I've always been fascinated by the philosophical divide among the nations with the best educational systems. Some (i.e. Finland) deemphasize tests, while others such as India and South Korea place strong emphasis on them. The issue is not black and white enough for us to categorically embrace or condemn tests. Different cultural and environmental factors are obviously in play.
Idk if the Indian educational system is considered the best (I'm actually pretty sure that it's not). As for the South Korean educational system, idk if it should be considered a model to follow considering kids there spend like 16 hours a day in school but perform comparatively well as kids in Finland who spend maybe less than half of that time in school.
As for the South Korean educational system, idk if it should be considered a model to follow considering kids there spend like 16 hours a day in school but perform comparatively well as kids in Finland who spend maybe less than half of that time in school.
The point is that the children who learn through standardized tests are performing just as well as, if not better than, those in an environment where tests are deemphasized.
Having said that, it's hard to make the cross-cultural argument that standardized testing is always positive or negative.
Yes, I'm aware of IIT. That does not really disprove my point though. Your argument is similar to if someone was said that Mexico is a poor country and someone made a counterargument saying that Carlos Slim Helu lives in Mexico ergo it's not a poor country.
> The point is that the children who learn through standardized tests are performing just as well as, if not better than, those in an environment where tests are deemphasized.
Kinda hard to argue with this if you are not providing what sort of metrics you are using.
I'm glad I applied to college before ubiquitous digital storage was a thing...I vaguely remember trying to come up with contrived ways to sound like an interesting person. I think I even mentioned listening to country music.
In terms of measurements of character, has there been much thought given to looking at longitudinal data of a student's career? An easy one would be improvement on standardized tests over time. I qualified for the National Merit Award because I happened to be doing prep work for SATs...I didn't even know the PSATs meant anything, but just because I performed well on one test, I was the beneficiary of all those schools who use "Number of National Merit Scholars in freshman class" as some kind of quality metric...My SAT performance afterward was only middling-good. I don't think that should've put me in a higher bracket than someone who struggled at first, but rather than giving up, did what it took to reach a higher percentile.
But what about students who go from being a flop, say, in their sophomore year, to really growing in terms of performance and accomplishments by their senior years? A student who, over 4 years, has a C average, but by senior year, is acing AP classes...that takes a considerable amount of grit. And the student is on the upswing. Hell, even a student who did no sports until senior year, but manages to letter on a high-performing team in a sport new to him/her...that to me is about as outstanding as a state champion who competed since the age of 5.
Maybe it isn't quite time for this question to be asked, but I'm damn well sure that the day is soon.
Why are students being admitted into undergraduate programs any more?
There are no admissions for the Khan Academy, and a huge chunk of undergraduate college education could be done in exactly that format.
Donate to have units created very well (and a few different versions just because) copyleft. Pay Universities for thorough certification of knowledge for particular subjects, and finally join Universities for capstone activities which will grant degrees upon completion. Keep graduate-level education as is.
Why do there have to be tens of thousands of Calculus I courses every year when one or two online would do? The technology and market exists for this, but the inertia of an antiquated undergrad education system gets in the way.
Let's ignore all of the arguments around Khan Academy (pretend it's a foregone, rational conclusion that it is superior) and focus on people:
Humans are irrational by default (https://www.youtube.com/watch?v=GDAzsZLvfPw). This is more or less proven in many behavioral experiments (Too lazy right now to source them) when people will do the obviously wrong thing based on internal logics.
Because of this, we can look to the irrational beliefs of people that influence the schools. Namely, social and physical proof. And the social proof has a major benefit: it's less likely that someone is bullshiting credentials if they have an actual degree to back it. http://xkcd.com/451/
Still, it will be a while before irrational opinions catch up (or are leapfrogged by locations like Ghana that embrace the new medium) and that there are ways to catch cheaters.
is a fine word substitution. but the problem with your observation (assuming it is correct), is that nobody is paying money for commodity education. people are lining up to pay money (in the hope that) a "peer reviewed" credential will set them apart.
