Last month 2,400 University of California faculty asked for admissions to resume using the SAT "to ensure foundational fluency." Of course many employers want to ensure that too, especially when college degrees don't anymore.
The widening abilities gap followed the 2020 elimination of the SAT/ACT, a temporary measure that has now become a permanent vulnerability. This outcome was explicitly predicted by the Academic Senate’s 2020 Standardized Testing Task Force (STTF) report, which warned that removing these tests would eliminate a vital predictor of college success and obscure the impact of severe high-school grade inflation. Unfortunately, the outcomes cautioned against in that report have now materialized in the data across our campuses. All other leading STEM institutions, including the UC’s primary peers, have resumed using SAT/ACT in their admissions to ensure foundational fluency. For the University of California to remain a global leader in STEM, it is essential to restore these objective benchmarks. -- https://ucstudentsuccess.org/
The UC faculty opposed the SAT requirement being discarded in the first place. They were overruled by the UC Regents, and that may happen again. And even if the SAT is brought back, I'm sure it will be given much less weight and subjected to the "in a local context" process in the name of equity.
Why is a lossy testing filter better than just failing out those who can't make it? Maybe allow for larger freshmen classes and smaller latter classes or adopt community colleges and have all students start there and advance into the UC system sophomore year on. Instead they bring back what is basically an IQ test for admission.
SAT score is known to be predictive of college grades. Is it also predictive for whether a mid-to-late career candidate will pass a phone screen? It is used for early career candidate filtering in finance, but I have not heard of anybody caring beyond that because of the availability of signal on the actual tasks they will be performing.
how many elected leaders are in STEM? would high SAT scores and grades exclude many US presidents and congressmen? (yes) winning elections seems kind of important to me. so if you were just like, selecting for leadership - and many of our leaders are brilliant people, just not in the sense of being good at taking tests - would that be good or bad? or... what is your real opinion? what are you actually mad about?
obviously the UC system should give spots to the kids who will use those spots the best. but it is very hard to define what "using spots the best" means.
I have seen company descriptions in job ads that list college achievements of founders. They are invariably young Asian men. I understand that it's a cultural signifier and don't judge them. But, I also understand that I will never hear back from them because I don't share that background. So, I never apply to any job listing that references college experience of either side, other than wanting a degree in general.
Someone once told me that American work culture used to be based more on intern-ship / apprentice type hiring but now obsess with formal degrees. I wonder how much of this shift in culture is influenced by the Korean, Chinese and Indian immigrants, as a formal education is a prerequisite to compete in these countries' job market? For example, it is quite common in India for employers to ask for our 10th and 12th standard marks / grade (because these are national exams) along with college grades - to apparently gauge "Consistency". Fluctuating performance, a break or dropout years all negatively impact you and can be nerve wracking for many freshers, until they manage to get some work experience. It is somewhat disappointing to see this culture permeate to America too, even though I feel quite conflicted about it - after all, everyone does want to hire the best / most competent / reliable candidate; but the other approach - a vocational kind of training - also has its merits and seems to have served American companies well too. (Zoho in India is experimenting with this kind of hiring in India where they are hiring high-school students, mainly from rural areas, and offering them a work cum study program. They don't get any formal diploma or degree though - https://www.zohoschools.com/ ).
> I wonder how much of this shift in culture is influenced by the Korean, Chinese and Indian immigrants
I think there’s an influence, but it’s amplifying a pre-existing trend. Bureaucratic societies favor formal credentials. The U.S. has become much more bureaucratic since the mid-20th century, and credentialism has grown. Reliance on degrees and other formal credentials also enables the universities to achieve political goals through admissions and grading policies. Asian immigrants in the U.S. have readily adapted to that system.
I think it is a positive for an employer to ask for an SAT because it tells me right away I don't want to work for them. Once (a long time ago) I tried to upload my resume to apply for a job. The web page started asking me very basic questions, like a basic aptitude test. I was out. Tell me you do not know how to find and evaluate talent!
I think this is highly age-dependent. I took the SAT well over a decade ago and have significant work experience since then. It would be odd to require me to put down my SAT scores, which I don't even precisely remember.
But if I were < 5 years out of college, and especially if I had gone to school during COVID times (when SATs were not required by many colleges), I would completely understand why an employer might ask.
Basically, colleges used to act as a filter for SAT and other attributes. During the 2020-2025 period, they admitted students under fairly different standards, due in part to testing challenges and social movements.
It makes sense for an employer to want to do a little more diligence to ensure that students who were admitted during this period are similar to students admitted during the prior several decades.
It was required at my first job in 1989, for entry level actuaries.
In fairness, part of job performance was passing the actuarial exams, the first two of which were calculus and statistics. I imagine testing well on the SATs for a math or EE degree (what they hired) was a good indicator of passing tests.
As a manager, there are several qualities that I value highly in an engineer, and they all happen to begin with the letter C: Competent, Consistent, Curious, Caring, and Clear Communicators.
While SAT scores might act as a proxy for competency and possibly curiosity, they're not going to tell you much about whether the person is consistently reliable, whether they care about others and cooperate well, or whether their vocabulary or literary analysis skills have any correlation with their ability to read the room and tailor their communication to their audience.
If I were giving these job posters the benefit of the doubt, I would guess they're including this requirement for the same reason that musicians request particular colors of M&Ms in their riders. They want to weed out people (or bots) who aren't paying attention. Nevertheless, there are better ways to do that than demanding (and presumably filtering by) teenage performance metrics.
I met an HR manager who had worked for a local but well known company with a reputation for caring about things like GPA and SAT scores. She told me that remembering your SAT scores after college was a sign of a competitive attitude.
I had a friend with an MS show up for first-day of work for a job that asked for SAT scores on the application. HR said "we never got documentation for your SAT scores, can you provide that?" He was on the phone with his mother, having her go through a filing cabinet when he realized that he didn't want to work for a company that was this serious about SAT scores when hiring someone with a post-graduate degree.
That does seem wild. Out of curiosity, did he also have to submit GRE scores, which would be closer in time and more representative of his current knowledge/skills?
Worker w/PhD was asked by Quebec during residency interview to produce High School grades. Bureaucrats will be bureaucrats independent of language I guess.
The job description touts investment from some members of the "PayPal Mafia." For some odd reason that fact that and the SAT requirement combine to make this whole thing feel kind of normal.
> Please note that we will also rely significantly on both solicited references (where you introduce us) as well as unsolicited or "back-end" references (where we do our own). For the latter, please rest assured that we will never contact a current employer without first getting your permission.
References already give me goosebumps. Having them reach out to people who haven't given you permission to be a reference sounds like a recipe for disaster.
If they do sensitive work for the government, it'd make sense that they'd do those back-end references. Also if they are in high end finance where you want to weed out people who have demonstrated moral flexibility was well as total lack of it for certain things.
> If they do sensitive work for the government, it'd make sense that they'd do those back-end references.
If they do government work that requires clearances, the clearance process already covers this sort of investigation on its own.
In any case, they are free to do whatever background checks they want within legal limits, but I'd never apply to a company with such ridiculous hiring processes.
You cannot use the SAT as a metric to compare different cohorts. SAT scoring has been revised many times over the years. When I took it the highest possible score was 1600. From 2004 through 2016 the highest score was 2400. Now it is back to 1600 again. Plus, both the content and the format of the exam has changed many times over the years. At times, there was no essay requirement, at times the essay was required, and at times it was optional. Hence, each year the examination produces a different distribution/histogram of scores even if you normalize the 1600 vs 2400 difference.
Even the scaled score is not that informative (and perhaps crosses the line on age discrimination) because for older workers the population of people taking the SAT was much smaller as a percentage of high school grads (and presumably weighted towards higher IQs). It's also why there were so many fewer perfect SAT scores -- smaller population in the bell curve.
Which requires them to explicitly ask your age outside the bounds of qualification for a job (over 18 etc). Which ends up opening them to age discrimination lawsuits.
It does not require them to ask about your age, just the year in which you took the SAT. As other commenters have pointed out, this can range from 12 to 17.
Also, they could just ask for your SAT score and any relevant info (if you took it during COVID from your car, etc.) and then you could disclose whatever context you wanted.
It's weird that people think AI breaks the concept of work sample testing. Work sample testing isn't "about" programming, and predates the profession of programming. You can (and some companies do) work-sample test sales account managers, customer support, accountants, whatever.
AI changes the underlying job you're testing for. So, obviously, the tests you might have been using pre-AI won't work anymore; they're testing something that isn't really the job anymore. Update your tests so they're about the real work again, that's all. For coding, that probably means assuming (or requiring) candidates use AI to do your assessment.
What AI really does mess with is conversational/interactive interviewing. We do all our interactive scripted interview on Slack, but I can imagine us having to end that practice and return to face-to-face.
I would never ask for them since its so cringe. But SAT scores correlate to IQ at .81, and IQ is one of the few things that strongly correlates to knowledge work performance positively. There is probably a lot of alpha from knowing candidates SAT scores. Its more useful than knowing the college they went to.
> and IQ is one of the few things that strongly correlates to knowledge work performance positively
i think you mean that it correlates to pay. nobody knows what you mean by "knowledge work performance." reviews of your peers also correlated with pay. often it is not the smartest person who is the most popular. so... do you see how you said something kind of meaningless?
