Top Colleges Discriminating Against Asian Students:Thoughts?

[Alyrium Denryle]

If you got an A in organic chemistry and can't write code for shit though, applying for a job in IT is probably not the best idea regardless.

CS degrees dont require anything like Organic Chemistry. My BS in biology required Organic Chemistry. A CS degree might require a few general science elective courses. One might take introductory chem or physics, maybe biology. But no more than a class or two. Just one of the introductory survey courses.

Coming from someone who just defended his dissertation in biology, there is immense value in the liberal arts curriculum in US universities. They dont just let you "find yourself", that is what flitting around for a few years and taking 6 years to graduate is for. It helps make you a roundly educated person. I spent my time in those courses learning ethics (General, Biomedical, Environmental) and formal logic (oh yeah, truth tables baby), psychology (given that I study behavior, good choice), comparative religion, and womens studies. All of which were extremely valuable. I am a better person for taking those courses.

What else would you do if you majored in computer science but wasted your time getting an A in chemistry in order to inflate your GPA? If it were a four unit class, it would have a larger impact on your GPA than the presumably 3 unit programming class.

If you think organic chemistry is something to pad your GPA with, you are either insane, or a complete fucking moron.

[Channel72]

I'm glad the corporatist point of view is being so well-represented in this discussion. We should abandon the liberal arts university, cornerstone of knowledge and learning since essentially the Enlightenment forward, because it purportedly makes hiring more difficult. Hint: there's a reason they're called universities, not technical schools.

I'm sorry you don't like reality, but what you call the "corporatist point of view" basically translates to "what corporations want from applicants" - i.e. what applicants need to do to be successful and land jobs. You can dismiss that because you find it crude or unenlightened, but that's what it is. Ultimately, a technology company is interested in solving problems that require specific, technical skills. If candidates don't have those skills, we just can't hire them. And the current curricula coming out of US colleges provides us with a lot of candidates with patchy, scattered knowledge of CS. This is because (1) colleges require them to take lots of non-CS courses, and (2) the CS courses they take are so fucking scatter-brained - they don't comprise a coherent curriculum that starts at the basics and builds. Rather they're just a bunch of scattered topics, like Introductory courses, Databases, machine learning, etc. all of which aren't really connected very well by a single comprehensive curriculum that builds on previous classes. So we get candidates that have giant gaping holes in their knowledge, and so we just can't hire them.

As for liberal arts, I didn't fucking say that there's no value in learning anything outside your field - although the case could be made that this sort of exposure to a wide variety of knowledge is something that should mostly happen prior to the college level. Anyway, your complaint doesn't amount to anything useful, and your comment about the etymological origin of the word "University" is also pointless. The fact is, after graduation, people need to actually, you know... get jobs? And what American Universities are doing right now isn't working very well, and it's definitely not serving the best interests of students - because ultimately, students are going to have to deal with the "evil corporatist viewpoint" when they get out in the real world, unless they want to stay in academia. And as an evil corporatist, I'm just not going to hire someone that doesn't excel specifically in their chosen field.

CS degrees dont require anything like Organic Chemistry.

Yeah, I know - I used it as an example because anecdotally I find that a lot of CS and EE majors for some reason also take organic chemistry and physics to satisfy DEC science requirements.

[Gaidin]

I'm glad the corporatist point of view is being so well-represented in this discussion. We should abandon the liberal arts university, cornerstone of knowledge and learning since essentially the Enlightenment forward, because it purportedly makes hiring more difficult. Hint: there's a reason they're called universities, not technical schools.

I'm sorry you don't like reality, but what you call the "corporatist point of view" basically translates to "what corporations want from applicants" - i.e. what applicants need to do to be successful and land jobs. You can dismiss that because you find it crude or unenlightened, but that's what it is. Ultimately, a technology company is interested in solving problems that require specific, technical skills. If candidates don't have those skills, we just can't hire them. And the current curricula coming out of US colleges provides us with a lot of candidates with patchy, scattered knowledge of CS. This is because (1) colleges require them to take lots of non-CS courses, and (2) the CS courses they take are so fucking scatter-brained - they don't comprise a coherent curriculum that starts at the basics and builds. Rather they're just a bunch of scattered topics, like Introductory courses, Databases, machine learning, etc. all of which aren't really connected very well by a single comprehensive curriculum that builds on previous classes. So we get candidates that have giant gaping holes in their knowledge, and so we just can't hire them.