I suspect that the only aspect of undergraduate education that requires the reputation, commercialism, or exclusivity of "traditional" universities is testing. Scoring well on an electrical engineering test administered by MIT might be just as good as taking the course there.
I doubt that I would have survived an online education, though I can't say for sure.
I know someone who got most of an undergrad degree online, in a substantive field, by necessity -- she was a teenage mom working a job and caring for an infant in a rural town. Her mom told me the hardest things about online college for her daughter were the loneliness and crushing boredom. It's great that she had that opportunity. I'm glad she's tougher and more focused then I'll ever be.
The atmosphere at a small college is where I was able to keep up with my coursework. It got me out of my hometown. I got over my crippling shyness, and learned to interact with people in a civilized way. Graded assignments "forced" me to do the work.
In my view one thing we still don't know about online education is how many people can actually get through it. We also still don't genearlly know what an "educated person" will consist of, who sat in their parents basement for 4 years staring at a computer screen.
Thanks for clearing that up. As I was reading your comment it seemed like you were of the opinon that online education is of less value than face to face. My mistake :)
While I can understand that for basic stuff, there is no way I would ever attend an online course on some advanced undergraduate topics in physics or mathematics online.
There is a benefit of being in a classroom environment when learning an advanced subject. You have a direct connection to the professor and other students, which is essential for advanced material. Plus you can easily go and talk to them afterwards for more details. I learnt a lot by going to professors office hours which doesn't work online.
Also, what about labs? Every science program I am aware of has a laboratory component. That cannot be done at home. Plus good luck trying to get universities to agree on what labs to do. Each university has different types of researchers who focus on certain areas, which is reflected in the labs selected. My former university, for example, had no particle physics people so any lab related to particle physics would be misplaced. Instead they had a lot of quantum people so the labs were more geared towards that. I know of another university that had a lot of particle physics people and their labs reflected this.
Another thought is research. How would undergraduates get exposed to research in your system?
Thinking about it more, having TA'd first year calculus, you need to be able to interact in real time with students. I think online resources are great and I recommended them to my students, but they cannot replace face to face interaction. This is something online technologies cannot replicate, so I say it is a bad idea to even do it for basic things like calculus, which is very fundamental to the sciences.
Personally, I think it is a noble goal, but unlikely. Instead, I think the universities should integrate online stuff in addition to in classroom lectures.
EDIT: I forgot to mention the university environment. Being surrounded by my peers in the same field really helped. I learnt a lot just by hanging around the various clubs. That absolutely cannot be replicated online.
It would be hard to do online courses for subjects that require a lot of physical tools for teaching, or loads of physical activity. Some examples are physics, mechanical engineering, computer hardware design, dancing etc.
But it might be useful promote online education more since it increases accessibility and affordability.
> You have a direct connection to the professor and other students, which is essential for advanced material.
You can ask the student to write down their questions and mail the professor. This way, a good professor can be shared by many more students.
>How would undergraduates get exposed to research in your system?
The professors can guide them to where interesting conversation happens on the Internet, and they can get exposed to interesting research there.
>I forgot to mention the university environment. Being surrounded by my peers in the same field really helped.
There is no law that this atmosphere of dense population of smart people should be limited to universities.
By the way, I am agreeing to your central point, which is that online education should be a complement and not a substitute to conventional education.
> The flaws in standardized testing are well-documented at this point.... their predictive powers only forecast a student’s progress as far as the first semester of their freshman year.
Let's be honest. Raise your hand if you needed markedly
more than your first semester of freshman year to figure
out which of your friends at college were going to cut it
in some traditional academic sense.
Not saying that's the only sense, but as far as the
college should be able to measure, academics for 4(+) years are the only fair initial barometer. Of course, academics can't say anything about who can write a 7+
figure donation to the college 30 years after graduation...
If you asked high-schoolers (not necessarily even seniors) which of their peers would do the best in college, do you think they would give useful answers?