> nobody knows what you mean by "knowledge work performance"
I actually was pretty easily able to deduce what they meant by "knowledge work performance".
It's understandable to be frustrated by not knowing something, but to claim "I don't understand that and therefore no one does and you're being nonsensical" is a bad look.
Consider responding with curiosity rather than defensiveness.
A very good metastudy is "The Validity and Utility of Selection Methods in Personnel Psychology" (by Frank L. Schmidt and John E. Hunter). It summarizes 100 years of research on predicting job and training performance. It makes a very strong case that General Mental Ability (GMA, their word for IQ) is the single most valid predictor of employee success on the job, not just income.
Between 2005 and 2015 the maximum score was 2400 instead of 1600. Assuming anyone who got <1600 during that period wouldn't admit it, you now have three well defined buckets.
But of course this is a lot of unnecessary steps compared to the usual method: length of work and education history +18 years.
Hah, you did the one at 12 also? I too remember both, and it was also 30 years ago. I don't know why, probably because they were really important numbers to teenagers, and they say you remember things that happened to you as a teen more than any other part of your life.
Sneaky age filter? You must be young enough to remember your SAT scores.
I can remember mine just fine.
If you're really looking for smart people, use "Answer this word problem in two or more paragraphs. Write your answer on the sheet of paper provided. In cursive."
Are people who were forced to learn cursive smarter? They're just older.
Doing this may well expose you to age discrimination lawsuits, since it's just sneaky indirect age filtering.
Another example would be if you required a minimum SAT score of 1601. Sure, someone could have gone off and taken the SAT as an adult or a young child but in reality it is mainly an age filter.
The signal being that the smart people will refuse to jump through this hoop in your inane process because they have a lot of other opportunities to choose from.
I learned in Brazil. Here in the California I asked my son's kindergarten teacher if she would teach cursive, and she said they don't teach calligraphy and I've never seen it described this way, but she's right.
They are obviously using the SAT as a safer alternative to the legally dubious practice of IQ testing which can lead to running afoul of the ADA and EEOC. I'm not sure it's much safer, but I am positive it's less safe than doing timed leetcode. At least leetcode problems can be painted as relating to the job.
Additionally, the SAT is a shitty IQ test that is constantly crammed for and cheated on. I remember my SAT test. I was the only person in the room not openly cheating. The teacher proctor didn't care. Higher scores mean better students, more funds, higher home prices, bonuses, and a litany of secondary effects. That's not even including people that pay professional test-takers to do it for them.
The software industry needs to let go of their obsession with finding 10X ROCKSTAR L33T programmers. They never will though. It has gotten worse every few years for decades, and the problems are almost entirely managerial.
IQ testing isn't legally dubious. The idea that it is is an Internet myth. There are a couple household-name corporations that administer general cognitive tests for candidates for some roles.
More companies don't do it because it doesn't work well.
This by default includes a whole bunch of people who didn't take any kind of standardized tests (most notably, immigrants).
The (albeit small) country I'm from doesn't do any. Reasoning was that standardized tests create an environment where teaching is merely done to create good test scores, not to actually teach.
They made me do an online, timed IQ test, stressing that I should take it in my native language to not waste precious seconds understanding it in english.
It was horribly translated, every sentence was written like something this: "A and B are two broters/sisters. A gives B 3 apples and he/she/them eats one and returns one to he/she/them…" at some point one section had the instructions wrong so I did all the questions wrong. There was no way to change the language or re-read the instructions to try to understand what the original text might have actually been.
They're known for asking odd questions like "how did you perform in math in high school" and "please justify your performance in math in high school". It was actually my guess as well before I read the post.
I just applied to Epic, the EHR company from Wisconsin, and I can confirm that they also ask for SAT scores. Thankfully I have my collegeboard credentials saved
Their pre-screen test was awful. Brain teasers, including some infamous ones like “if you have two coins that make 15 cents and one is not a nickel, how is this possible,” and moderately involved programming questions like “parse phone numbers from a file and record those with any of these area codes OR every other digit is a 2 OR they have multiple pairs of consecutive digits” and you’re given a blank text box with no formatting, IDE, or even non-word processing style indentation help.
Thankfully, the coding assessment does have syntax highlighting for plenty of languages, but in general I feel like face-to-face assessments are more productive. TFA seems to make a similar point
Out of curiosity, did they require SATs, or just ask? And how many years out of HS are you? Seems like it would be crazy to ask for them once someone has finished college and has a UGPA to report.
I like it. I also like it when companies ask for 10 years of [5 year old technology] experience, or say "there's more to working here than the salary!", or other red flags that make it easy to move to the next listing.
If you think my decades old SAT score is relevant, then I know all I need to know about your company.
I only had one company (D. E. Shaw & Co.) ask for my SAT scores. I was late-20s and had to have the recruiter repeat herself two more times before I understood what she was asking.
It was also the single highest density of talent I’ve ever worked, by a long shot. Crazy talented coworkers.
I would think it an inappropriate question if you're asking a PhD (or MS). You could just ask for a copy of the paper(s) or the dissertation/thesis. Some people improve over the time in the uni environment. Some people don't test well when they're bored, etc. Some just grow up.
A lot of responses pointing out various flaws with this question, including the fact that it can be used as a proxy for ageism, the fact that the grading scale has not been consistent over time, or that most foreigners will not have gone through the US education system. However, is it really that uncommon for Americans to never have had reason to take the SAT/ACT, such as, simply not going to uni, or going straight to work after graduating high school?
You only need to SAT/ACT for going to get into an university. And these days it seems to matter less. I looked it up and less than 50% of high school students take the SAT/ACT. Personally I never had any employer ask for my SAT/ACT scores since the 90s.
> Take-home projects or a trial period of some kind. This makes the most intuitive sense by far: having candidates do a representative slice of the job gives you a solid idea of whether they'd be any good at it. Combining this with structured interviews was (before AI) considered a gold standard; you'd get a sense of who they are and how they work by talking, have a way to compare them pretty objectively to other candidates because of the structured and consistent nature of the interview process, and then you'd get a sense of how they apply their attributes practically to the job via the work exercise.
Unfortunately a lot of companies have over the last several years been using this to get candidates to do a project for free for them. If it's going to take more than a few hours of my time, I don't take project style interviews seriously unless compensation is added (which some companies do offer and is a big green flag).
Definitely been tricked into working for free a time or two.
Best middle ground I ever had was an interview where they impose a strict 1 hour (maybe it was 90 minutes, idk) time limit between when I got the prompt and when I emailed them back. Then they spent some time looking over it, then I had an interview with an engineer who had read my code and we chatted about it. Why I had made certain decisions, what corners were cut because of the time limit, etc.
Felt very fair. Not enough time to assign a valuable task, enough time and privacy that I wasn't under the gun like you are in a whiteboard interview, and it was pretty applicable to what I would be doing at the company. Solid interview. Didn't get the job, but respected the process.
I took both (since I could just use whichever I scored better on for applications), and the ACT is a much less miserable format. The test is broken up into much larger time chunks, meaning I could take a more useful nap after finishing a 50 minute section in 20 minutes than I could finishing a 20 minute section in 8.
Difficulty wise, the ACT was easier, but not by a lot. They seemed to have pretty similar predictive strengths, based on percentiles of people I knew lining up between the two, but obviously the ACT loses precision because of how coarsely it's scored.
> Why would you ask for a self-reported, unverifiable test score that could be decades old at this point?
Because many colleges that used to reliably filter for them no longer do (or didn't during a several-year period).
It's true that self-reported scores are not the most accurate, but if I were applying for a job I would report honestly, on the assumption that they could easily request for the scores to be sent by the College Board. The risk/reward of lying does not make sense, at least in my case.
> > Why would you ask for a self-reported, unverifiable test score that could be decades old at this point?
> It's true that self-reported scores are not the most accurate, but if I were applying for a job I would report honestly, on the assumption that they could easily request for the scores to be sent by the College Board.
No, they couldn't, except by going through you (the College Board doesn't take third-party score requests.) You might be able to request that if they are recent enough, but not if they are literally decades old (well, not if they are ~21 years old or older.)
I am aware that third parties can't request scores. I was referring to the employer asking to have the scores sent, which the applicant would be compelled to do (or look like they fudged their original score reporting).
I'm also aware that the College Board doesn't hang onto scores forever. I doubt there are any employers who require SAT scores for applicants who took it prior to 200% (the cutoff indicated in your linked article).
My problem is it's self reported, so it ends up being a "are they smart or did they lie?" game. You can't easily verify it for anybody over 23 or so.
Canonical has a job application where you are supposed to rank yourself on a percentile (up to like 1 in 10,000) on how good you were at math in high school. It's a very easy way to incentivize lying, and also to hire people with an excessively high appraisal of themselves. There are a lot of people who are reasonably good at math, and have avoided humbling environments like the Putnam, and have convinced themselves they are God's gift to math, when in reality they were just the brightest kid in a class of 100 high school students.
Another thing making a comeback -- reference checks. I had to supply references for my first job in 1999. Then I wasn't asked for them again until 2024, and then for every job after that.