That's because it's a computer science program. Literally. There's plenty of ways to get what the corporations want. But the electrical engineering department at my college started facepalming when they started teaching python instead of C in the introductory programming course but still insisted on teaching it to the EE students as well. Seriously? The corporations are loving some of these changes, at least at my old university, but its screwing over the other departments when they're having to practically teach lower level programming to their students because of the changes being made but the Engineering College politics being played. It's insane. There is a literal difference between Computer Science and Programming sometimes. The corporations need to pick their damn poison.

[Channel72]

That's because it's a computer science program. Literally. There's plenty of ways to get what the corporations want. But the electrical engineering department at my college started facepalming when they started teaching python instead of C in the introductory programming course but still insisted on teaching it to the EE students as well. Seriously? The corporations are loving some of these changes, at least at my old university, but its screwing over the other departments when they're having to practically teach lower level programming to their students because of the changes being made but the Engineering College politics being played.

Um... I can tell you with absolute certainty that many corporations (at least the one I work for) would much rather that students are taught C over Python. If an applicant can't code pointer arithmetic, or doesn't understand the difference between the stack and the heap, they are not getting hired. Hell, I'd prefer if students were taught assembly from the beginning, and not just MIPS, so they ultimately understand all the shit the CPU is actually doing when they write Python one-liners.

[Gaidin]

Um... I can tell you with absolute certainty that corporations (at least the one I work for) would much rather that students are taught C over Python. If an applicant can't code pointer arithmetic, or doesn't understand the difference between the stack and the heap, they are not getting hired. Hell, I'd prefer if students were taught assembly from the beginning, and not just MIPS, so they ultimately understand all the shit the CPU is actually doing when they write Python one-liners.

That came in the second course. Not the intro course. But by then the student's broken off into EE. That's why the EE professor's were pissed. They had no grounding in what could be said was medium to low level syntax. Teaching pointers and the stack and heap was something they did all the time, but they always had the grounding in the C syntax from that first course. When our CS department decided to go off into experimental lala land, and I don't know if they're still doing it, I just know they spent at least two years doing it while I was getting an EE degree, boy were my professors going batshit.

[Channel72]

Yeah, I can understand. Universities have been dumbing down the CS curriculum for years now - first it was Java, now Python. The problem is that for the "evil corporatist", it's way harder to assess the skill of a candidate who only knows high-level programming concepts. Whereas, if a candidate can say, code a memory allocator in C, it's much more obvious that they know what they're doing.

[Gaidin]

But I think we've sort of gone off from what I meant to say in the first place. Computer Science and, well, Programming can be two different things. Computer Science is often a very theoretical field, forget if you're going for research for any of these companies. The companies that want to take a design of theirs that they know works and tweak it for their customer, update it, or update it heavily, they're more looking for a programmer. The programmer may have a computer science degree, yes. But there that is. The heavy updates that require total overhauls may require actual computer science theory to get it to come together from another angle, but the Computer Science departments seem to be losing something unless the students are willing to stay and do some form of research for a Masters these days. It's, well, sad.

[Ralin]

Undergraduate admissions. We're mostly talking about teenagers. Hobbies are potential ways to show their ability to learn or show initiative or whatever since they generally don't have a much else.

Unless you're not too well off in which case those potential hobby times can be filled with a job or other duties their better off classmates may not have.

In which case they write an essay about how they worked after school to help support their family and overcame adversity by pulling themselves up by their bootstraps or something. The specifics are less important than finding some way to stand out.

How? He doesn't have any hobbies at all. The only thing he does is drink beer and watch football. Honestly, the guy can name any player in Europe, but that is about it.

His research/work credentials, presumably? If he has something better than his mad guitar skills to show he's smart and motivated then he's better off than most of the teenagers we're talking about

[ArmorPierce]

And this is why corporations input on academics should be limited. Corporations want future employees that are trained in very focused areas and systems. They don't want to foot the bill however and if systems become obsolete rendering those skills students trained in useless. .. though luck.

The point of a university education is not only to prepare students for a job in corporations, it's so that they can be productive contributing members of society with a deep and broad enough knowledge base top draw from.

[Channel72]

And this is why corporations input on academics should be limited. Corporations want future employees that are trained in very focused areas and systems. They don't want to foot the bill however and if systems become obsolete rendering those skills students trained in useless. .. though luck.

The cliche of obsoleted systems rendering a programmer's skills useless is false. The reality is that, apart from say, concurrency, computer science hasn't changed in a real fundamental sense since the fucking 60s. Corporations (at least, non-shitty corporations) shouldn't be looking for employees who specifically know how to write code in one particular version of one particular language. They should be looking for employees who understand the fundamental concepts and theory behind computer programming - someone who understands that will have no trouble picking up any new-fangled language.

[Gaidin]

The cliche of obsoleted systems rendering a programmer's skills useless is false. The reality is that, apart from say, concurrency, computer science hasn't changed in a real fundamental sense since the fucking 60s. Corporations (at least, non-shitty corporations) shouldn't be looking for employees who specifically know how to write code in one particular version of one particular language. They should be looking for employees who understand the fundamental concepts and theory behind computer programming - someone who understands that will have no trouble picking up any new-fangled language.

I...err...

There's been at least three generations of languages since the 60's specifically because Computer Science has changed fundamentally. Well, 50's is more accurate when FORTRAN was introduced and the concept of a high level programming language took over. This let the computer take care of non-essential details. In the 1970's the fourth generation was introduced with languages like Perl, favoring a higher level of abstraction of internal hardware details and programmer friendliness. The 1980s brought about the fifth generation and languages designed to make the computer solve a given problem without the programmer. LISP is usually given as the best known example, for how the programmer needs to worry about the problem that needs to be solved and and conditions that need to be met, without worrying about how to implement a routine.

Now, depending on your profession there is still a lot of low level assembly being done. And your language is chosen based on what you're doing. Or should be. More later. Predictions were made, insanely wrong, about how one generation(Fifth) would take the world by storm, and all others would fall off except the necessary assembly. But then, hell, it was discovered that it was useless for what some people were trying to do, but then it was great for what others were trying to do. But saying it hasn't changed? Especially since the 60's?

The truth is that Corporations tend to pick something like Java because Java is Cool!(TM), not really because it's exactly the best thing for their job sometimes. It's design used to be no more than fifty pages, but it's ballooned up go quite a few hundred because the corporations want a God Language that can do everything and the kitchen sink. Some don't like Java, but need to pay bills. Others do. We all have bosses. I tend to hate its guts because it doesn't give me the fine control I grew to like from C/C++. I never had to experience it's full functionality so the ballooning design was not something that effected me. So when I had to use it for two years I did nothing but twitch for eight hours a day. But, as I said, I had bills.

[Channel72]

There's been at least three generations of languages since the 60's specifically because Computer Science has changed fundamentally.

I don't agree that computer science has changed fundamentally in any significant way. Most "new" things like lambdas, closures, OO programming etc., all that has existed since LISP and Smalltalk/Simula. LISP had features in the 50s and 60s that languages like Java are just *now* getting. And the basic data structures and algorithms we use today were mostly invented in the 60s and 70s.

The only fundamentally new thing that has changed in computer science in any meaningful way is concurrency, due to the prevalence of multicore architectures and the way modern CPU cache lines work. And really, we still haven't figured out the best way to do concurrency anyway.

[Thanas]

His research/work credentials, presumably? If he has something better than his mad guitar skills to show he's smart and motivated then he's better off than most of the teenagers we're talking about

What fucking research do you think he did in highschool? And what fucking guitar skills? Guy likes to drink beer and watch football. That is it.

