Making universities obsolete
matt-welsh.blogspot.com
matt-welsh.blogspot.com
If you get down to it, probably the biggest function of a university is social: if you majored in an engineering discipline, think about your industry connections. How many of them were classmates in undergrad? Elite universities have served as social hubs for the upper class for centuries and the networking function of college and university is invaluable. I'm not using this as an argument for keeping the existing university structure, but if the author really wants to make universities obsolete, he needs to think about how one might facilitate this and other functions. (I haven't even touched the subject of research, which many professors at R1 universities think of as their primary purpose, not teaching. "Teaching" is an unpleasant job you offload on adjuncts and TAs.)
We've had institutions that function like the ones that Matt Welsh envisions for thousands of years. They're called libraries.
[1] http://www.insidehighered.com/news/2011/01/18/study_finds_la...
edit: added citation ;)
I had a commuter college experience and I bet unless you have the social skills, fortitude and setup (such as living in dorms) to take advantage of the social benefits of college, you wont get many social benefits out of it.
For me college was primarily personal social growth, professional credentialing, and educational support network. I think the last two of these can be offloaded to the internet for any abstract subject (not anything that requires creating a physical artifact instead of a digital representation).
Think about that.
I had a calculus student, begging me to pass him. He'd failed my course, he'd failed the same course twice previously, and if he failed a third time he would never be allowed into the engineering school.
It gives me no pleasure to crush dreams or to fail students, but do you want this guy, say, designing bridges?
You in particular may be incorruptible, but if you were untenured and the student had a lawyer I'd bet on him "passing" your class.
The counterargument: If the student can pay cash and wants to take an engineering class why not indulge him? Shall we deny all those without math skill the opportunity to even see, say, robotics, just because we're sure they'll never be qualified to practice it at a pro level? Why is the purpose of calculus class to sort people into silos rather than simply... Teach calculus? And why can't a student of means be allowed to fail the classes of his choice?
One answer is that engineering education is so expensive to deliver, and demand so high, that we can't afford to lavish it on hobbyists. (If you've had the pain of trying to teach a class containing students who are over their heads, you can feel this argument in your bones, because the resource being wasted is your time, for which you are not well compensated.) But... Maybe we can fix that? Productivity-boosting technology has many uses.
Realistically it'll be at least 5 years from calculus class until he'll even be the most junior assistant to someone designing a actual bridge and from there probably at least 10 years until he'll be in charge of a bridge design. That's plenty of time of time to catch up. So why not just let him study calculus until he gets good enough to move on? Remove the three fails and you're out rule and the whole problem of students 'needing' to pass that particular course goes away.
If he really doesn't have the necessary talent then he'll never pass all the courses necessary to become an engineer and will never design bridges, but that should be up to him to discover, and if he wants to spend 12 years discovering that then that should very much be his prerogative.
Essentially, under this model everyone is expected to eventually master the material--so in many ways this is harsher than traditional grading. It would be equivalent to him taking calculus until he got an A--not just taking it until he managed to pass.
[Edit: after reading _delirium's correction below, I made a couple small changes. I've got a cold, and I'm afraid I haven't made my point well.]
I'm perfectly fine if students aren't graded (or if grades are a strictly private matter between the student and teacher). Many teachers think so too, so-called "radical" teachers, but of course they unfortunately have bosses... a few teachers get away with it, particularly in more democratic schools where students are more in control, but not enough do.
Personally, I learn all sorts of subjects, and the notion of someone conventionally grading me... would degrade the whole experience.
In fact that's mostly why the "radical" teachers' bosses force them to give grades; UC Santa Cruz's 1950s-80s experiment in giving narrative evaluations instead of grades was mostly ended because employers demanded grades. If it hadn't been for outside pressure on the administration, mainly from employers and students worried about employment prospects, much of the faculty would've liked to continue the previous approach (though support was less strong among engineers and computer scientists, who for various reasons were more in favor of a letter-grading model, and saw the narrative-evaluation model as too "soft").