If you want the job, can’t you just lie? Or are SAT scores something that cannot be faked? I dunno, I also say I know Kotlin when I have more experience in Java (and honestly I couldn’t care less about specific tech stacks), or that I know about tcp/udp when all I have is read a couple of (good) books about it.
I don’t feel bad lying about some stupid requirement
> If you want the job, can’t you just lie? Or are SAT scores something that cannot be faked?
I would assume that if you progress to the point of an offer, they would ask you to have the official scores sent by the College Board. Apparently they hang onto scores back to 2005 and can send them for a fee.
Many years ago I was walking around a college job fair when a recruiter was yelling at a college student. His crime: "wasting everybody's time" by presenting a Resume with a 3.4 GPA when the company clearly listed a 3.5 GPA as the minimum they would accept.
If this is how they treat people that don't yet work for them, it doesn't bode well for how they will treat people that work for them.
As I mentioned elsewhere on HN [0], younger generations are much more competitive now.
Visit and talk with undergrads at a top CS program like Stanford, Cal, UIUC, MIT, etc. The culture is different because this is a much more competitive generation. When the acceptance rate into a top CS program is in the 1-5% range and laurels like being a Valedictorian, NHS member, JV or Varsity sports team member in HS, getting a 2100/1500+ on the SAT, and taking 6-7 APs are now table stakes, you get a degree of viciousness, competitiveness, and steel-eyed execution that a lot of older Americans just aren't used to.
This mindset is the norm across Asia though - from the Gaokao to the JEE to SKY-or-bust. Honestly, I'm glad that younger generations are much more competitive now - pressure makes diamonds.
And honestly, the top 40-50 STEM programs nationally graduate around 30-40k new grads a year. Add to that respected regional programs and Veteran-to-Employment pipelines and you have a self-sustaining talent pipeline.
I disagree with the increase in competitiveness being a good thing. Excessive filtering at all levels has meant that eccentrics or absent minded professor types are not making it into research roles, and creatives or mad geniuses are filtered out before they have the chance to make an impact. There are a lot of people who are extremely bright and creative, but just don't have it all together the whole time from ages 14-25, and these days they have no chance of making it into research positions.
The system is rewarding conscientiousness and consistency over creativity.
> The system is rewarding conscientiousness and consistency over creativity
This assumes that you can get to the top via rote skills alone. Rote learning only gets you so far and most of those kinds flame out.
It's hard to describe, but once you meet actually talented people what you end up seeing is that they're just extremely diligent and deeply passionate about a topic and will continuously execute.
For example, when I was in HS I wrestled. Yes there were physical differences that could impact a sparring round, but technique and preparation was almost always able to outcompete base innate talent. Later, I ended up learning ballet the Russian style and it was the same - the truly creative types who were at Vaganova or Paris had already built strong fundamental and technical skills which allowed them to mix and match and create.
You cannot be creative without also being diligent and understanding fundamentals.
The "eccentrics" and "mad geniuses" are few and far between, and to find people with talent, you do need to use exclusionary tactics like scores and interview performance.
Not really. Outliers of that sort get dealt with as an exceptional case and it works. I've seen amazingly bright batchmates get into all sorts of programs without most of the qualifications, because they were a genuine math wizard (the kind that submitted errata to a standard textbook on some weird number theory stuff at age 15), but they didn't always score the best on many exams, especially when aggregated across subjects. IIRC he's a researcher now. All trends point to him being an eccentric old hag at age 50 ;)
The filtering system is meant for the majority case and there it works. The outliers get dealt with as outliers, which also works. In this case, he later asked the author of that textbook who he emailed with the errata, to connect him with the group he wanted to work in. Needless to say it was a very strong referral.
The US education has always been competitive. In sports.
Its school system has always been a state-sponsored daycare.
SAT/ACT tests reflect this. I can get a perfect score in SAT math easily. And I likely could do that as a kid (I never took standardized tests at school). I wouldn't have been able to get the perfect scores in the Chinese gaokao or Korean/Japanese tests.
> There are a lot of people who are extremely bright and creative, but just don't have it all together the whole time from ages 14-25, and these days they have no chance of making it into research positions.
This is just nonsense. Are you saying that we should kick out smart kids with high test scores to let in absent-minded students who care about only getting drunk so that they _might_ become great researchers in their 30-s?
To the topic at hand: it's way too easy to fluff your resume with nonsense like "Coordinated a responsible team for an implementation of cross-cutting concerns improving customer retention change by 12.23% across the organization". Test scores provide at least some objective measurement.
I'm not opposed to the use of test scores, it's more the requirement that kids have the whole package of grades+activities+honors societies+test scores consistently through a phase of life that is widely understood as tumultuous for many. We're selecting for robots.
There's a reason the west was so productive in terms of new scientific and technological discoveries in the twentieth century, and it's not that our scientists were the most consistent conscientious students who prepared extensively for exams and padded their resumes in just the right way.
And to add to this, learning itself is a _skill_. Working on a complex problem, looking at it from different angles, spending time memorizing facts, working on learning to paint fine lines - these are all skills that you need to master.
By not motivating children to do that during their formative years, you set them up for failure later in life.
Yes, there will always be exceptions, humans are extremely variable. But for the general case just letting children float along without any goals or competition is not a great general strategy.
I didn't really state my whole view, but hard exams like Oxford and Cambridge use make sense for elite colleges to use (rather than unrelated extracurriculars), but reforming the whole education system to be oriented around a single high stakes test like in China or Korea has its own severe costs. I do not want high schoolers to spend 20 hours a week at hagwons, and the current resume filler system is also terrible. I do not think declining admissions rates at elite universities reflect that the students are any smarter or more prepared than they were 20 years ago, but rather they are much more cutthroat about many things that are orthogonal to being successful adults.
Ideally we'd follow a more exam focused system more like the UK, though I wouldn't want to require all students to only study 3-4 subjects towards the end of high school. But something in the european model of IB/Abitur/A Levels, where there are serious exams in various subjects at the end of high school for all uni bound students, plus some special higher level exams for the most elite unis (in the vein of Cambridge's Sixth Term Examination Paper). We could probably repurpose AP exams to fill a similar admissions role to A Levels, and possibly use the AMC/AIME/USAMO more explicitly for admission.
Edit: this doesn't let me reply again, I think the chain got too deep. But the point is that we're not just using AP exams and USAMO scores, we're also using a pile of other metrics around extracurriculars, GPA, and honors societies, and the end result is stressed out children and not actually having better outcomes than we did two decades ago. Declining admissions rates at elite colleges do not reflect smarter and more productive incoming students.
Yeah, Chinese-style gaokao would be impossible in the US. And I think it goes _too_ far into the "competitive high pressure" direction.
There are many ways to make a more competitive and objective system. I honestly don't have a lot of professional experience with any particular one, so I don't have strong opinions on a particular form it should take. A European model is good, some kind of mix of Chinese+European would also be great. And ultimately, these systems would be more fair for applicants.
And the current topic just highlights the ridiculousness of the status quo. For most people in the US, SAT is the _only_ objective test score that they have.
> We could probably repurpose AP exams to fill a similar admissions role to A Levels
They are already being used like that in college admissions today.
> possibly use the AMC/AIME/USAMO more explicitly for admission
Already in use explicitly at most of the top CS programs today
> I do not think declining admissions rates at elite universities reflect that the students are any smarter or more prepared than they were 20 years ago, but rather they are much more cutthroat about many things that are orthogonal to being successful adults
The cutthroatness is what we should be optimizing for long term. Competition is what begets innovation.
Also, even at solid middle tier universities like UC Riverside, the calibre of student has increased dramatically over the past decade [0]
> There's a reason the west was so productive in terms of new scientific and technological discoveries in the twentieth century
> it's not that our scientists were the most consistent conscientious students who prepared extensively for exams and padded their resumes in just the right way.
Instead, a large portion were immigrants or the children of immigrants who arrived in the US as part of trans-national brain drain from countries with strict education systems (eg. Hungarian Jewish Americans in WW2, Eastern Europeans in the 1980s to present, Asian Americans today).
There's a reason Asian Americans, Eastern European Americans, and immigrant African Americans are overrepresented in leadership and white collar industries despite the very real handicap of having extended periods time without US citizenship or a greencard.
Instead of optimizing for feel-happy edge cases, we should be optimizing for building the best talent where possible, and that requires being competitive.
> We're selecting for robots.
Frankly, this is insulting as well. Yes there are some late bloomers, but they are outliers. If they can truly succeed they would stil find a non-beaten path to succeed in a competitive ecosystem.
> consistently through a phase of life that is widely understood as tumultuous for many
Only to y'all "heritage" Americans. For those of us who are kids of immigrants, we learnt that life is a race, either you compete or you fall to the wayside.
Of late, US folks coming out of university optimize for money and head for things like finance - people who hire STEM people. Those more interested in the field than the bucks go for advanced STEM degrees. And if the Yanks don't go for the bucks, they go for the MD.
> US folks coming out of university optimize for money and head for things like finance...
Not really, and I say this as someone who works in VC with peers in PE, Growth Equity, and other segments of high finance.
If you have the resume to get hired as an IB Analyst you will also get hired as a new grad SWE or APM at OpenAI, Google, or Roblox where they would earn the same or more than as an IB analyst with chiller work hours.