Does anyone know how/where one might get access to a data set like the one he uses in this paper? (1993 student demographic data, high school GPAs, SAT scores, and freshman GPAs)
There is a nonsense propaganda behind these critique on standardized testing.
Test scores will always always always always always always always always always always reflect socioeconomic disparities, because there is and will always always always always always always always always always always be socioeconomic difference in student abilities. It is insane to say standardized tests "punish disadvantaged students and minorities". Tests are tests, they do not and should not care about students' socioeconomic background. Whether universities should account for socioeconomic status of applicants or not, that's their consideration, not test makers' consideration.
Of course standardized tests do not measure all aspects of a student, but they never claim they do. This critique is also silly. The problem is, it is extremely difficult to measure "traits such as optimism, curiosity, resilience" of a person, especially when you want to use the results in college admission. Current psychological instruments for those traits are not developed for that purpose. No serious psychologist will suggest colleges to use those psychological instruments in college admission.
Do no blame standardized testing for those silly reasons!
> It is insane to say standardized tests "punish disadvantaged students and minorities"
A test can be written with a bias that favors or punishes any group. I don't think its intentional, as it is quite difficult to create them without these issues.
A simple example to show the point:
Which of the following is not like the others:
A) Chicken
B) Moose
C) Cow
D) Pig
E) Dog
Many people would answer A: A chicken is a bird and everything else is a mammal.
But if you come from an agricultural background, in addition examining the animals based on their Biological Kingdom, you might be tempted to say B, Moose: It's the only animal that has no role on a farm. B is not an illogical answer from that perspective, but within the context of the test it's likely wrong.
When people say a test has a socioeconomic bias, they're saying a question can become easier or harder depending on a student's economic background. The question above is easy for most of us, but would probably be solved slower by anyone who grew up on a farm.
> When people say a test has a socioeconomic bias, they're saying a question can become easier or harder depending on a student's economic background.
That's not their argument.
However, actually the issue you mentioned has long been noticed by psychologists / test makers. Experts in SAT / ACT know a lot about these things. Please refer to Differential Item Functioning https://en.wikipedia.org/wiki/Differential_item_functioning
SAT / ACT both have best experts (psychometricians and statisticians). Those are very basic things for them. Can you tell me what is your reference? From what I read, I think SAT is addressing other issues.
The new changes haven't been tested at all from what I've read. They're dropping vocabulary tests, for example, which are astonishingly predictive. From the media coverage it doesn't look like the College Board are as interested in scholastic aptitude as plaudits for Doing the Right Thing™.
Colleges are attempting to deal with claims of racial bias by watering down their standards. The bias claims are there because disadvantaged groups are well... Disadvantaged. They have a shitty education in primary school that poisons the well for them.
Why should character or "heart" matter? What if you're a smart kid who is a real asshole? Or a dumb kid with a good sales pitch? Or you didn't participate in extra-circular activities for whatever reason? Why should some guy wading through 10,000 applications be making some evaluation and judgement of my character?
They should drop the pretense of some sort of fair selection process and just do a pure lottery. If you can't read and get into Harvard, you'll wash out quickly.
For elite schools atleast, I don't think it has anything to do with watering down their standards. These schools have the pick of the crop and if 3.3million high school students graduate in the US alone every year, the top 1% of students make up about 33,000 applicants, far more than the 2000 or so Harvard admitted in last year.
The issue really morphs from "How do I select the best students for my school?" to "What sort of arbitrary metrics can I use to filter down this list of already amazing students?" If I had to guess however, I'd say these schools, under the guise of character, are trying to select students that have the highest potential to give back to the university. "Oh this student does part-time acting? Well we will accept her on the chance she becomes the next Natalie Portman. This guy has a best selling app? Better put him on the list of potential Zuckerbergs."