[Ralin]

His research/work credentials, presumably? If he has something better than his mad guitar skills to show he's smart and motivated then he's better off than most of the teenagers we're talking about

What fucking research do you think he did in highschool? And what fucking guitar skills? Guy likes to drink beer and watch football. That is it.

Guitar skills being shorthand for his hypothetical hobbies that he doesn't have

[Starglider]

I don't agree that computer science has changed fundamentally in any significant way.

The entire field of data science was invented. Huge amounts of research on effectively processing 'big data' have been done that wasn't possible prior to the mid 90s when cluster supercomputers and cheap compute grids in general became practical. No this is not the same thing as conventional statistics. Distributed architectures are vastly more elaborate now; the fact that we still use TCP/IP does not mean that a late 60s understanding of networking would be adequate to understand cloud application development. The state of the art in many other subfields has advanced beyond all recognition.

Most "new" things like lambdas, closures, OO programming etc., all that has existed since LISP and Smalltalk/Simula. LISP had features in the 50s and 60s that languages like Java are just *now* getting.

This is how research works, things are invented and prototyped in academic settings, are filtered and refined and trickle out into general use. Meanwhile, academics have moved on from Lisp and Smalltalk to hundreds of newer research languages.

And the basic data structures and algorithms we use today were mostly invented in the 60s and 70s.

By that standard, no scientific or engineering field ever 'fundamentally changes', because by and large basic useful things continue to be basic and useful. By your standard relativistic and quantum physics were not a 'fundamental change' because Newtonian physics is still used most of the time for most applications.

[Channel72]

The entire field of data science was invented. Huge amounts of research on effectively processing 'big data' have been done that wasn't possible prior to the mid 90s when cluster supercomputers and cheap compute grids in general became practical. No this is not the same thing as conventional statistics. Distributed architectures are vastly more elaborate now; the fact that we still use TCP/IP does not mean that a late 60s understanding of networking would be adequate to understand cloud application development. The state of the art in many other subfields has advanced beyond all recognition.

Yeah I mentioned concurrency as a significant exception. And processing big data is basically just an application of shared-nothing concurrency. Yes, MapReduce is an impressive and promising paradigm. And using TCP/IP for all this is actually pretty recent - before they were using Infiniband and other lower-latency protocols for this stuff.

This is how research works, things are invented and prototyped in academic settings, are filtered and refined and trickle out into general use. Meanwhile, academics have moved on from Lisp and Smalltalk to hundreds of newer research languages.

Please name one language that is radically more powerful or expressive than Lisp. I'm not saying Lisp is some kind of be-all end-all of computer languages (I don't even really like it that much) - just that today's languages aren't really that more interesting or productive than what you got with S expressions and macros - and very few languages these days actually let you manipulate ASTs as a first-class construct. Yeah, we've got Haskell with monads, and lots of upcoming languages trying to invent new abstractions for concurrency, but at the end of the day there's very little that's radically different from what we had with Lisp 40 years ago.

And the basic data structures and algorithms we use today were mostly invented in the 60s and 70s.

By that standard, no scientific or engineering field ever 'fundamentally changes', because by and large basic useful things continue to be basic and useful. By your standard relativistic and quantum physics were not a 'fundamental change' because Newtonian physics is still used most of the time for most applications.

There has been no revolution anywhere near the scale or significance as the transition from Newtonian to Einsteinien or Quantum physics in computer science. Perhaps qbits will change that, but I'm still bit shifting it old school like they did in the 70s. Again, as I've said many times now, the only real exception to this is in the field of concurrency - what with our fancy lockfree compare and swap hash tables these days.

Although, perhaps another area besides concurrency that has shown significant theoretical advancement is VM and JIT technology - although even that is a lot older than most people think.

[Exonerate]

I'm not sure what the point of this discrimination would be, if it exists. Surely it would be in the best interest of colleges to be able to boast a large number of highly-graded graduates. This almost sounds like reverse Affirmative Action to me.

In practical terms, there are a limited number of slots available for new students. Certain demographics (Asians and Jews especially) produce a glut of qualified candidates, and would utterly dominate incoming classes if selection was based on blind merit. Conversely, you'd have very few Black or Latino students. If you want to bring up those groups, someone else has to lose out.

So why the fuck is it that the spots are being taken from Asian Americans, who have been historically disadvantaged and continue to face discrimination today, instead of Whites who face no such barriers?

[Thanas]

Are Asian americans facing discrimination today? Their average income seems to surpass that of white people.

[AniThyng]

Are Asian americans facing discrimination today? Their average income seems to surpass that of white people.

AFAIK this only holds true for certain subsets of "asian american", and disregards all others.

[GuppyShark]

I remain grateful for the Australian UAI/ATAR system. Imperfect, but better than the mess I've seen whenever this comes up.

I think it differs from state to state, but I know that I (on paper at least) benefited from affirmative action as a white male Australian.

Is that crazy? Let me explain.

When I completed my high school years I received bonus points on my final TER (Tertiary Entrance Rank - I assume some sort of equivalent to the SAT) due to the following:

* Lived in an area with low socioeconomic status.
* Attended a poorly performing public high school.
* Neither of my parents attended a tertiary institution.

Unfortunately, I rewarded the country for this positive discrimination by being inconsistent and ultimately dropping out when the government cut off my student income.

I'm smart, but I had never learned any sort of work ethic coasting on raw talent my entire schooling, so having to actually work to accomplish things was foreign to me. I was a lazy shit and I often just couldn't be bothered going to class.

It's only now, ten years later, that I have gone back to study properly and I am kicking butt and scoring grades. Maybe if I'd not gotten those bonus points I'd have gotten my shit together sooner.

[Borgholio]

Are Asian americans facing discrimination today? Their average income seems to surpass that of white people.

Not from what I have personally seen. My wife's family is Japanese and they are quite well off. In fact, THEY are the ones doing most of the discriminating (her grandmother HATES Koreans).

[Channel72]

I mean, any visible minority is going to be the target of some discrimination. It's just that currently, Asians experience less discrimination on average than say, blacks, because Asians tend to be successful people and so they're not associated with crime and poverty the way blacks/latinos are.

[Borgholio]

It's just that currently, Asians experience less discrimination on average than say, blacks, because Asians tend to be successful people and so they're not associated with crime and poverty the way blacks/latinos are.

This is true. Cities in Orange County like Irvine, Lake Forest, Mission Viejo, and surrounding areas have very high wealthy Asian populations. In some communities, even whites are in the minority. And due to the overall wealth of these places, they tend to be very well manicured, safe, clean, and retain high property values. Hard to complain about any of that...

[Exonerate]

Are Asian americans facing discrimination today? Their average income seems to surpass that of white people.

It is, but it's also a highly deceptive statistic. For one, it's an incredibly large grouping that encompasses everything from Indian Americans, who are the highest income ethnic group in the US, to Hmong Americans who have a poverty rate greater than one in three. Income distribution is is quite bimodal; highly educated immigrants from East Asia and India are doing quite well, but those who immigrated by something like lottery or as refugees are not.

The average household income is largely buoyed by high educational attainment, concentration in high cost of living areas, and larger household size. When comparing with equalized education, Asian Americans (along with other minorities) get paid less. There's many other studies about ethnic names impacting employer callback rates, East Asian/Indian names receiving less attention from faculty members, under-representation in corporate leadership adjusted for the proportion of the workforce, etc. All evidence I've seen is that AA success is despite discrimination. Returning closer to the original topic, there's something perverse about citing higher AA incomes, mostly due to higher education, as a reason for disadvantaging them in education.

[ArmorPierce]

Are Asian americans facing discrimination today? Their average income seems to surpass that of white people.

Asians do face discrimination, difference being that as a group they are fairly educated and wealthy and support each other. Also not mentioning that asian immigrants often comprise the already relatively well off of their respective societies.