People overestimate finance salaries - it's the same as big tech with worse hours.
> And if the Yanks...
I don't think you live here in the States or Canada and as such haven't experienced our job market.
Please butt out of the convo.
---
Edit: can't reply
> Have you experienced the US job market outside of the Valley and NYC
Yes.
Before I switched to VC, I've managed teams and hiring for teams or product lines that reported to me in North Carolina, Georgia, Virgina, Texas, Washington, Massachusetts, and Colorado, and helped open my previous employer's Prague and Warsaw offices.
I also started my career outside of Bay Area or NYC tech before my stint as a staffer.
Additionally, the majority of tech hiring in the US remains consolidated in a handful of geographical locales [0].
> [...] you get a degree of viciousness, competitiveness, and steel-eyed execution
I think there's a lot of truth to that. (Aside: Many manage without the viciousness part. It's not their fault their parents lined them up with an internship and a research paper co-author in high school, and they're not jerks about it.)
Though the current generation of students didn't invent hyper-competitive. Before software engineering jobs (and startups) were high-income and high-status, you'd see that mentality among many people on track for Wall Street, for example.
Another example: Before CS was a go-to for the hyper-competitive, a mentor of mine actually switched from pre-med to CS, at an Ivy, because a percentage of pre-med students were outright sabotaging other students, and it turned him off of the field.
> that a lot of older Americans just aren't used to.
Though, there have been -- and hopefully will remain to be -- people doing it for the love of the field, who are not impressed.
Other than the genuine people being crowded out of admissions slots and fratbro interviews by Wall Street types...
If a Palo Alto helicopter-parented overachiever McDojo black belt tries to pick a fight... with a humble rope-belted person in Asia, who's studied martial arts for the love of it... the latter will chuckle good-naturedly, and help the Californian up off the ground.
> a Valedictorian, NHS member, JV or Varsity sports team member in HS, getting a 2100/1500+ on the SAT, and taking 6-7 APs are now table stakes
This is very true in my experience, except I subbed out Valedictorian with multiple varsity sports/student government and the SAT with ACT and I didn’t even get waitlisted at top schools.
> One of the least effective predictors was unstructured interviews or 'chats’
Yeah those are the worst, one time I had an “interview” with a company that I really liked, the founder is also an awesome guy and we chatted few times and all is well. Then I got invited to their facility, great place and team, some of them were structured on how they evaluate, but most of them were an absolute mess, and some of them were hostile as if I would get hired it will get them fired the day after (the passive aggressive of trying to belittle your projects or work and not trying to understand your approach it but to attack it instead) and when I would ask them in a good faith about something they did, you would get a fake halo effect with “oh I can’t tell it’s secret! NDA bla bla” as if they did a patented work.. it was horrible method to hire people despite the great founder I knew.
In my opinion, the best way is what I usually do, after initial screening, I give them an assignment that they can do in few days and then return the work, the quality of the output will determine that, and it’s exactly how you will do in real work anyway, and you get to measure their critical thinking and problem solving rather than how would they sell or articulate something on the spot (maybe they are overwhelmed and their head went blank), as I am looking for an engineer not a sales dude, and they would tale some time to build and solve it.
I've been talking to a few companies lately, and one just keeps stringing it along, having me talk to manager after manager. It's been 6 weeks, and still no end in sight
If it's been 10+ years and an employer wants your SAT scores, 1600 is as good an answer as any. Anyone asking for that data point doesn't actually care about the accuracy, they just want to see if you'll compliantly jump through a pointless hoop.
(Save the "but that's fraud!" replies. It's not material to the job, so it isn't).
That cost-benefit analysis makes no sense if you had a pretty good scores. What's the point in fudging a few extra points if it means that diligence reveals you to be a liar?
It would also be somewhat suspicious if you went to a so-so college but allegedly had a perfect SAT. It would only make sense to lie if your score was well under 1600, you went to a college that makes sense for someone with a perfect SAT, and you didn't think it was likely they would follow up with a request for the official score report.
SAT isn’t an IQ test, and probably all sorts of people took it before they had cultural awareness or a diagnosis that would have lead to different testing conditions had it been taken after diagnosis, let alone the fact that test scores are not comparable.
GPAs similarly not comparable over large time ranges, schools, or degrees without normalization you can’t get.
SAT isn't a perfect IQ test but it's not bad for a first-pass filter. Nearly all who scored 1600 will be bright and nearly all who score 800 won't be.
It's a bit like BMI. Yes, if you're Peter Dinklage or Arnold Schwarzenegger it will be pretty meaningless. But most people aren't and BMI works pretty well for them.
And the meaning of the score changes over the years based on the test itself changing. Same goes for the company's GPA requirements where there have clearly been shifts across schools on the amount of grade inflation allowed or even encouraged.
As an aside, I'm not sure if I or the College Board can prove my score at this point.
Not only that, but the SAT is not an IQ test and you can definitely study for it. Students with wealthy or motivated parents can get study books or tutors which makes a huge difference in score.
The Princeton Review promises a 200 point score improvement with some of their packages. And they can fairly-reliably achieve it too.
And I guarantee you that claim is based on an intentionally flawed experiment where they take students who have never seen the test before vs after completing the program. The actual control should be against students who have taken a couple of cheaply available practice tests.
I never done an iq test, but just curious, can’t one simply “rehearse/study” for such tests? I dunno, let’s say you do a few past IQ tests (with answers available), I guess one could get a higher score just by doing that
I dislike this argument. I think in some dimensions these types of tests can work, but I’ve never been the type of person who’s been able to score well, and I don’t test particularly well in general, yet I did my PhD work at <IVY LEAGUE> and have had a great career despite this. I think that testing is good for people who can be adequately evaluated, but for people like me it just leads to a lifetime of feeling like something’s wrong with you.
Test-taking may only roughly correlate to intelligence, and is just one dimension of a human... but they likely care less about false negatives (like you) than they do about false positives in alternative assessments.
I can't speak for others here and we don't have SATs in this country, but these things can be very unfair. When I was seventeen, I had a lot of things to deal with which were not of my own making, such as caring for a terminally ill relative and wondering if I would even have a roof over my head a few months down the line. That kind of thing tends to take your mind off school work. Several years later and I was in a much better place.
I think it's fair enough to say teenagers in general have more instability in their life even without this.
Last month 2,400 University of California faculty asked for admissions to resume using the SAT "to ensure foundational fluency." Of course many employers want to ensure that too, especially when college degrees don't anymore.
https://www.insidehighered.com/news/faculty/learning-assessm...The UC faculty opposed the SAT requirement being discarded in the first place. They were overruled by the UC Regents, and that may happen again. And even if the SAT is brought back, I'm sure it will be given much less weight and subjected to the "in a local context" process in the name of equity.
I took the SAT 50 some years ago. I kinda doubt I would do so well on it today without doing remedial prep work on the math.
Apologies, but which part of my comment was this a reply to?
Why is a lossy testing filter better than just failing out those who can't make it? Maybe allow for larger freshmen classes and smaller latter classes or adopt community colleges and have all students start there and advance into the UC system sophomore year on. Instead they bring back what is basically an IQ test for admission.
High graduation rates are an important metric to administrators. If a professor gave a failing grade to 1/3 of the class they would be in hot water.
From the students' perspective, it is better to not be allowed in than fail out midway through. One test is cheaper than years in college.
I always enjoy the advocates who claim that students have mastered their subjects, but "don't test well".
Would you want a pilot on your flight who flunked flying school exams, but somehow "really knew how to fly!"?
SAT score is known to be predictive of college grades. Is it also predictive for whether a mid-to-late career candidate will pass a phone screen? It is used for early career candidate filtering in finance, but I have not heard of anybody caring beyond that because of the availability of signal on the actual tasks they will be performing.
It is correlated tightly with IQ, so yes it will likely be a strong predictive signal for passing a phone screen.
Maybe Maths, but english is probably not correlated tightly with IQ as it is more affected by language background and education.
>english is probably not correlated tightly with IQ as it is more affected by language background and education.
No, see: https://en.wikipedia.org/wiki/G_factor_(psychometrics) English and math has a 0.64 correlation.
> Is it also predictive for whether a mid-to-late career candidate will pass a phone screen?
The answer is probably no. I got many friend they got good marks in SAT, but they were average.
how many elected leaders are in STEM? would high SAT scores and grades exclude many US presidents and congressmen? (yes) winning elections seems kind of important to me. so if you were just like, selecting for leadership - and many of our leaders are brilliant people, just not in the sense of being good at taking tests - would that be good or bad? or... what is your real opinion? what are you actually mad about?
obviously the UC system should give spots to the kids who will use those spots the best. but it is very hard to define what "using spots the best" means.
> not in the sense of being good at taking tests
The trick to doing well on the SATs is to pay attention in class.
We certainly don't want them to fail out, which is what is happening. Berkeley reported 10% failure rate in the intro CS course and 35% in the pre-intro course. https://www.dailycal.org/news/campus/academics/failing-grade...
I have seen company descriptions in job ads that list college achievements of founders. They are invariably young Asian men. I understand that it's a cultural signifier and don't judge them. But, I also understand that I will never hear back from them because I don't share that background. So, I never apply to any job listing that references college experience of either side, other than wanting a degree in general.