Beginning with his "failure #2";
>The common argument is that we need grades in order to differentiate the "good" from the "bad" students.
No, no its not. Its to show that you've learned the material. He goes on to say:
> Presumably the idea is that if you can't get through a course in the 12-to-13 week semester then you deserve to fail, regardless of whatever is going on in your life and whether you could have learned everything over a longer time span, or with more help, or whatever.
12-13 weeks is plenty long to learn the material for a course. A student is at a college to learn. It should be their top priority to learn the material during the course.
As for "failure #3", lectures have been used for hundreds of years because they work. You gain insight from someone very knowledgeable in a field that you might not get at first without it being pointed out to you.
He also seems to miss the fact that for science courses, you need labs. You cannot do chemistry labs online. You won't be able to do field work in biology by watching a video.
A somewhat agree with point #1, but more so to the fact about how expensive college is. That needs to be changed.
You lose the overall atmosphere of a learning environment through online courses, and that is not something I'd want to give up.
The course has presumably been designed so that a motivated and intelligent student can understand it in this time.
But what happens when life interferes with a student's plans? I can think of lots of reasons why a student may be distracted by legitimate life events in this period of time, e.g. illness, or a family wedding/birth/death etc.
Online courses do give you the flexibility to work around things like this. They also allow people with full time jobs to access the material at a pace which suits them.
But it doesn't do that. It merely shows you were able to perform well in a contrived testing environment that has little to do with how you'd actually approach problems in the "real world" (that is, everywhere but a test).
>12-13 weeks is plenty long to learn the material for a course. A student is at a college to learn. It should be their top priority to learn the material during the course.
I think you're missing the author's point here. Yes, a student is at college/university to learn. However, the job of the institution should be to teach, not to weed out students, and that's where the true failure is taking place. And again, nobody is always in full control of their life. Shit happens, as they say, and just two weeks of illness (or really any unforeseen circumstances) can cost you a whole semester. Some students need to juggle studies, work and getting enough sleep while pretty much having no life because of the horrendous costs of attending higher education (which is a whole other problem entirely).
>As for "failure #3", lectures have been used for hundreds of years because they work.
That's an appeal to tradition. They haven't been used because they work, but because there was no viable alternative. But now, we have such an alternative - video lectures and online courses.
>He also seems to miss the fact that for science courses, you need labs. You cannot do chemistry labs online. You won't be able to do field work in biology by watching a video.
That's true indeed. I don't think it's productive to attack universities as a whole, either. They serve a lot more purpose than just holding lectures. You can meet people in the same or adjacent fields as you, you can form connections, work in research (and maybe even stay there after graduation) and many other things. But that doesn't change the point of the article, which is that lectures as a teaching format suck when compared to the alternative.
>You lose the overall atmosphere of a learning environment through online courses, and that is not something I'd want to give up.
You'll have to excuse me for being a bit snarky now, but if by "atmosphere of a learning environment" you mean cramped lecture rooms, long and drawn-out lectures without much (if any) interaction with the lecturer with the overall process often being more of a hassle than anything else for all involved parties, then no, I don't think that's something worth clinging to because we've done it this way for the last thousand years.
I think that at some level an A or B indicates something related to knowledge and capabilities. My personal experience is that tests are an effective method of determining who should proceed and who should not. It's not fair to other students and teachers if a student takes Calc II without knowing Calc I. Some mechanism is needed to weed out those who shouldn't move on.
Lectures do work. They have been working for hundreds of years. This isn't an appeal to tradition. This is a statement about evidence that they work. Lectures work so effectively that people want to put them online so anyone can watch them.
But guess what happens with students in online classes? My anecdotal experience is that they won't watch the lectures. They won't sit through a 20 minute lecture on how to solve quadratic equations. (Except a core group of good students.) The passing rate of students in online math courses at my college is lower than for the face-to-face courses.
But that's not a counterargument. Your simply restating your own point, which I disagree with. It's pretty obvious that someone who would score well in an artificial testing environment A would score well in such an environment B if B is a followup of A. That doesn't disprove the point that grades are a fundamentally flawed concept in education.
>My personal experience is that tests are an effective method of determining who should proceed and who should not.
Again, based on what? Conformance to a completely artificial testing environment? Again, tests have nothing to do with reality and are a flawed concept. If anything, they show if you can work under an extreme time constraint with limited resources available. Are we really basing success or failure in all of our education on this limited and arguably pretty unimportant skill-set?
>Lectures do work. They have been working for hundreds of years. This isn't an appeal to tradition. This is a statement about evidence that they work.
You're still ignoring the context, and that is that we have a better alternative to physical lectures now. Let me present an analogy:
"Books work. They have been working for hundreds of years. This is a factual statement about evidence that they work."
While principally true, that doesn't change the fact that printed letters on dead trees are no longer a viable model of distributing knowledge. It's an appeal to tradition because you ignore the context at hand. Your only argument is "we've always done it that way, therefore it must be right, why change it?".
>Lectures work so effectively that people want to put them online so anyone can watch them.
What are you trying to argue with this? Yes, online lectures are better than traditional lectures. To extend on the book analogy, while it's fundamentally still the same as reading a physical book (you're parsing words), an electronic representation of text is in every way superior to a printed book. The same holds true for online lectures.
>They won't sit through a 20 minute lecture on how to solve quadratic equations. (Except a core group of good students.)
Let me present you with my anecdotal experience as a student myself: Next to no one but a core group of good students will sit through any physical lecture unless it's absolutely mandatory to attend, either. As I've put it in another comment, if you're not at the university because you want to learn, what the fuck are you doing at the university?
The main problem with his argument is not that the "testing environment" is flawed. He is talking about learning as if it were a race, and suggesting that only the hares should be allowed to run it.
Often those who do extremely well in math classes have a pretty good head start, due to circumstances that are not a result of their abilities, or even their own choices. And, of course, such students will do better later; they can maintain a more comfortable pace than those who are racing behind b/c they are trying to make up for their late start.
It also disfavors those who are distracted not by parties and social scenes, but distracted from the main course by branches of the subject under study which catch their interest. That is, it favors the compliant student who is after good grades, and disfavors those who desire knowledge, skill and understanding for their own sake, or for any purpose not foreseen by those who designed the "course".
The great benefit to these on-line study programs is that it serves the interests of all of these kinds of students. Neither necessity, influence, affluence nor preference need induce the educators to favor one over the others.
I can give them a test and when they get an 'F' on it there is no argument. The test is a good measure of who knows the subject and who doesn't. For the most part. It's the most effective tool that I know of.
The absolutely mandatory part is the crux of the issue with regard to face-to-face classes. It forces students to meet at a set time and work on a given topic. In the absence of this forcing mechanism human nature takes over and most students fail miserably at the time management aspect of online classes. This is all anecdotal though. I don't have any research to back this up.
But does this have to be the case? In the current educational environment, I would agree that a number of students benefit from being made to attend class. The question is if this is just a result of human nature, or something else. Where that something else could be having students go to college when they are not ready for it, or high schools and earlier killing students intrinsic desire to learn, or some other reason.
Really? I give non-compulsory lectures, and get > 80% attendance rates. I also encourage audience participation, and get it. I think my lectures would be less useful via the internet. Not useless of course, but not as useful.
>No, no its not. Its to show that you've learned the material.
I disagree with both of you. It's to motivate the student to learn.
In my opinion, if your not at the university to learn first and foremost, then what the fuck are you doing at the university in the first place?
People, by and large, need artificial encouragement to help them do what they want to do in the first place, but lack the willpower or procrastinate too much. People subscribe to classes so they can learn in a formal setting material they could perfectly well study on their own before, but didn't. People take out gym membership so they have no excuse not to go. People ask relatives and friends to help force them prepare for something important. You may be a champion of iron will (though I doubt it), but most people aren't, and they're still quite capable of learning nevertheless.
In the recent ML/AI course offerings from Stanford, the exercises were a joke - they were painfully easy. But if they didn't have graded exercises with deadlines, probably five times fewer people would have finished the course. In fact, similar courses with entire lectures and homework have been posted on the web for years, with little engagement, while this new offering, with a formal setting that graded you on exercises, has seen tremendous success.
Your rhetorical question is exceedingly naive.
The contrived setting of the classroom is much better than an online setting. This may change as technology gets better. Right now, online is not a better format than face-to-face. Online might be the most cost effective format but it isn't the pedagogical optimum.
I don't think you've made a compelling argument on that score. It is precisely because life is complicated that lower cost and more flexible method might be considered optimum.
There are however a lot of students who see academics as something of a game and measure their achievement by score, the fact that they may learn a few things along the way is secondary to their desire to be able to say "hey , look at me I'm a high achiever please pay me lots of money". These people would probably be lost without a grading system of some kind and a lot of them are actually very smart people who probably should be at university.
And while this more of a philosophical point than anything, I think that this form of competitiveness ("look, I'm better than you") is a major problem of today's society. Education should teach to work together, not against each other.
Experiments in reducing the grade-centricness of universities have mostly foundered on employer objections. UC Santa Cruz for a while was famous for giving narrative evaluations instead of grades. Clases were formally graded pass/fail to indicate whether you met the minimum requirements to move on, and then in either case the professor wrote a one-page or so explanation of what you did well and poorly in the class, some advice, etc. In a sense the evaluation was still quite rigorous, just not boiled down to a single letter.
Companies hated this because they wanted to know which students had the 4.0 GPAs, and which had the 2.0 GPAs: a string of passes plus a stack of narrative reports didn't fit their evaluation model. So UCSC moved to allowing students who wanted to receive a grade along with the evaluation to opt into an A/B/C/D/F grading model alongside the narrative evaluation; eventually they made that mandatory; and finally just dropped the narrative evaluations.
You would probably just have recruiters who would grade candidates on the basis of how many times the word "excellent" appeared on the report or similar, no doubt startups would pop up offering machine learning type solutions to these problems as well.
Not to mention it would be easier to corrupt the system as there would be more plausible deniability if a professor simply "forgot" to include some negative factors on the report.
Much easier just to filter numerically based on grade. Also most of the smart people who do enjoy the learning as an end in itself should know the material well enough to do well in an exam anyway and just view the process of revising for the exam and tweaking their coursework for good marks as "playing the game".
I'm not convinced that academics are better qualified to predict the needs of society regards knowledge and education any more than the free market is.
Regards to helping students evaluate themselves , there is no reason that a professor cannot provide feedback in addition to assigning grades indeed I'm sure many would be happy to and it's more likely a case that many students are simply happy to get their grade and move on rather than evaluate themselves.
On the contrary, I'm pretty sure the free market is the last thing qualified for that. It will always be a race to the bottom, and that's not something education should be.
Would Stackoverflow.com have the participation it has if it didn't have badges, numbers, and karma? I think your perspective ignores a great deal of research on games and human behavior.
And that's exactly the problem with the entire system. We need other ways to give incentives and motivation. The point is that plain numbered grades suck at giving motivation, they suck at actually judging the performance and/or understanding of a student and they suck at being anything but a completely arbitrary method of weeding out people who are unable to conform to the contrived requirements of an artificial testing environments, whether they are actually intelligent or not (and being a slow learner doesn't mean someone is dumb, another thing this system sucks at).
A few points.
Firstly I entirely agree that having access to a video of a lecture is much better in many ways than attending live. I have a terrible attention span and I remember sitting in classes paying attention then having my mind slowly drift away for just a few seconds, by the time I'd snapped by I'd sometimes completely lose the context of the example the professor was explaining and it was really hard to mentally catch up, sometimes I couldn't and I'd just end up with a set of half comprehensible notes to revise from.
Of course your not going to raise your hand and say "sorry I wasn't paying attention could you please repeat", but a video does not judge you like that. Pause it or rewind and get yourself a coffee , look up something you didn't quite understand on wikipedia for an alternative explanation then return to the video.
Of course there is no reason for lectures to be an hour long either, simply break it into 2 half house sessions with space for questions in between which could be done via a chatroom which the lecturer answers when they return.
I'm sure I would have got much better grades at university had I had access to all of the lectures via video, whenever I want to brush up any CS I just watch youtube videos from MIT , Berkeley etc and I learn so much more per hour than I did from my "traditional" education.
Another point.
This part raised a question to me:
"Presumably the idea is that if you can't get through a course in the 12-to-13 week semester then you deserve to fail, regardless of whatever is going on in your life and whether you could have learned everything over a longer time span, or with more help, or whatever. "
How much of your university degree is there to prove what you know and how much is there to prove your capacity for learning?
I've often struggled with some concepts that others have found easy and sometimes vice versa, but I can get a bloodymindedness that I will get to grips with something even if it takes me 10x the time.
How much should learning speed be represented vs sheer determination? Sometimes in the real world you will run into situations where you need to learn something very quickly because there simply isn't the time.
Should someone who takes 5 years to get a degree be considered lesser than someone else who does the same degree in 2? Assuming they get the same marks awarded.
To be honest, I think they should (unless they are taking the degree part time).
If it takes you over twice as long to learn something, that means we can probably expect you to take twice as long to learn new things?
In practice, I have never seen a student who took 5 years (and therefore repeated 2 years on a 3 year degree) who ended up getting a good final grade (although I am sure there exist exceptions somewhere).
1) Should speed be in the grade?
2) Should speed be considered when measuring a graduate?
3) Is a faster graduate better?
My answer would be no, yes, and it depends. I want grades to be on coursework, not character. I think speed can be inferred (from enrollment dates, additional minors/majors/projects/clubs/association/etc.). I don't think speed is everything (determination means a lot), but I don't ignore it either.
They have managed to adapt many times before. So while their current methods might be slightly outdated, I doubt they will be becoming obsolete anytime soon, they will merely change their mode of operation to include new technologies.
[edit] More seriously though I think 'universities' as we know them are important for particular things. What is likely to take off in the future however is good independent accreditation for private-study. The most important failing of universities at present is that you can hardly ever fail...
But this approach creates a rift between teacher and students. Nobody cares about actual learning, students see the teacher only as a source for getting exam answers and teachers only see this glamorous profession as a chore to get paid.
This approach however, loses the fat and makes the education more leaner IMHO. No money and no diploma is involved and you are there because you really really want to see/learn/do whatever the teacher is trying to convey. As for the teacher this makes his/her work much less tiring, you record it once and can change it after you got exposed to biomass. You can change stuff in your teaching and get instant feedback. If this is not amazing for the both of the elements involved in the teaching (teacher & student) I do not know what is.
The Linear Algebra lecture, on the other hand, is what lectures should be. Dr. Hong (who I am naming because he is awesome) is engaged with the class. He answers everyone's questions, and will call you out if you aren't paying attention. And more importantly, he always thoroughly explains the process and makes sure you understand it (and for all the matrix operations we have learned so far, he also makes sure you understand how to generalize it to larger matrices) before you walk out of the classroom.
Then, the homework is optional (though he does joke about killing you if you don't do it.) The goal of the homework in this case isn't to help you understand the material...that's what the lectures are for. He gives homework so students can become efficient at the process after they understand it.
So I don't think it's the lecture/homework model that's really broken here. A disinterested instructor will kill the course no matter whether you're lecturing in a classroom and doing problems at home, or lecturing on the Internet and doing problems during class. What's broken is that lecturers don't focus enough on the process and the concepts, instead running through multiple examples and hoping the students pick something up along the way.
(At least, for mathematics. I'm not sure how well this teaching style would translate to other disciplines such as history, composition, or even biology.)
Matt could have expanded his risks a bit more, at least to point out that no system will ever be perfect, and, ultimately, every student is different. Completely razing the current system may not be a great idea, because some students need the classroom setting (of course, as Matt points out, many of these students are left behind anyway because the playing field is often tilted toward the strongest students). That said, I suspect many other students would do very well in a more open setting (or, ideally, some combination of the two).
If standardized testing and related work are good enough for college admissions, would something similar be sufficient for more specific course work?
With that said, most people do place a lot of value in the connections they made during their time in school. I'm not sure those connections can be built by online institutions. That may have a dramatic effect on opportunities one has later in life.
On an unrelated note, I had a teacher in high school several years ago that figured universities were already obsolete at that time. It's interesting it is a topic still in debate. Perhaps he was a man ahead of his time; he teaching style certainly was.
Outside of simply procuring a job, they matter even more. As an example, the company I work at offers professional education services on CUDA and OpenCL (with other HPC architectures in the works). One of the guys who teaches it did his masters in EE at Stanford. I don't think the impact of that can be overstated when selling the service to companies.
Many parents are very conservative in their approach to education, and the parents see the University path as the most certain way forward. Since the parents are largely the motivators toward higher education and sometimes pay the bill, they're a large target audience to reach.
Once they go online not only will they have to justify their fees but students may get this crazy idea that the university should be there to help them fulfill their own learning goals.
I suppose that some of the more mercenary universities (or those that can't attract a lot of physically present students) may do the math and realise that they can make more money in total by going for the "pile 'em high and sell 'em cheap (on the internet" approach.
And university sports should just be professionalized. The big-time sports are a cartel designed to avoid paying young athletes while still selling tickets to watch them play, and the small-time sports like the one I played could be moved to public or private sports clubs.
Why are academic education and sports playing so closely tied together?
Most of our football players for example would be a complete waste of a seat in any academic institution.
I think big-time college football and basketball may be pure historical accidents. The sport of American football had its earliest success as an Ivy League college sport; back when the rules were being firmed up Harvard was the national champion. Wikipedia:
In the late 19th and early 20th centuries, gameplay developments by college coaches such as Eddie Cochems, Amos Alonzo Stagg, Knute Rockne, and Glenn "Pop" Warner helped take advantage of the newly introduced forward pass. The popularity of collegiate football grew as it became the dominant version of the sport in the United States for the first half of the 20th century.
And basketball was literally invented by a school sports instructor as an indoor sport suitable for play in school gymnasiums:
http://www.kansasheritage.org/people/naismith.html
... and the inventor immediately joined the University of Kansas.
So basketball and American football may have been born and raised in schools, which explains why they're historically tied to schools in their native land.
My impression is that the other big-time American college sports are late additions to an existing big-time college-sports system. For example, baseball grew up with a complete minor league system operating independently of colleges, and college baseball has only become big within the last few decades. And ice hockey is a regional college sport in the USA but in Canada, where they take it seriously, it's a club sport with a hockey team in every town larger than a critical size. (My impression is that the critical size may be as small as "six people". ;)
I'd go so far as to say that actually teaching students things is the least important service performed by universities even for undergraduates. Far more important for students' future careers is signalling their performance over the course of three years of relatively autonomous study within a controlled environment of their choosing, and the commitment that entailed. Not to mention the huge variety of extracurricular opportunities offered that are particularly hard to replicate through the medium of a website.
Performance could still be measured by grades even in a purely online university.
Extracurricular activities are still possible, people could join study groups within their local area and socialise there, possibly also discounted travel to visit with study groups elsewhere. In addition you could spend the money you saved via cheaper college fees to indulge in some pretty kick ass hobbies.
A replacement is needed. But what could be a better system for providing high level skills with a measure of ability?
"The real question is whether broadening access ends up reinforcing the educational caste system: if you're not smart or rich enough to go to a "real university," you become one of those poor, second-class students with a certificate Online U."