I'm not an admissions officer (not even close), but that would be my best guess on what someone an admissions officer would do faced with the problem of having so many amazing students to fill their undergrad class 10 times over. At some point it no longer makes sense to filter on raw academic ability (see also, Google's "GPA does not matter"), and a purely random lottery would 1.) never actually be a random lottery (thank you alumni) and 2.) may massively favor one group of people (if I have to choose 2000 our of 33,000, and 60% of that 33,000 are asian from a 150,000k+ income family, you run the risk of homogenous class - if you have to ask whats wrong with this, consider this, would you like to work at a job that hired exclusively white men (or whatever race/gender you are)).
Another way to look at it would be this way: if you were given the task to hire a new team of software engineers, and once the resume's came in they all had, 3.9+ college GPAs, Ph.ds in CS, and multiple years of experience on your tech stack of choice, would you just "do a pure lottery" to see who works for your company? Chances are you might end up filtering them on the most circumstantial of metrics (i.e. "This guy was a Yankees fan, I don't think we would get along very well")
The pool of qualified applicants is not as deep as you think. I am doing a lot of college recruiting interviews and students at Harvard, MIT, Berkley are pretty good but the quality goes down very quickly after top 10. We consistently make offers to ~50% of students from MIT but only to 10%-15% of students from top 20-30 university.
> If you can't read and get into Harvard, you'll wash out quickly.
A college does a disservice to students to admit them when they will wash out.
My experience as a student at Caltech in the 70's was they did a pretty good job with the admissions. They selected for smart & motivated & interested in tech. As a result, living in the dorms full of this kind of people, there was never a dull moment.
I only knew a couple of students who clearly didn't belong there, and they soon left.
One friend of mine left his sophomore year, as he was doing badly. About 10 years later, he asked if he could come back and try again, and Caltech said sure. He then proceeded to get straight A's and graduate. I asked him if he felt that he did better the second time because he'd gotten smarter. He laughed, and said no, it was just that in his late 20's he was then ready to work.
They are doing that now to disadvantaged students. An acquaintance in my state university was a smart guy from a notoriously awful high school.
He was completely overwhelmed and dropped out his first year -- he just wasn't prepared for the material, and the remedial help offered by the college wasn't very useful to someone who wanted to be in a STEM track. He dropped out, joined the Army, and flourished... we're Facebook friends, and he has an MS in CS now.
The question is whether higher education should perpetuate pre-existing socio-economic disparities, or seek to compensate or overcome them. The vast majority of educators believe in the latter.
Thus they need to see beyond the present circumstances of each applicant and discern which will be most likely to excel, taking into account the impact of their 4-year educational process.
Some students from great backgrounds will fail to keep up in college. While some others who come from disadvantaged backgrounds are able to "catch up" and pass others in achievement by the end of college.
How can they predict those outcomes in advance? That's the question that the SAT and ACT were created to answer, and now we're finding that they don't actually answer it better than alternatives. In part because it is becoming clear that there is more to success in life than how well a person solves word problems.
Can you provide more background on your statement that the SAT and ACT are not achieving the goals for which they were designed? My suspicion has been that things are changing because aptitude tests are unpopular for socio-political reasons rather than for any want of predictive power.
Basically, the tests were supposed to provide an objective means to predict student outcomes. But they are no better predictors than high school grades alone. And the scoring shows significant correlations with factors like gender, race, language, family income, etc.--so they're not objective.
From this article:
> The University of California, Berkeley1 economist Jesse M. Rothstein has found that the combination of a student’s high school grades and demographic information predicted first-year grades in college about as well as her high school grades and SAT scores do.
> DePaul is implementing their own tests for non-cognitive skills, with a series of essay questions. For the entering class of 2012, about 10 percent of applicants (or about 5 percent of the freshman class) chose not to send ACT or SAT scores. Instead they completed four short-answer questions, designed to measure their leadership skills and their ability to meet long-term goals. Systematically scoring the responses to those questions, DePaul reported that the freshman-to-sophomore retention rate was almost identical for those who submitted standardized test scores (85 percent) and those who did not (84 percent).
I have said this before, but the admissions system at so many US colleges looks very strange to British eyes.
There is special treatment for people who play sport, whose parents went to the same college (especially if they give money) and for those who have the right sort of hobbies.
If that was not bad enough, there are different rules for different ethnic groups.
41 comments
[ 5.8 ms ] story [ 109 ms ] threadhttp://en.wikipedia.org/wiki/Indian_Institutes_of_Technology
As for the South Korean educational system, idk if it should be considered a model to follow considering kids there spend like 16 hours a day in school but perform comparatively well as kids in Finland who spend maybe less than half of that time in school.
The point is that the children who learn through standardized tests are performing just as well as, if not better than, those in an environment where tests are deemphasized.
Having said that, it's hard to make the cross-cultural argument that standardized testing is always positive or negative.
> The point is that the children who learn through standardized tests are performing just as well as, if not better than, those in an environment where tests are deemphasized.
Kinda hard to argue with this if you are not providing what sort of metrics you are using.
In terms of measurements of character, has there been much thought given to looking at longitudinal data of a student's career? An easy one would be improvement on standardized tests over time. I qualified for the National Merit Award because I happened to be doing prep work for SATs...I didn't even know the PSATs meant anything, but just because I performed well on one test, I was the beneficiary of all those schools who use "Number of National Merit Scholars in freshman class" as some kind of quality metric...My SAT performance afterward was only middling-good. I don't think that should've put me in a higher bracket than someone who struggled at first, but rather than giving up, did what it took to reach a higher percentile.
But what about students who go from being a flop, say, in their sophomore year, to really growing in terms of performance and accomplishments by their senior years? A student who, over 4 years, has a C average, but by senior year, is acing AP classes...that takes a considerable amount of grit. And the student is on the upswing. Hell, even a student who did no sports until senior year, but manages to letter on a high-performing team in a sport new to him/her...that to me is about as outstanding as a state champion who competed since the age of 5.
Why are students being admitted into undergraduate programs any more?
There are no admissions for the Khan Academy, and a huge chunk of undergraduate college education could be done in exactly that format.
Donate to have units created very well (and a few different versions just because) copyleft. Pay Universities for thorough certification of knowledge for particular subjects, and finally join Universities for capstone activities which will grant degrees upon completion. Keep graduate-level education as is.
Why do there have to be tens of thousands of Calculus I courses every year when one or two online would do? The technology and market exists for this, but the inertia of an antiquated undergrad education system gets in the way.
Humans are irrational by default (https://www.youtube.com/watch?v=GDAzsZLvfPw). This is more or less proven in many behavioral experiments (Too lazy right now to source them) when people will do the obviously wrong thing based on internal logics.
Because of this, we can look to the irrational beliefs of people that influence the schools. Namely, social and physical proof. And the social proof has a major benefit: it's less likely that someone is bullshiting credentials if they have an actual degree to back it. http://xkcd.com/451/
Still, it will be a while before irrational opinions catch up (or are leapfrogged by locations like Ghana that embrace the new medium) and that there are ways to catch cheaters.
Here is some writing feedback for you: this style of discourse isn't wanted here.
The vast majority of STEM undergraduate courses fit very well into the format and are very arguably better suited to computer delivery.
is a fine word substitution. but the problem with your observation (assuming it is correct), is that nobody is paying money for commodity education. people are lining up to pay money (in the hope that) a "peer reviewed" credential will set them apart.
I know someone who got most of an undergrad degree online, in a substantive field, by necessity -- she was a teenage mom working a job and caring for an infant in a rural town. Her mom told me the hardest things about online college for her daughter were the loneliness and crushing boredom. It's great that she had that opportunity. I'm glad she's tougher and more focused then I'll ever be.
The atmosphere at a small college is where I was able to keep up with my coursework. It got me out of my hometown. I got over my crippling shyness, and learned to interact with people in a civilized way. Graded assignments "forced" me to do the work.
In my view one thing we still don't know about online education is how many people can actually get through it. We also still don't genearlly know what an "educated person" will consist of, who sat in their parents basement for 4 years staring at a computer screen.
There is a benefit of being in a classroom environment when learning an advanced subject. You have a direct connection to the professor and other students, which is essential for advanced material. Plus you can easily go and talk to them afterwards for more details. I learnt a lot by going to professors office hours which doesn't work online.
Also, what about labs? Every science program I am aware of has a laboratory component. That cannot be done at home. Plus good luck trying to get universities to agree on what labs to do. Each university has different types of researchers who focus on certain areas, which is reflected in the labs selected. My former university, for example, had no particle physics people so any lab related to particle physics would be misplaced. Instead they had a lot of quantum people so the labs were more geared towards that. I know of another university that had a lot of particle physics people and their labs reflected this.
Another thought is research. How would undergraduates get exposed to research in your system?
Thinking about it more, having TA'd first year calculus, you need to be able to interact in real time with students. I think online resources are great and I recommended them to my students, but they cannot replace face to face interaction. This is something online technologies cannot replicate, so I say it is a bad idea to even do it for basic things like calculus, which is very fundamental to the sciences.
Personally, I think it is a noble goal, but unlikely. Instead, I think the universities should integrate online stuff in addition to in classroom lectures.
EDIT: I forgot to mention the university environment. Being surrounded by my peers in the same field really helped. I learnt a lot just by hanging around the various clubs. That absolutely cannot be replicated online.
But it might be useful promote online education more since it increases accessibility and affordability.
> You have a direct connection to the professor and other students, which is essential for advanced material.
You can ask the student to write down their questions and mail the professor. This way, a good professor can be shared by many more students.
>How would undergraduates get exposed to research in your system?
The professors can guide them to where interesting conversation happens on the Internet, and they can get exposed to interesting research there.
>I forgot to mention the university environment. Being surrounded by my peers in the same field really helped.
There is no law that this atmosphere of dense population of smart people should be limited to universities.
By the way, I am agreeing to your central point, which is that online education should be a complement and not a substitute to conventional education.
> The flaws in standardized testing are well-documented at this point.... their predictive powers only forecast a student’s progress as far as the first semester of their freshman year.
Let's be honest. Raise your hand if you needed markedly more than your first semester of freshman year to figure out which of your friends at college were going to cut it in some traditional academic sense.
Not saying that's the only sense, but as far as the college should be able to measure, academics for 4(+) years are the only fair initial barometer. Of course, academics can't say anything about who can write a 7+ figure donation to the college 30 years after graduation...
If you asked high-schoolers (not necessarily even seniors) which of their peers would do the best in college, do you think they would give useful answers?
Test scores will always always always always always always always always always always reflect socioeconomic disparities, because there is and will always always always always always always always always always always be socioeconomic difference in student abilities. It is insane to say standardized tests "punish disadvantaged students and minorities". Tests are tests, they do not and should not care about students' socioeconomic background. Whether universities should account for socioeconomic status of applicants or not, that's their consideration, not test makers' consideration.
Of course standardized tests do not measure all aspects of a student, but they never claim they do. This critique is also silly. The problem is, it is extremely difficult to measure "traits such as optimism, curiosity, resilience" of a person, especially when you want to use the results in college admission. Current psychological instruments for those traits are not developed for that purpose. No serious psychologist will suggest colleges to use those psychological instruments in college admission.
Do no blame standardized testing for those silly reasons!
A test can be written with a bias that favors or punishes any group. I don't think its intentional, as it is quite difficult to create them without these issues.
A simple example to show the point:
Many people would answer A: A chicken is a bird and everything else is a mammal.But if you come from an agricultural background, in addition examining the animals based on their Biological Kingdom, you might be tempted to say B, Moose: It's the only animal that has no role on a farm. B is not an illogical answer from that perspective, but within the context of the test it's likely wrong.
When people say a test has a socioeconomic bias, they're saying a question can become easier or harder depending on a student's economic background. The question above is easy for most of us, but would probably be solved slower by anyone who grew up on a farm.
That's not their argument.
However, actually the issue you mentioned has long been noticed by psychologists / test makers. Experts in SAT / ACT know a lot about these things. Please refer to Differential Item Functioning https://en.wikipedia.org/wiki/Differential_item_functioning
Experts at ACT maybe. But didn't the SAT recently announce it's re-tooling much of its test to address these issues?
That'd be an easy start.
Why should character or "heart" matter? What if you're a smart kid who is a real asshole? Or a dumb kid with a good sales pitch? Or you didn't participate in extra-circular activities for whatever reason? Why should some guy wading through 10,000 applications be making some evaluation and judgement of my character?
They should drop the pretense of some sort of fair selection process and just do a pure lottery. If you can't read and get into Harvard, you'll wash out quickly.
The issue really morphs from "How do I select the best students for my school?" to "What sort of arbitrary metrics can I use to filter down this list of already amazing students?" If I had to guess however, I'd say these schools, under the guise of character, are trying to select students that have the highest potential to give back to the university. "Oh this student does part-time acting? Well we will accept her on the chance she becomes the next Natalie Portman. This guy has a best selling app? Better put him on the list of potential Zuckerbergs."
I'm not an admissions officer (not even close), but that would be my best guess on what someone an admissions officer would do faced with the problem of having so many amazing students to fill their undergrad class 10 times over. At some point it no longer makes sense to filter on raw academic ability (see also, Google's "GPA does not matter"), and a purely random lottery would 1.) never actually be a random lottery (thank you alumni) and 2.) may massively favor one group of people (if I have to choose 2000 our of 33,000, and 60% of that 33,000 are asian from a 150,000k+ income family, you run the risk of homogenous class - if you have to ask whats wrong with this, consider this, would you like to work at a job that hired exclusively white men (or whatever race/gender you are)).
Another way to look at it would be this way: if you were given the task to hire a new team of software engineers, and once the resume's came in they all had, 3.9+ college GPAs, Ph.ds in CS, and multiple years of experience on your tech stack of choice, would you just "do a pure lottery" to see who works for your company? Chances are you might end up filtering them on the most circumstantial of metrics (i.e. "This guy was a Yankees fan, I don't think we would get along very well")
A college does a disservice to students to admit them when they will wash out.
My experience as a student at Caltech in the 70's was they did a pretty good job with the admissions. They selected for smart & motivated & interested in tech. As a result, living in the dorms full of this kind of people, there was never a dull moment.
I only knew a couple of students who clearly didn't belong there, and they soon left.
He was completely overwhelmed and dropped out his first year -- he just wasn't prepared for the material, and the remedial help offered by the college wasn't very useful to someone who wanted to be in a STEM track. He dropped out, joined the Army, and flourished... we're Facebook friends, and he has an MS in CS now.
Thus they need to see beyond the present circumstances of each applicant and discern which will be most likely to excel, taking into account the impact of their 4-year educational process.
Some students from great backgrounds will fail to keep up in college. While some others who come from disadvantaged backgrounds are able to "catch up" and pass others in achievement by the end of college.
How can they predict those outcomes in advance? That's the question that the SAT and ACT were created to answer, and now we're finding that they don't actually answer it better than alternatives. In part because it is becoming clear that there is more to success in life than how well a person solves word problems.
From this article:
> The University of California, Berkeley1 economist Jesse M. Rothstein has found that the combination of a student’s high school grades and demographic information predicted first-year grades in college about as well as her high school grades and SAT scores do.
> DePaul is implementing their own tests for non-cognitive skills, with a series of essay questions. For the entering class of 2012, about 10 percent of applicants (or about 5 percent of the freshman class) chose not to send ACT or SAT scores. Instead they completed four short-answer questions, designed to measure their leadership skills and their ability to meet long-term goals. Systematically scoring the responses to those questions, DePaul reported that the freshman-to-sophomore retention rate was almost identical for those who submitted standardized test scores (85 percent) and those who did not (84 percent).
There is special treatment for people who play sport, whose parents went to the same college (especially if they give money) and for those who have the right sort of hobbies.
If that was not bad enough, there are different rules for different ethnic groups.
It really is quite extraordinary.