Someone once told me that American work culture used to be based more on intern-ship / apprentice type hiring but now obsess with formal degrees. I wonder how much of this shift in culture is influenced by the Korean, Chinese and Indian immigrants, as a formal education is a prerequisite to compete in these countries' job market? For example, it is quite common in India for employers to ask for our 10th and 12th standard marks / grade (because these are national exams) along with college grades - to apparently gauge "Consistency". Fluctuating performance, a break or dropout years all negatively impact you and can be nerve wracking for many freshers, until they manage to get some work experience. It is somewhat disappointing to see this culture permeate to America too, even though I feel quite conflicted about it - after all, everyone does want to hire the best / most competent / reliable candidate; but the other approach - a vocational kind of training - also has its merits and seems to have served American companies well too. (Zoho in India is experimenting with this kind of hiring in India where they are hiring high-school students, mainly from rural areas, and offering them a work cum study program. They don't get any formal diploma or degree though - https://www.zohoschools.com/ ).
> I wonder how much of this shift in culture is influenced by the Korean, Chinese and Indian immigrants
I think there’s an influence, but it’s amplifying a pre-existing trend. Bureaucratic societies favor formal credentials. The U.S. has become much more bureaucratic since the mid-20th century, and credentialism has grown. Reliance on degrees and other formal credentials also enables the universities to achieve political goals through admissions and grading policies. Asian immigrants in the U.S. have readily adapted to that system.
I think it is a positive for an employer to ask for an SAT because it tells me right away I don't want to work for them. Once (a long time ago) I tried to upload my resume to apply for a job. The web page started asking me very basic questions, like a basic aptitude test. I was out. Tell me you do not know how to find and evaluate talent!
I think this is highly age-dependent. I took the SAT well over a decade ago and have significant work experience since then. It would be odd to require me to put down my SAT scores, which I don't even precisely remember.
But if I were < 5 years out of college, and especially if I had gone to school during COVID times (when SATs were not required by many colleges), I would completely understand why an employer might ask.
Basically, colleges used to act as a filter for SAT and other attributes. During the 2020-2025 period, they admitted students under fairly different standards, due in part to testing challenges and social movements.
It makes sense for an employer to want to do a little more diligence to ensure that students who were admitted during this period are similar to students admitted during the prior several decades.
It was required at my first job in 1989, for entry level actuaries.
In fairness, part of job performance was passing the actuarial exams, the first two of which were calculus and statistics. I imagine testing well on the SATs for a math or EE degree (what they hired) was a good indicator of passing tests.
You basically described Canonical's hiring process !
The filter works both ways so it makes sense. Those employers do not want to hire people like you either.
As a manager, there are several qualities that I value highly in an engineer, and they all happen to begin with the letter C: Competent, Consistent, Curious, Caring, and Clear Communicators.
While SAT scores might act as a proxy for competency and possibly curiosity, they're not going to tell you much about whether the person is consistently reliable, whether they care about others and cooperate well, or whether their vocabulary or literary analysis skills have any correlation with their ability to read the room and tailor their communication to their audience.
If I were giving these job posters the benefit of the doubt, I would guess they're including this requirement for the same reason that musicians request particular colors of M&Ms in their riders. They want to weed out people (or bots) who aren't paying attention. Nevertheless, there are better ways to do that than demanding (and presumably filtering by) teenage performance metrics.
C = Competitiveness.
I met an HR manager who had worked for a local but well known company with a reputation for caring about things like GPA and SAT scores. She told me that remembering your SAT scores after college was a sign of a competitive attitude.
I had a friend with an MS show up for first-day of work for a job that asked for SAT scores on the application. HR said "we never got documentation for your SAT scores, can you provide that?" He was on the phone with his mother, having her go through a filing cabinet when he realized that he didn't want to work for a company that was this serious about SAT scores when hiring someone with a post-graduate degree.
That does seem wild. Out of curiosity, did he also have to submit GRE scores, which would be closer in time and more representative of his current knowledge/skills?
Worker w/PhD was asked by Quebec during residency interview to produce High School grades. Bureaucrats will be bureaucrats independent of language I guess.
I'll ask him. This was about 20 years ago.
Company is Alpha Vantage:
https://beaverhand.com/apply/alpha-vantage-gtm-team-various-...
The job description touts investment from some members of the "PayPal Mafia." For some odd reason that fact that and the SAT requirement combine to make this whole thing feel kind of normal.
> Please note that we will also rely significantly on both solicited references (where you introduce us) as well as unsolicited or "back-end" references (where we do our own). For the latter, please rest assured that we will never contact a current employer without first getting your permission.
References already give me goosebumps. Having them reach out to people who haven't given you permission to be a reference sounds like a recipe for disaster.
Every job does that, whether they tell you or not.
If they do sensitive work for the government, it'd make sense that they'd do those back-end references. Also if they are in high end finance where you want to weed out people who have demonstrated moral flexibility was well as total lack of it for certain things.
> If they do sensitive work for the government, it'd make sense that they'd do those back-end references.
If they do government work that requires clearances, the clearance process already covers this sort of investigation on its own.
In any case, they are free to do whatever background checks they want within legal limits, but I'd never apply to a company with such ridiculous hiring processes.
airbnb asked me why i didnt go iit in india and interviever soured after that. it was a chinese guy.
Seriously? This made me lol.
You cannot use the SAT as a metric to compare different cohorts. SAT scoring has been revised many times over the years. When I took it the highest possible score was 1600. From 2004 through 2016 the highest score was 2400. Now it is back to 1600 again. Plus, both the content and the format of the exam has changed many times over the years. At times, there was no essay requirement, at times the essay was required, and at times it was optional. Hence, each year the examination produces a different distribution/histogram of scores even if you normalize the 1600 vs 2400 difference.
Yeah they’re much easier now.
The scores have changed, but ideally they are asking for the percentiles. Those are scaled to the current year.
Even the scaled score is not that informative (and perhaps crosses the line on age discrimination) because for older workers the population of people taking the SAT was much smaller as a percentage of high school grads (and presumably weighted towards higher IQs). It's also why there were so many fewer perfect SAT scores -- smaller population in the bell curve.
Number of perfect scores is also affected by the increase in the number of students who spend 20 hours each a week or more doing SAT prep.
You can look at historical percentile by year and score though.
Which requires them to explicitly ask your age outside the bounds of qualification for a job (over 18 etc). Which ends up opening them to age discrimination lawsuits.
It does not require them to ask about your age, just the year in which you took the SAT. As other commenters have pointed out, this can range from 12 to 17.
Also, they could just ask for your SAT score and any relevant info (if you took it during COVID from your car, etc.) and then you could disclose whatever context you wanted.
It's weird that people think AI breaks the concept of work sample testing. Work sample testing isn't "about" programming, and predates the profession of programming. You can (and some companies do) work-sample test sales account managers, customer support, accountants, whatever.
AI changes the underlying job you're testing for. So, obviously, the tests you might have been using pre-AI won't work anymore; they're testing something that isn't really the job anymore. Update your tests so they're about the real work again, that's all. For coding, that probably means assuming (or requiring) candidates use AI to do your assessment.
What AI really does mess with is conversational/interactive interviewing. We do all our interactive scripted interview on Slack, but I can imagine us having to end that practice and return to face-to-face.
I would never ask for them since its so cringe. But SAT scores correlate to IQ at .81, and IQ is one of the few things that strongly correlates to knowledge work performance positively. There is probably a lot of alpha from knowing candidates SAT scores. Its more useful than knowing the college they went to.
> and IQ is one of the few things that strongly correlates to knowledge work performance positively
i think you mean that it correlates to pay. nobody knows what you mean by "knowledge work performance." reviews of your peers also correlated with pay. often it is not the smartest person who is the most popular. so... do you see how you said something kind of meaningless?
> nobody knows what you mean by "knowledge work performance"
I actually was pretty easily able to deduce what they meant by "knowledge work performance".
It's understandable to be frustrated by not knowing something, but to claim "I don't understand that and therefore no one does and you're being nonsensical" is a bad look.
Consider responding with curiosity rather than defensiveness.
Nope, I meant what I said.
A very good metastudy is "The Validity and Utility of Selection Methods in Personnel Psychology" (by Frank L. Schmidt and John E. Hunter). It summarizes 100 years of research on predicting job and training performance. It makes a very strong case that General Mental Ability (GMA, their word for IQ) is the single most valid predictor of employee success on the job, not just income.
Sneaky age filter? You must be young enough to remember your SAT scores.
Sneaky immigration filter? Most wont have an SAT score at all
Or racial filter: https://nces.ed.gov/fastfacts/display.asp?id=171
After all, it would be illegal to make a hiring decision based on the race you specify you are when you apply. But SAT scores...
Between 2005 and 2015 the maximum score was 2400 instead of 1600. Assuming anyone who got <1600 during that period wouldn't admit it, you now have three well defined buckets.
But of course this is a lot of unnecessary steps compared to the usual method: length of work and education history +18 years.
I remember mine (both from when I was 12 and when I was 17) close to 30 years later.
Hah, you did the one at 12 also? I too remember both, and it was also 30 years ago. I don't know why, probably because they were really important numbers to teenagers, and they say you remember things that happened to you as a teen more than any other part of your life.
Sneaky age filter? You must be young enough to remember your SAT scores.
I can remember mine just fine.
If you're really looking for smart people, use "Answer this word problem in two or more paragraphs. Write your answer on the sheet of paper provided. In cursive."
Having to write the "I did not cheat" pledge in cursive was the most difficult part of the SAT for me.
Are people who were forced to learn cursive smarter? They're just older.
Doing this may well expose you to age discrimination lawsuits, since it's just sneaky indirect age filtering.
Another example would be if you required a minimum SAT score of 1601. Sure, someone could have gone off and taken the SAT as an adult or a young child but in reality it is mainly an age filter.
> Are people who were forced to learn cursive smarter? They're just older.
My kids are learning cursive in elementary school right now, FWIW.
The signal being that the smart people will refuse to jump through this hoop in your inane process because they have a lot of other opportunities to choose from.
"In cursive" is just filtering for people old enough to have been taught cursive.
sadly my kids were just recently taught cursive in elementary school for some unknown reason
Learning it is mostly useful for being capable of reading it, esp. when encountering historical documents (or when encountering old people)
Or encountering California license plates. Or finding Walgreens.
Not wanting to learn cursive is like not wanting to know lower case just because caps lock exists.
I learned in Brazil. Here in the California I asked my son's kindergarten teacher if she would teach cursive, and she said they don't teach calligraphy and I've never seen it described this way, but she's right.
It’s an approximation for an IQ test
They are obviously using the SAT as a safer alternative to the legally dubious practice of IQ testing which can lead to running afoul of the ADA and EEOC. I'm not sure it's much safer, but I am positive it's less safe than doing timed leetcode. At least leetcode problems can be painted as relating to the job.
Additionally, the SAT is a shitty IQ test that is constantly crammed for and cheated on. I remember my SAT test. I was the only person in the room not openly cheating. The teacher proctor didn't care. Higher scores mean better students, more funds, higher home prices, bonuses, and a litany of secondary effects. That's not even including people that pay professional test-takers to do it for them.
The software industry needs to let go of their obsession with finding 10X ROCKSTAR L33T programmers. They never will though. It has gotten worse every few years for decades, and the problems are almost entirely managerial.
IQ testing isn't legally dubious. The idea that it is is an Internet myth. There are a couple household-name corporations that administer general cognitive tests for candidates for some roles.
More companies don't do it because it doesn't work well.
> Higher scores mean better students, more funds, higher home prices, bonuses, and a litany of secondary effects.
Sounds like an IQ test
This by default includes a whole bunch of people who didn't take any kind of standardized tests (most notably, immigrants).
The (albeit small) country I'm from doesn't do any. Reasoning was that standardized tests create an environment where teaching is merely done to create good test scores, not to actually teach.
The SAT doesn’t test course material. It is literally just applying middle school math and English proficiency.
It's not just middle school math. It covers geometry and algebra 2, which very few students complete in middle school.
Starting in 2016 College Board said that they were aligning the test with Common Core, so this might not be true any more.
Canonical?
Heard nothing but bad things about their hiring process.
They made me do an online, timed IQ test, stressing that I should take it in my native language to not waste precious seconds understanding it in english.
It was horribly translated, every sentence was written like something this: "A and B are two broters/sisters. A gives B 3 apples and he/she/them eats one and returns one to he/she/them…" at some point one section had the instructions wrong so I did all the questions wrong. There was no way to change the language or re-read the instructions to try to understand what the original text might have actually been.
That's when I closed the tab.
I'm a Debian Developer.
wild to call out a random org like that…
They're known for asking odd questions like "how did you perform in math in high school" and "please justify your performance in math in high school". It was actually my guess as well before I read the post.
Look at any of their job applications, they're all like this:
https://canonical.com/careers/3752633/application
I just applied to Epic, the EHR company from Wisconsin, and I can confirm that they also ask for SAT scores. Thankfully I have my collegeboard credentials saved
Their pre-screen test was awful. Brain teasers, including some infamous ones like “if you have two coins that make 15 cents and one is not a nickel, how is this possible,” and moderately involved programming questions like “parse phone numbers from a file and record those with any of these area codes OR every other digit is a 2 OR they have multiple pairs of consecutive digits” and you’re given a blank text box with no formatting, IDE, or even non-word processing style indentation help.
Thankfully, the coding assessment does have syntax highlighting for plenty of languages, but in general I feel like face-to-face assessments are more productive. TFA seems to make a similar point
Out of curiosity, did they require SATs, or just ask? And how many years out of HS are you? Seems like it would be crazy to ask for them once someone has finished college and has a UGPA to report.
I see we (as a society) are now full Primeagen: "Epic as in health records, not Epic as in Fortnite."
1600 and your GPA was 4.0. Any other answer proves you aren't trying hard enough to get the job.
4.0, what no APs? Insta-rejected. /s
I like it. I also like it when companies ask for 10 years of [5 year old technology] experience, or say "there's more to working here than the salary!", or other red flags that make it easy to move to the next listing.
If you think my decades old SAT score is relevant, then I know all I need to know about your company.
I only had one company (D. E. Shaw & Co.) ask for my SAT scores. I was late-20s and had to have the recruiter repeat herself two more times before I understood what she was asking.
It was also the single highest density of talent I’ve ever worked, by a long shot. Crazy talented coworkers.
I would think it an inappropriate question if you're asking a PhD (or MS). You could just ask for a copy of the paper(s) or the dissertation/thesis. Some people improve over the time in the uni environment. Some people don't test well when they're bored, etc. Some just grow up.
When I graduated in the 1970s, and looked for my first job, it was expected that some company might ask for such scores and, iirc, one did.
A lot of responses pointing out various flaws with this question, including the fact that it can be used as a proxy for ageism, the fact that the grading scale has not been consistent over time, or that most foreigners will not have gone through the US education system. However, is it really that uncommon for Americans to never have had reason to take the SAT/ACT, such as, simply not going to uni, or going straight to work after graduating high school?
You only need to SAT/ACT for going to get into an university. And these days it seems to matter less. I looked it up and less than 50% of high school students take the SAT/ACT. Personally I never had any employer ask for my SAT/ACT scores since the 90s.
> Take-home projects or a trial period of some kind. This makes the most intuitive sense by far: having candidates do a representative slice of the job gives you a solid idea of whether they'd be any good at it. Combining this with structured interviews was (before AI) considered a gold standard; you'd get a sense of who they are and how they work by talking, have a way to compare them pretty objectively to other candidates because of the structured and consistent nature of the interview process, and then you'd get a sense of how they apply their attributes practically to the job via the work exercise.
Unfortunately a lot of companies have over the last several years been using this to get candidates to do a project for free for them. If it's going to take more than a few hours of my time, I don't take project style interviews seriously unless compensation is added (which some companies do offer and is a big green flag).
Definitely been tricked into working for free a time or two.
Best middle ground I ever had was an interview where they impose a strict 1 hour (maybe it was 90 minutes, idk) time limit between when I got the prompt and when I emailed them back. Then they spent some time looking over it, then I had an interview with an engineer who had read my code and we chatted about it. Why I had made certain decisions, what corners were cut because of the time limit, etc.
Felt very fair. Not enough time to assign a valuable task, enough time and privacy that I wasn't under the gun like you are in a whiteboard interview, and it was pretty applicable to what I would be doing at the company. Solid interview. Didn't get the job, but respected the process.
The SAT vs ACT preference map on Wikipedia is something I had not seen before.
https://en.wikipedia.org/wiki/File:SAT-ACT-Preference-Map.sv...
I took both (since I could just use whichever I scored better on for applications), and the ACT is a much less miserable format. The test is broken up into much larger time chunks, meaning I could take a more useful nap after finishing a 50 minute section in 20 minutes than I could finishing a 20 minute section in 8.
Difficulty wise, the ACT was easier, but not by a lot. They seemed to have pretty similar predictive strengths, based on percentiles of people I knew lining up between the two, but obviously the ACT loses precision because of how coarsely it's scored.
I'm surprised they didn't at least attempt to email them to ask why.
Ignorance is always a possibility here, as it might be their first time hiring.
Soon people will be taking proctored IQ tests just to be allowed to submit a resume.
> Why would you ask for a self-reported, unverifiable test score that could be decades old at this point?
Because many colleges that used to reliably filter for them no longer do (or didn't during a several-year period).
It's true that self-reported scores are not the most accurate, but if I were applying for a job I would report honestly, on the assumption that they could easily request for the scores to be sent by the College Board. The risk/reward of lying does not make sense, at least in my case.
> > Why would you ask for a self-reported, unverifiable test score that could be decades old at this point?
> It's true that self-reported scores are not the most accurate, but if I were applying for a job I would report honestly, on the assumption that they could easily request for the scores to be sent by the College Board.
No, they couldn't, except by going through you (the College Board doesn't take third-party score requests.) You might be able to request that if they are recent enough, but not if they are literally decades old (well, not if they are ~21 years old or older.)
https://satsuite.collegeboard.org/scores/sending-sat-scores/...
I am aware that third parties can't request scores. I was referring to the employer asking to have the scores sent, which the applicant would be compelled to do (or look like they fudged their original score reporting).
I'm also aware that the College Board doesn't hang onto scores forever. I doubt there are any employers who require SAT scores for applicants who took it prior to 200% (the cutoff indicated in your linked article).
I'd rather that than leet code dancing around: "It's not an IQ test, since those are bad, but this is okay though!"
My problem is it's self reported, so it ends up being a "are they smart or did they lie?" game. You can't easily verify it for anybody over 23 or so.
Canonical has a job application where you are supposed to rank yourself on a percentile (up to like 1 in 10,000) on how good you were at math in high school. It's a very easy way to incentivize lying, and also to hire people with an excessively high appraisal of themselves. There are a lot of people who are reasonably good at math, and have avoided humbling environments like the Putnam, and have convinced themselves they are God's gift to math, when in reality they were just the brightest kid in a class of 100 high school students.
Fair enough, but the vast majority companies that interview like this would be totally fine with the brightest 1 of 100, if not the top 20 of 100.
Another thing making a comeback -- reference checks. I had to supply references for my first job in 1999. Then I wasn't asked for them again until 2024, and then for every job after that.
Whatchoo talkin about, Willis? I've had to supply references for every job I've had.
Somehow it seems every other hackernews was living in a much better timeline, job-wise, than I during the ZIRP era.
If you want the job, can’t you just lie? Or are SAT scores something that cannot be faked? I dunno, I also say I know Kotlin when I have more experience in Java (and honestly I couldn’t care less about specific tech stacks), or that I know about tcp/udp when all I have is read a couple of (good) books about it.
I don’t feel bad lying about some stupid requirement
> If you want the job, can’t you just lie? Or are SAT scores something that cannot be faked?
I would assume that if you progress to the point of an offer, they would ask you to have the official scores sent by the College Board. Apparently they hang onto scores back to 2005 and can send them for a fee.
I’ll do you one better:
I was denied a role with a major engineering firm based on my 3.something GPA!
They needed a 3.4 or 3.5.
Many years ago I was walking around a college job fair when a recruiter was yelling at a college student. His crime: "wasting everybody's time" by presenting a Resume with a 3.4 GPA when the company clearly listed a 3.5 GPA as the minimum they would accept.
If this is how they treat people that don't yet work for them, it doesn't bode well for how they will treat people that work for them.
As I mentioned elsewhere on HN [0], younger generations are much more competitive now.
Visit and talk with undergrads at a top CS program like Stanford, Cal, UIUC, MIT, etc. The culture is different because this is a much more competitive generation. When the acceptance rate into a top CS program is in the 1-5% range and laurels like being a Valedictorian, NHS member, JV or Varsity sports team member in HS, getting a 2100/1500+ on the SAT, and taking 6-7 APs are now table stakes, you get a degree of viciousness, competitiveness, and steel-eyed execution that a lot of older Americans just aren't used to.
This mindset is the norm across Asia though - from the Gaokao to the JEE to SKY-or-bust. Honestly, I'm glad that younger generations are much more competitive now - pressure makes diamonds.
And honestly, the top 40-50 STEM programs nationally graduate around 30-40k new grads a year. Add to that respected regional programs and Veteran-to-Employment pipelines and you have a self-sustaining talent pipeline.
[0] - https://news.ycombinator.com/item?id=48506001
I disagree with the increase in competitiveness being a good thing. Excessive filtering at all levels has meant that eccentrics or absent minded professor types are not making it into research roles, and creatives or mad geniuses are filtered out before they have the chance to make an impact. There are a lot of people who are extremely bright and creative, but just don't have it all together the whole time from ages 14-25, and these days they have no chance of making it into research positions.
The system is rewarding conscientiousness and consistency over creativity.
> The system is rewarding conscientiousness and consistency over creativity
This assumes that you can get to the top via rote skills alone. Rote learning only gets you so far and most of those kinds flame out.
It's hard to describe, but once you meet actually talented people what you end up seeing is that they're just extremely diligent and deeply passionate about a topic and will continuously execute.
For example, when I was in HS I wrestled. Yes there were physical differences that could impact a sparring round, but technique and preparation was almost always able to outcompete base innate talent. Later, I ended up learning ballet the Russian style and it was the same - the truly creative types who were at Vaganova or Paris had already built strong fundamental and technical skills which allowed them to mix and match and create.
You cannot be creative without also being diligent and understanding fundamentals.
The "eccentrics" and "mad geniuses" are few and far between, and to find people with talent, you do need to use exclusionary tactics like scores and interview performance.
Not really. Outliers of that sort get dealt with as an exceptional case and it works. I've seen amazingly bright batchmates get into all sorts of programs without most of the qualifications, because they were a genuine math wizard (the kind that submitted errata to a standard textbook on some weird number theory stuff at age 15), but they didn't always score the best on many exams, especially when aggregated across subjects. IIRC he's a researcher now. All trends point to him being an eccentric old hag at age 50 ;)
The filtering system is meant for the majority case and there it works. The outliers get dealt with as outliers, which also works. In this case, he later asked the author of that textbook who he emailed with the errata, to connect him with the group he wanted to work in. Needless to say it was a very strong referral.
The US education has always been competitive. In sports.
Its school system has always been a state-sponsored daycare.
SAT/ACT tests reflect this. I can get a perfect score in SAT math easily. And I likely could do that as a kid (I never took standardized tests at school). I wouldn't have been able to get the perfect scores in the Chinese gaokao or Korean/Japanese tests.
> There are a lot of people who are extremely bright and creative, but just don't have it all together the whole time from ages 14-25, and these days they have no chance of making it into research positions.
This is just nonsense. Are you saying that we should kick out smart kids with high test scores to let in absent-minded students who care about only getting drunk so that they _might_ become great researchers in their 30-s?
To the topic at hand: it's way too easy to fluff your resume with nonsense like "Coordinated a responsible team for an implementation of cross-cutting concerns improving customer retention change by 12.23% across the organization". Test scores provide at least some objective measurement.
I'm not opposed to the use of test scores, it's more the requirement that kids have the whole package of grades+activities+honors societies+test scores consistently through a phase of life that is widely understood as tumultuous for many. We're selecting for robots.
There's a reason the west was so productive in terms of new scientific and technological discoveries in the twentieth century, and it's not that our scientists were the most consistent conscientious students who prepared extensively for exams and padded their resumes in just the right way.
> We're selecting for robots.
And to add to this, learning itself is a _skill_. Working on a complex problem, looking at it from different angles, spending time memorizing facts, working on learning to paint fine lines - these are all skills that you need to master.
By not motivating children to do that during their formative years, you set them up for failure later in life.
Yes, there will always be exceptions, humans are extremely variable. But for the general case just letting children float along without any goals or competition is not a great general strategy.
> the whole package of grades+activities+honors societies+test scores
This is _also_ a very US thing. Without true competition, students have to fluff their "resumes" with nonsense to get admitted into good colleges.
Other countries have tougher tests that can provide a better signal.
> We're selecting for robots.
I disagree. We're selecting for people who can set a goal and follow it.
Apply the same arguments to sports. Should we not stop all the competitions until the age of 25?
I didn't really state my whole view, but hard exams like Oxford and Cambridge use make sense for elite colleges to use (rather than unrelated extracurriculars), but reforming the whole education system to be oriented around a single high stakes test like in China or Korea has its own severe costs. I do not want high schoolers to spend 20 hours a week at hagwons, and the current resume filler system is also terrible. I do not think declining admissions rates at elite universities reflect that the students are any smarter or more prepared than they were 20 years ago, but rather they are much more cutthroat about many things that are orthogonal to being successful adults.
Ideally we'd follow a more exam focused system more like the UK, though I wouldn't want to require all students to only study 3-4 subjects towards the end of high school. But something in the european model of IB/Abitur/A Levels, where there are serious exams in various subjects at the end of high school for all uni bound students, plus some special higher level exams for the most elite unis (in the vein of Cambridge's Sixth Term Examination Paper). We could probably repurpose AP exams to fill a similar admissions role to A Levels, and possibly use the AMC/AIME/USAMO more explicitly for admission.
Edit: this doesn't let me reply again, I think the chain got too deep. But the point is that we're not just using AP exams and USAMO scores, we're also using a pile of other metrics around extracurriculars, GPA, and honors societies, and the end result is stressed out children and not actually having better outcomes than we did two decades ago. Declining admissions rates at elite colleges do not reflect smarter and more productive incoming students.
Yeah, Chinese-style gaokao would be impossible in the US. And I think it goes _too_ far into the "competitive high pressure" direction.
There are many ways to make a more competitive and objective system. I honestly don't have a lot of professional experience with any particular one, so I don't have strong opinions on a particular form it should take. A European model is good, some kind of mix of Chinese+European would also be great. And ultimately, these systems would be more fair for applicants.
And the current topic just highlights the ridiculousness of the status quo. For most people in the US, SAT is the _only_ objective test score that they have.
> We could probably repurpose AP exams to fill a similar admissions role to A Levels
They are already being used like that in college admissions today.
> possibly use the AMC/AIME/USAMO more explicitly for admission
Already in use explicitly at most of the top CS programs today
> I do not think declining admissions rates at elite universities reflect that the students are any smarter or more prepared than they were 20 years ago, but rather they are much more cutthroat about many things that are orthogonal to being successful adults
The cutthroatness is what we should be optimizing for long term. Competition is what begets innovation.
Also, even at solid middle tier universities like UC Riverside, the calibre of student has increased dramatically over the past decade [0]
[0] - https://news.ycombinator.com/item?id=39369724
> There's a reason the west was so productive in terms of new scientific and technological discoveries in the twentieth century
> it's not that our scientists were the most consistent conscientious students who prepared extensively for exams and padded their resumes in just the right way.
Instead, a large portion were immigrants or the children of immigrants who arrived in the US as part of trans-national brain drain from countries with strict education systems (eg. Hungarian Jewish Americans in WW2, Eastern Europeans in the 1980s to present, Asian Americans today).
There's a reason Asian Americans, Eastern European Americans, and immigrant African Americans are overrepresented in leadership and white collar industries despite the very real handicap of having extended periods time without US citizenship or a greencard.
Instead of optimizing for feel-happy edge cases, we should be optimizing for building the best talent where possible, and that requires being competitive.
> We're selecting for robots.
Frankly, this is insulting as well. Yes there are some late bloomers, but they are outliers. If they can truly succeed they would stil find a non-beaten path to succeed in a competitive ecosystem.
> consistently through a phase of life that is widely understood as tumultuous for many
Only to y'all "heritage" Americans. For those of us who are kids of immigrants, we learnt that life is a race, either you compete or you fall to the wayside.
Of late, US folks coming out of university optimize for money and head for things like finance - people who hire STEM people. Those more interested in the field than the bucks go for advanced STEM degrees. And if the Yanks don't go for the bucks, they go for the MD.
> US folks coming out of university optimize for money and head for things like finance...
Not really, and I say this as someone who works in VC with peers in PE, Growth Equity, and other segments of high finance.
If you have the resume to get hired as an IB Analyst you will also get hired as a new grad SWE or APM at OpenAI, Google, or Roblox where they would earn the same or more than as an IB analyst with chiller work hours.
People overestimate finance salaries - it's the same as big tech with worse hours.
> And if the Yanks...
I don't think you live here in the States or Canada and as such haven't experienced our job market.
Please butt out of the convo.
---
Edit: can't reply
> Have you experienced the US job market outside of the Valley and NYC
Yes.
Before I switched to VC, I've managed teams and hiring for teams or product lines that reported to me in North Carolina, Georgia, Virgina, Texas, Washington, Massachusetts, and Colorado, and helped open my previous employer's Prague and Warsaw offices.
I also started my career outside of Bay Area or NYC tech before my stint as a staffer.
Additionally, the majority of tech hiring in the US remains consolidated in a handful of geographical locales [0].
[0] - https://www.bls.gov/oes/2022/may/oes151252.htm
Have you experienced the US job market outside of the Valley and NYC?
> [...] you get a degree of viciousness, competitiveness, and steel-eyed execution
I think there's a lot of truth to that. (Aside: Many manage without the viciousness part. It's not their fault their parents lined them up with an internship and a research paper co-author in high school, and they're not jerks about it.)
Though the current generation of students didn't invent hyper-competitive. Before software engineering jobs (and startups) were high-income and high-status, you'd see that mentality among many people on track for Wall Street, for example.
Another example: Before CS was a go-to for the hyper-competitive, a mentor of mine actually switched from pre-med to CS, at an Ivy, because a percentage of pre-med students were outright sabotaging other students, and it turned him off of the field.
> that a lot of older Americans just aren't used to.
Though, there have been -- and hopefully will remain to be -- people doing it for the love of the field, who are not impressed.
Other than the genuine people being crowded out of admissions slots and fratbro interviews by Wall Street types...
If a Palo Alto helicopter-parented overachiever McDojo black belt tries to pick a fight... with a humble rope-belted person in Asia, who's studied martial arts for the love of it... the latter will chuckle good-naturedly, and help the Californian up off the ground.
> a Valedictorian, NHS member, JV or Varsity sports team member in HS, getting a 2100/1500+ on the SAT, and taking 6-7 APs are now table stakes
This is very true in my experience, except I subbed out Valedictorian with multiple varsity sports/student government and the SAT with ACT and I didn’t even get waitlisted at top schools.
Extreme competition isn't good. It will just lead to race to bottom.
“ getting a 2100/1500+ on the SAT” Typo? 2100/2400 << 1500/1600 in terms of rarity.
> One of the least effective predictors was unstructured interviews or 'chats’
Yeah those are the worst, one time I had an “interview” with a company that I really liked, the founder is also an awesome guy and we chatted few times and all is well. Then I got invited to their facility, great place and team, some of them were structured on how they evaluate, but most of them were an absolute mess, and some of them were hostile as if I would get hired it will get them fired the day after (the passive aggressive of trying to belittle your projects or work and not trying to understand your approach it but to attack it instead) and when I would ask them in a good faith about something they did, you would get a fake halo effect with “oh I can’t tell it’s secret! NDA bla bla” as if they did a patented work.. it was horrible method to hire people despite the great founder I knew.
In my opinion, the best way is what I usually do, after initial screening, I give them an assignment that they can do in few days and then return the work, the quality of the output will determine that, and it’s exactly how you will do in real work anyway, and you get to measure their critical thinking and problem solving rather than how would they sell or articulate something on the spot (maybe they are overwhelmed and their head went blank), as I am looking for an engineer not a sales dude, and they would tale some time to build and solve it.
I've been talking to a few companies lately, and one just keeps stringing it along, having me talk to manager after manager. It's been 6 weeks, and still no end in sight
If it's been 10+ years and an employer wants your SAT scores, 1600 is as good an answer as any. Anyone asking for that data point doesn't actually care about the accuracy, they just want to see if you'll compliantly jump through a pointless hoop.
(Save the "but that's fraud!" replies. It's not material to the job, so it isn't).
That cost-benefit analysis makes no sense if you had a pretty good scores. What's the point in fudging a few extra points if it means that diligence reveals you to be a liar?
It would also be somewhat suspicious if you went to a so-so college but allegedly had a perfect SAT. It would only make sense to lie if your score was well under 1600, you went to a college that makes sense for someone with a perfect SAT, and you didn't think it was likely they would follow up with a request for the official score report.
"You're partly making your decision based on who someone was as a 17 year old."
Sure, but IQ tests show a high degree of stability over a person's life. It's not unreasonable to be interested in it for sorting.
SAT isn’t an IQ test, and probably all sorts of people took it before they had cultural awareness or a diagnosis that would have lead to different testing conditions had it been taken after diagnosis, let alone the fact that test scores are not comparable.
GPAs similarly not comparable over large time ranges, schools, or degrees without normalization you can’t get.
SAT isn't a perfect IQ test but it's not bad for a first-pass filter. Nearly all who scored 1600 will be bright and nearly all who score 800 won't be.
It's a bit like BMI. Yes, if you're Peter Dinklage or Arnold Schwarzenegger it will be pretty meaningless. But most people aren't and BMI works pretty well for them.
SAT isn't an IQ test but IQ is very correlated with SAT scores.
While edge cases may exist, for the most part, a test score / GPA / anything where the numerator is almost as big as the denominator is a good signal.
Can you say the same for SAT tests, where the score is best of N and N is however many the person can afford and varies between candidates?
And the meaning of the score changes over the years based on the test itself changing. Same goes for the company's GPA requirements where there have clearly been shifts across schools on the amount of grade inflation allowed or even encouraged.
As an aside, I'm not sure if I or the College Board can prove my score at this point.
Not only that, but the SAT is not an IQ test and you can definitely study for it. Students with wealthy or motivated parents can get study books or tutors which makes a huge difference in score.
The Princeton Review promises a 200 point score improvement with some of their packages. And they can fairly-reliably achieve it too.
And I guarantee you that claim is based on an intentionally flawed experiment where they take students who have never seen the test before vs after completing the program. The actual control should be against students who have taken a couple of cheaply available practice tests.
You can study for most IQ tests as well.
Yes; though SAT is less prep-resistant and it'd be smart to apply a "+/- 100pts" fuzz to a score.
I never done an iq test, but just curious, can’t one simply “rehearse/study” for such tests? I dunno, let’s say you do a few past IQ tests (with answers available), I guess one could get a higher score just by doing that
You can. IQ scores are not stable throughout one’s life.
I dislike this argument. I think in some dimensions these types of tests can work, but I’ve never been the type of person who’s been able to score well, and I don’t test particularly well in general, yet I did my PhD work at <IVY LEAGUE> and have had a great career despite this. I think that testing is good for people who can be adequately evaluated, but for people like me it just leads to a lifetime of feeling like something’s wrong with you.
Test-taking may only roughly correlate to intelligence, and is just one dimension of a human... but they likely care less about false negatives (like you) than they do about false positives in alternative assessments.
I can't speak for others here and we don't have SATs in this country, but these things can be very unfair. When I was seventeen, I had a lot of things to deal with which were not of my own making, such as caring for a terminally ill relative and wondering if I would even have a roof over my head a few months down the line. That kind of thing tends to take your mind off school work. Several years later and I was in a much better place.
I think it's fair enough to say teenagers in general have more instability in their life even without this.
If you think it matters (I don’t) and think you scored unfairly low at 17, you can take (or retake) the SAT at any age.