program.
Playing off of what I mentioned elsewhere, everyone should have the opportunity to learn the fundamentals in a free public education setting (both in high school and community college). But that’s just not an option for all students unless they self-teach.
I wound up self-teaching a lot of the practical concepts before and during high school and was fortunate to attend a high school with some absolutely astounding computer science teachers- they were better than nearly every CS professor I had.
I'll toss out a partial counter to that: Universities tend to also have massive amounts of financial aid available. On top of government grants and aid.
A lottery process would eliminate a lot of those systemic problems. Mandating higher percentages of aid going to lower-income students and banning higher-income students from recieving performance-based aid would free up a ton.
You're not wrong, but once a lottery system is in place smoothing out the budget/loan gap is a bit easier.
How are you defining "higher income" here? No one who otherwise could have gotten a full academic scholarship should have to take out loans merely because their parents made too much money.
Wait, there is something very off about this statement. Asian Americans have the highest mean college GPAs (the source tracks from 1990 to 2009, but we can assume the trends track similarly with recent high school GPA data), and they're definitely not perceived as white.
Latino Americans are significantly culturally and genetically white-adjacent yet have the second lowest mean GPA. I'm getting a little in the weeds here, but a 2015 paper published in the American Journal of Human Genetics estimates that Latinos in the US on average carry 65.1% European ancestry. Many Latinos are able to pass with some degree of whiteness — it doesn't seem to help them much.
If there were structural racism in the college environment, then its effects must be very weak and easily surmountable.
Latino Americans are not generally perceived as white.
https://www.pewresearch.org/hispanic/2021/11/04/measuring-the-racial-identity-of-latinos/
I don’t know what genetics have to do with structural racism, seeing as race is not really a biologically meaningful concept.
That's true, but they are perceived as whiter than Asians. Some are perceived as somewhat white; and some present as unambiguously white.
Genetics determine appearance. Whether or not race has any biological meaning, people racially classify others by their appearance.
An average 65.1% European ancestry across the Latino American population means that a significant number of Latino American individuals present as partially or very European/white.
At the very end of the "very" spectrum is my best friend from college. He's Latino American: Colombian. His heritage is mixed indigenous and European. He looks and passes as totally white. He identifies as a Colombian American. I doubt that anyone has visually recognized and treated him as Latino in a discriminating manner.
In college, I met a number of Latino Americans like him who I wouldn't have clocked as such until they told me about their heritage and how they identified as such.
Anyway, my original thought is that the statement that students get better grades when they're perceived as white stands at odds with the reality. If it were true, we'd see white students with the highest average GPA; Latinos, who are not generally perceived as white but some of whom are perceived as white and others are perceived at least whiter than Asians, with the second highest; and Asians in third.
I think that’s more complicated than simply being perceived as “whiter” and I don’t know if I’d feel comfortable characterizing it in those terms at all.
This is an unwarranted extrapolation from what @alden actually said. They brought up 7 separate factors, but you have created an argument around how one single factor is not totally determinative when we shouldn’t expect it to be in the first place, particularly when four of the other points are clearly correlated with the results: