next »

[twofish]

Online courses, of which many are simply glorified correspondence courses, are now being portrayed as equal to traditional education. 

Strange, because I dont' see any difference between most online courses and correspondence courses.

That's because you haven't been looking very hard.  Education is fundamentally a social process, and with the University of Phoenix you get social interaction between the students and the teacher that it is hard to do with correspondence courses.  I'd argue in fact that for basic math courses, the student-student interaction at University of Phoenix is probably better than what I saw at University of Texas at Austin (which accounts for the huge dropout rate in the lower division math classes).

The problem with most models of education is that they look only at the lecture hall where only a small fraction of learning actually takes place.

University of Phoenix is the educational equivalent to McDonald's.  It gives a consistent product that is decent to large numbers of people.  The reason that UoP is not going to produce Nobel-prize researchers anytime soon has to do with internal structure.  In order to produce high level research, you need deep networks between the mostly adjunct faculty, and UoP actively discourages adjuncts from organizing, because if you organize the adjuncts, the administration would very quickly lose control of the university, and the assembly line McDonald's model, falls apart.

A lot of education involves instilling values, and one of basic values one learns at MIT is that when someone says "we can't do this" the response is "why the hell not?" and if the response to that is "because no one else has done it" the response is "so we'll be the first." 

Also, I don't think that pure online interaction is optimal, but at the same time I think that pure brick and mortar works either. 

[jonesey]

Look, it's apples and oranges.  UoP et. al. are Career Education.  They don't do research, they don't want to do research, etc.  They want to provide an education for the 99% of students who aren't going into academia, they need a degree to advance in their job, gain a better job, etc.  UoP fills that niche, as do most state Us. 

What, no one at UoP or Nova or National University is going to cure cancer?  Write (another) book on The Enlightenment that no one will read?  Wow, that's a shame.  Are they going to complete their degrees and advance in their careers?  Absolutely.  Good for them.

Are these colleges equal to the R1s in producing future PhDs and researchers?  Of course not, but they're not trying to be. 

[daniel_von_flanagan]

Academia is the gold standard for the generation, preservation, and transmission of human knowledge.  Some distance education models hope to challenge B&M universities for the "transmission" aspect, but I think we are very far indeed from, say, U. of Phoenix faculty conducting Nobel prizewinning research.

The issue here is that U of Phoenix doesn't consider that it be their core mission. 

An issue, not the issue.

You might take issue with this, but I think your argument will be much stronger when you can point to some serious pure-research breakthroughs that come from outside of academics.

Ummm... 1981 and 1987 Nobel Prize in Physics went to someone who was working for IBM research when they did their work.  1978 Nobel went to two people who discovered the cosmic background radiation while working at Bell Labs.

Bell Labs/Lucent and IBM Watson are traditional, B&M, academic research laboratories (which happen to be privately funded, just as Brookhaven and Los Alamos are similar labs which are publicly funded).  FWIW, Watson was traditionally affiliated with Columbia University (I don't know if this is still the case).  My friends at Lucent, IBM, even Microsoft Research would be astonished to learn that they are being used as examples of the success of online learning.

The latest research in astrophysics is all transmitted online now and all the major journals are now archived online, and any astrophysicist isn't e-mailing their colleagues constantly is now just out of the game.  When people run programs on supercomputers, they don't physically sit in front of the computer.

Again, nothing new, I was using a supercomputer over the phone lines 20 years ago, and a non-super distant mainframe 30 years ago. Online transmission of data is not the same thing as online degrees. You are really stretching here. - DvF

[twofish]

Bell Labs/Lucent and IBM Watson are traditional, B&M, academic research laboratories (which happen to be privately funded, just as Brookhaven and Los Alamos are similar labs which are publicly funded).  FWIW, Watson was traditionally affiliated with Columbia University (I don't know if this is still the case).  My friends at Lucent, IBM, even Microsoft Research would be astonished to learn that they are being used as examples of the success of online learning.

I'm losing track of what we are arguing about here. 

My belief is that the internet is new and can be used to create high level research institutes that are a mix of online and B&M modalities.  What makes the internet different from correspondence courses is that the internet allows for the social networks and interactions that are required for high level research and development in a way that correspondence courses do not.  One thing that excites me is that the internet allows for the ability to bring those social interactions much earlier in a students education.

Part of the reason I think this is the case is that I basically work in a financial research lab with staff scattered across the world.  You do need some face to face communication, but you can interleave it with online interaction.  It's not an either or.

Personally, I don't think that there are any basic technological issues that prevent someone from creating reputable Ph.D. in numerical astrophysics.  The issues are funding and personnel.

[twofish]

Look, it's apples and oranges.  UoP et. al. are Career Education.  They don't do research, they don't want to do research, etc.  They want to provide an education for the 99% of students who aren't going into academia, they need a degree to advance in their job, gain a better job, etc.  UoP fills that niche, as do most state Us. 

I disagree somewhat with state universities.  A lot of state universities (like my alma mater the University of Texas at Austin) have research as their core mission.  The issue is that for the real high level stuff, teaching and research become intertwined, and there is a limit to the skills that you can teach if you don't have contact with people who are active researchers.  The real high-paying jobs require research skills.

For the state universites like UT Austin, this becomes an issue because they want their universities to be the core of high-tech industries.

What, no one at UoP or Nova or National University is going to cure cancer?  Write (another) book on The Enlightenment that no one will read?  Wow, that's a shame.  Are they going to complete their degrees and advance in their careers?  Absolutely.  Good for them.

The problem is that without adding a research component, there is a limit to how far you can advance.  You can get an MBA to work at a big pharma company without knowing how to cure cancer, but if you want to be CEO then you will need exposure to the people that cure cancer.  This is going to start being a problem once the market gets flooded with MBA's.

This is where I think a strategic partnership between UoP and MIT or Harvard Medical School might be very useful.  If you are going to train people to be MBA's for some high tech firm, teaching them computer programming skills might be useful.

Are these colleges equal to the R1s in producing future PhDs and researchers?  Of course not, but they're not trying to be. 

But I think that as time passes that the lack of linkage with an R1 program is going to be a bigger and bigger problem with their core mission.  I can understand why a course on quantum mechanics wouldn't be useful to UoP, but a course on business calculus?

[twofish]

Let me just point out where I think I disagree, since we got sidetracked.

The internet provides something new over correspondence courses because it allows of social interaction, and social interaction is the crucial part of any high level education.  This conversation for example would not be possible with correspondence courses.  Also the internet provides two-way interaction which again is necessary for high level education.  This is new.

The other point is that once you have social interaction, you then have the crucial missing piece that makes high level research and education possible,

[daniel_von_flanagan]

My belief is that the internet is new and can be used to create high level research institutes that are a mix of online and B&M modalities.  What makes the internet different from correspondence courses is that the internet allows for the social networks and interactions that are required for high level research and development in a way that correspondence courses do not.  One thing that excites me is that the internet allows for the ability to bring those social interactions much earlier in a students education.

The internet is not all that new; I used to chat across country on the Arpanet over 25 years ago, and commercial online services appeared just a few years later.  A person with a Z80 machine and a bank of modems could have run a course in 1985 that is almost as interactive (though without graphical capabilities) as the ones today.

I think you underestimate traditional correspondence courses.  The best of them, for example extension at places like Wisconsin, were run by serious people and did have a social interaction component, though much slower than today's internet.

However, running a course remotely is not the same thing as having a virtual university, and right now the B&M anchor does not seem to be eliminable. 

Personally, I don't think that there are any basic technological issues that prevent someone from creating reputable Ph.D. in numerical astrophysics.  The issues are funding and personnel.

What a good B&M university brings to the table is a depth of resources (human and material) onsite.  This includes faculty not only in the narrow specialty, but in the range of specialties necessary to created such a degree.  (I assume that your PhD student in Numerical Astrophysics will be learning some Chemistry, some Physics, and will be taking mathematics courses in Analysis, Special Functions, Mathematical Statistics, and possibly Group Theory?)  Quality faculty want students in their specialties around, and the breadth of a B&M university provides the support for the graduate students who support all components of the program, for example in the form of teaching assistantships, which in turn means undergraduate students.  The traditional learing model likewise provides a revenue model, especially necessary since whatever grants are supporting the astrophysics research are not supporting the math and chemistry research.
Finally, you need labs - not just the place where the thesis research is taking place, but also where the students can be trained in the other subjects.

I am not saying this can't all be worked out, I just don't think we will be seeing it soon.  More likely current distance offerings at good current B&M universities will continue to evolve, and we will gradually see degrees from where the student gets his or her degree from such an institution but where the a substantial part of the effort is online.  - DvR

[dogvomit]

Online courses, of which many are simply glorified correspondence courses, are now being portrayed as equal to traditional education. 

Strange, because I dont' see any difference between most online courses and correspondence courses.

That's because you haven't been looking very hard.  Education is fundamentally a social process, and with the University of Phoenix you get social interaction between the students and the teacher that it is hard to do with correspondence courses.  I'd argue in fact that for basic math courses, the student-student interaction at University of Phoenix is probably better than what I saw at University of Texas at Austin (which accounts for the huge dropout rate in the lower division math classes).

The problem with most models of education is that they look only at the lecture hall where only a small fraction of learning actually takes place.

University of Phoenix is the educational equivalent to McDonald's.  It gives a consistent product that is decent to large numbers of people.  The reason that UoP is not going to produce Nobel-prize researchers anytime soon has to do with internal structure.  In order to produce high level research, you need deep networks between the mostly adjunct faculty, and UoP actively discourages adjuncts from organizing, because if you organize the adjuncts, the administration would very quickly lose control of the university, and the assembly line McDonald's model, falls apart.

A lot of education involves instilling values, and one of basic values one learns at MIT is that when someone says "we can't do this" the response is "why the hell not?" and if the response to that is "because no one else has done it" the response is "so we'll be the first." 

Also, I don't think that pure online interaction is optimal, but at the same time I think that pure brick and mortar works either. 

I couldn't agree with your assessment of UofP more!!!  there are several others that reak as well.

As for the MIT philosophy, now you make me wish I had gone there!  Because my approach has always been "why the hell not" and the response has always been "because it doesn't work that way here!" 

To which I usually want to respond, "actually, your right because nothing works here!"  Then I proved them wrong!
:)

I love your comments.

[twofish]

What a good B&M university brings to the table is a depth of resources (human and material) onsite.  This includes faculty not only in the narrow specialty, but in the range of specialties necessary to created such a degree.  (I assume that your PhD student in Numerical Astrophysics will be learning some Chemistry, some Physics, and will be taking mathematics courses in Analysis, Special Functions, Mathematical Statistics, and possibly Group Theory?)

They should have gotten the non-astrophysics stuff as an undergraduate.  That's one of the reasons I think it would be less administratively difficult to put a PhD. in theoretical astrophysics online than a good undergraduate program since for a Ph.D. program, you don't need a range of specialities.

The thing that the internet provides is that it allows one to disaggregate those functions across several institutions.

Finally, you need labs - not just the place where the thesis research is taking place, but also where the students can be trained in the other subjects.

Again, you don't need labs for a theoretical astrophysics Ph.D. program.  Just computer time, and that you can get over the internet.  You do need social contact with people working in the field, and that is the second hardest part, but the thing about University of Phoenix is that it demonstrates that it is doable.

The big bottleneck with having an online Ph.D. program in astrophysics is the revenue model.  Most universities use Ph.D. students as cheap labor to get grants and tuition, and there isn't the immediate cash that you would get with, for example, an MBA program. 

I don't doubt that bricks and mortar are necessary for some things, but having things online can change things quite drastically.  For example, an MIT undergraduate in thirty years could be physically located at Fermilab working as a lab technician while getting their general education requirements online, learning Spanish literature from a professor that is actually in Spain.

I am not saying this can't all be worked out, I just don't think we will be seeing it soon.

One of the things about "soon" is that it means different things to different people.  I think it will take about ten years to put in all of this infrastructure, but it is important to think ahead since bureaucracies are slow and ponderous and to react to changes that will happen ten years from now, involves some thinking about those changes now.

More likely current distance offerings at good current B&M universities will continue to evolve, and we will gradually see degrees from where the student gets his or her degree from such an institution but where the a substantial part of the effort is online.

I think the way things will evolve is some major collaboration that changes the rules of the game.  Something that might work is MIT + Thomas Edison State College + some major community college system + google + wikipedia.  The pressure for change has to come from the outside, since things won't change unless there is some outside forcing function.

Linking in with the community college system is particularly important since community college is where people get moved from lower class to lower middle, and linking in a community college with an Ivy League university or major research university, means that you now have class mobility from middle to upper.

[daniel_von_flanagan]

They should have gotten the non-astrophysics stuff as an undergraduate.

This is fantasy.  A strong student in a good university might have a research-ready background in one of these subjects by the time they start graduate work, but not all that they need, there will always be many gaps.  If your astrophysics students did all the lab physics and chemistry they needed as undergraduates, then they likely didn't have enough time to get to stochastic differential equations or spatial statistics or quantum mechanics at more than an introductory level. 

I don't know how much experience you have on graduate admissions committees in your field, but in mine even the students coming from Cal Tech and Harvard have such gaps.  Moreover, the gaps are bigger than they used to be, because the techniques used in research keep accumulating, especially in the theoretical/mathematics subjects (eg, someone coming up with ways to map neutrino distributions 30 years ago did not need to know Bootstrap or Wavelets).

The thing that the internet provides is that it allows one to disaggregate those functions across several institutions.

This disaggregation is fine for the actual research, but for the graduate training there is still the absence of revenue model.

A student comes to my department to study with me because I have a certain mix of interests and knowledge.  For each component of that knowledge, there are better people out there.  In your model, a student would do better to take each of the classes I teach from someone else.  This eliminates the market for generalists like me, and also the future market for many of the graduate students.

having things online can change things quite drastically.  For example, an MIT undergraduate in thirty years could be physically located at Fermilab working as a lab technician while getting their general education requirements online, learning Spanish literature from a professor that is actually in Spain.

Again, this is not new, people have been doing this for decades.

I am not saying this can't all be worked out, I just don't think we will be seeing it soon.

One of the things about "soon" is that it means different things to different people.  I think it will take about ten years to put in all of this infrastructure

You must be very young indeed.  I'm still waiting for the flying cars.

More likely current distance offerings at good current B&M universities will continue to evolve, and we will gradually see degrees from where the student gets his or her degree from such an institution but where the a substantial part of the effort is online.

 I think the way things will evolve is some major collaboration that changes the rules of the game.  Something that might work is MIT + Thomas Edison State College + some major community college system + google + wikipedia.  The pressure for change has to come from the outside, since things won't change unless there is some outside forcing function.

Except for the Wikipedia part, this is already happening without outside pressure.  My department (state RU/VH) has been coordinating with community colleges in my state for years; many of their faculty were students of ours.  We also make very substantial use of the web when it is appropriate, and my university has an active online classroom presence, again when appropriate.

I can't speak for the Ivies, though I know that some of the private top-10 research departments in my field, and all of the public ones, are actively engaged with various forms of such outreach. - DvF

[twofish]

This is fantasy.  A strong student in a good university might have a research-ready background in one of these subjects by the time they start graduate work, but not all that they need, there will always be many gaps.  If your astrophysics students did all the lab physics and chemistry they needed as undergraduates, then they likely didn't have enough time to get to stochastic differential equations or spatial statistics or quantum mechanics at more than an introductory level. 

The astrophysics program (UT Austin) that I attended was pretty self-contained, and it wasn't the case that courses outside the department were essential.  I took a few math courses for personal interest, but those weren't necessary for research.

Astrophysics doesn't require stochastic differential equations.  Spatial statistics is necessary, but that gets taught in the class on observational extragalactic astronomy.  Quantum mechanics is interesting because you need different things depending on what you are doing.  If you are doing stellar astrophysics, you go from intro quantum to radiation theory.  If you are doing cosmology, then you need field theory.

Moreover, the gaps are bigger than they used to be, because the techniques used in research keep accumulating, especially in the theoretical/mathematics subjects (eg, someone coming up with ways to map neutrino distributions 30 years ago did not need to know Bootstrap or Wavelets).

But most of those gaps are research area specific.  For example if you are doing neutrino radiation hydrodynamics in supernova, you don't need to know wavelets, but you do need to know monte carlo techniques.  In any case, the way to deal with those gaps, is to download papers from the web, and spring for books on amazon.

One reason I think it would be easier for an online Ph.D. program to exist is that hopefully any student in a Ph.D. program should be somewhat self-sustaining.  They should be able to figure out what the gaps in their knowledge are, and go out and find the resources to fill them.

A student comes to my department to study with me because I have a certain mix of interests and knowledge.  For each component of that knowledge, there are better people out there.  In your model, a student would do better to take each of the classes I teach from someone else.  This eliminates the market for generalists like me, and also the future market for many of the graduate students.

There isn't much of a present market for graduate students.  The system as it stands can only support 20% of the graduate students moving in.  The rest end up computer programmers for investment banks or something similar.

having things online can change things quite drastically.  For example, an MIT undergraduate in thirty years could be physically located at Fermilab working as a lab technician while getting their general education requirements online, learning Spanish literature from a professor that is actually in Spain.

Again, this is not new, people have been doing this for decades.

Actually no.  Undergraduate research basically didn't exist until the 1970's, and I don't know of anyone that has tried to redo the entire undergraduate physics curriculum from scratch with undergraduate research as the core.

Except for the Wikipedia part, this is already happening without outside pressure.  My department (state RU/VH) has been coordinating with community colleges in my state for years; many of their faculty were students of ours.  We also make very substantial use of the web when it is appropriate, and my university has an active online classroom presence, again when appropriate.

I can't speak for the Ivies, though I know that some of the private top-10 research departments in my field, and all of the public ones, are actively engaged with various forms of such outreach. - DvF

The politics might be different in your state, but I do get the sense in Texas that the research universities are somewhat annoyed at having to accept credits from two year colleges and only do so because the state mandates it.  When I mean "outside pressure" I'm talking in general terms.  State universities for example have political pressure to keep enrollment relatively open and to interact with the community in ways that the Ivies don't have, so community colleges have embraced technology more than the four year state universities and four year state universities more than the Ivies.

I do get the sense however that these programs are considered "extensions" to a core systems.  What I predict is that things will turn "inside out" and that these "extension" programs are going to turn into the standard way of doing things.

One big worry of mine as an MIT alumni is that MIT is getting left behind in all of this.  MIT is a strange mix.  It's basically a vocational-technical school that knew the right people, but while open courseware is a very good thing.  I don't see the alliances with two year community colleges that I think are going to be essential for MIT's survival in the coming decades.

[daniel_von_flanagan]

The astrophysics program (UT Austin) that I attended was pretty self-contained, and it wasn't the case that courses outside the department were essential.  I took a few math courses for personal interest, but those weren't necessary for research.

Astrophysics doesn't require stochastic differential equations.  Spatial statistics is necessary, but that gets taught in the class on observational extragalactic astronomy.  Quantum mechanics is interesting because you need different things depending on what you are doing.  If you are doing stellar astrophysics, you go from intro quantum to radiation theory.  If you are doing cosmology, then you need field theory.

Any field that avoids the latest theoretical tools is not going to advance very quickly.

I am at a space-grant institution, and the astrophysicists here do know SDEs.

A student comes to my department to study with me because I have a certain mix of interests and knowledge.  For each component of that knowledge, there are better people out there.  In your model, a student would do better to take each of the classes I teach from someone else.  This eliminates the market for generalists like me, and also the future market for many of the graduate students.

There isn't much of a present market for graduate students.  The system as it stands can only support 20% of the graduate students moving in.  The rest end up computer programmers for investment banks or something similar.

However, your model even removes the market for graduate students as graduate students, as their niche will be filled by the faculty who no longer teach.

Actually no.  Undergraduate research basically didn't exist until the 1970's

This is not correct.

The politics might be different in your state, but I do get the sense in Texas that the research universities are somewhat annoyed at having to accept credits from two year colleges and only do so because the state mandates it.

The articulation issues is a difficult one.  The resistance of the flagship campus to free articulation from the CCs comes from issues of inequivalent content, grading standards, and the desire of the receiving institution to control its degree.  That does not mean that there is no effort  to have a coordinated curriculum.

community colleges have embraced technology more than the four year state universities

Not in any of the states where I have taught! - DvF

[twofish]

Any field that avoids the latest theoretical tools is not going to advance very quickly.

The theoretical tools are different for each specialization.  There are too many different tools to know them all.  What you can do, and what a good undergraduate education can give you is the basic literacy you need so that if you have to learn algebraic topology, you just go to amazon and buy the book.

I am at a space-grant institution, and the astrophysicists here do know SDEs.

That's interesting.  I took a course on SDE's as a graduate student for fun.  Tried to find some use for them in my research but couldn't.  They get used a lot in finance.

PDE's are really, really important, since hydrodynamics is a core topic in astrophysics.  I wouldn't put SDE's as a core graduate requirement.  Also, if you have a working knowledge of PDE's, SDE's can be learned quickly.  Just add another term in the chain rule to get Ito's Lemma. 

Also in every case off-hand I can think of where SDE's would be useful, you can reduce them to PDE's with the Feymann-Kac formula.

I should point out that banks like astrophysics Ph.D.'s because they can usually figure things out for themselves.  Here the web, here are a few books, go out and make us some money.

However, your model even removes the market for graduate students as graduate students, as their niche will be filled by the faculty who no longer teach.

I don't think so.  Astrophysics graduate students can teach intro physics courses, and if you increase the demand for intro physics courses that increases the demand for astrophysics graduate students.

Actually no.  Undergraduate research basically didn't exist until the 1970's

This is not correct.

Interesting.  Were there organized undergraduate research programs before Margaret MacVicar started the UROP program at MIT in the late 1960's?  The proposal that I have is to make undergraduate research the core of an undergraduate physics curriculum.  I think that Olin College is doing something similar for engineering.

The articulation issues is a difficult one.  The resistance of the flagship campus to free articulation from the CCs comes from issues of inequivalent content, grading standards, and the desire of the receiving institution to control its degree.  That does not mean that there is no effort  to have a coordinated curriculum.

But

1) grading standards and quality has been the excuse for a lot of nonsense in higher education.  If quality was the real issue you can come up with an assessment that demonstrates that a student has a certain level of knowledge.  Most of the time, quality isn't the real issue, it's jobs and money.  I don't mind people looking after themselves and protecting their own jobs and income.  I do mind when they invoke high ideals to do it.  (And to be fair, state universities are much less bad about this than the Ivies.)

2) if there wasn't pressure from the state, the flagship universities would not even try to coordinate with CC's.  There are plenty of bureaucratic ways of getting nothing done, and flagship universities try all of them to keep from accepting CC credits, but ultimately the state government starts screaming at the universities to get something done.

[twofish]

One other thing.  I do get very annoyed when people try to use "fancy math" as a replacement for physical insight.  Math for the sake of math isn't a very good way of doing science.

[dogvomit]

Online courses, of which many are simply glorified correspondence courses, are now being portrayed as equal to traditional education. 

Strange, because I dont' see any difference between most online courses and correspondence courses.

That's because you haven't been looking very hard.  Education is fundamentally a social process, and with the University of Phoenix you get social interaction between the students and the teacher that it is hard to do with correspondence courses.  I'd argue in fact that for basic math courses, the student-student interaction at University of Phoenix is probably better than what I saw at University of Texas at Austin (which accounts for the huge dropout rate in the lower division math classes).

The problem with most models of education is that they look only at the lecture hall where only a small fraction of learning actually takes place.

University of Phoenix is the educational equivalent to McDonald's.  It gives a consistent product that is decent to large numbers of people.  The reason that UoP is not going to produce Nobel-prize researchers anytime soon has to do with internal structure.  In order to produce high level research, you need deep networks between the mostly adjunct faculty, and UoP actively discourages adjuncts from organizing, because if you organize the adjuncts, the administration would very quickly lose control of the university, and the assembly line McDonald's model, falls apart.

A lot of education involves instilling values, and one of basic values one learns at MIT is that when someone says "we can't do this" the response is "why the hell not?" and if the response to that is "because no one else has done it" the response is "so we'll be the first." 

Also, I don't think that pure online interaction is optimal, but at the same time I think that pure brick and mortar works either. 

University of Phoenix = McDonalds

Good analogy!

UP charges you for an education that may even hurt your credibility.
McDonalds offers a substitute for food that will eventually kill you if you eat too much, or at least you'll be buying more peptobismal!