A new online learning initiative made waves this morning, as Georgia Tech partnered up with Udacity and AT&T to launch a MOOC-based master’s program. Please read that link or the Chronicle’s story for background.
I’d like to offer a few first thoughts, based on my work over the past several years. These are provisional comments, not polished assessments:
First, the new degree program’s price point is very important. Inside Higher Ed reports that the normal, face-to-face program costs $40,000 (for out of state residents), and that the Udacity one costs 1/6th of that, or about $6,666.67. This will certainly appeal to some students, especially during a time when Americans are so concerned about price and debt. As with San Jose State’s Coursera experiment, we can imagine people willing to lose some personal contact in order to save money. G-tech is hoping for up to 10,000 paying students.
Second, the staffing aspect is also critical, as the new program relies on fewer and non-tenure-track faculty. IHE again:
Georgia Tech expects to hire only eight or so new instructors even as it takes its master’s program from 300 students to as many as 10,000 within three years, said Zvi Galil, the dean of computing at Georgia Tech.
The university will rely instead on Udacity staffers, known as “mentors,” to field most questions from students who enroll in the new program.
Obviously Georgia Tech controls costs in growing its program, and maybe saves money overall. Shifting away from tenure-line faculty is not a new thing, nor a MOOC invention. Universities and colleges have in fact been reducing the number of tenure-track faculty for years, making adjuncts the most common instructors in American higher education. Georgia Tech’s plan continues the trend.
Third, this is a new direction in attempting MOOC sustainability. We’ve seen partnerships before, but only at the level of classes, nor entire degree programs.
Fourth, this is a computer science degree, which has several implications. CS MOOCs tend to attract a strong online audience, from the Thrun AI class onwards. We might therefore bet that Georgia Tech’s program wins a large audience. On the other hand, computer science is an unusual field compared with other disciplines, at least in terms of engagement with computers. is a very strong limitation. Most academic leaders can dismiss CS as an outlying, unusually cyber-happy field.
Fifth, this program doesn’t appear to offer any pedagogical innovations. It might; we just haven’t heard of any.
Sixth, AT&T? Interesting move, presumably aimed at eliciting graduates for its workforce.
As a coda, I’d like to return to the possibility of higher education hitting an inflection point, my theme from last week. If this program succeeds and spawns imitators, the price of American higher education could well start to drop. Or the field bifurcates: online-only (or online-mostly) programs flourish, bringing average prices down, while some traditional (or blended) face-to-face colleges continue to increase their sticker prices. That’s a big if, but it’s possible. We can also imagine the program fizzling, or imitation limited to computer science programs nationwide – fascinating, but not having much of a broader impact.
What do you make of it?
A couple of thoughts:
1) Computer science really suffers from awful pedagogy. I’m willing to give the discipline a bit of a break — it’s pretty new after all. But what we do know about how best to teach/learn CS (see: Mark Guzdial and his students’ work, for example) runs counter to xMOOC teaching practices.
2) What is the job market for a Masters Degree in CS? I don’t know the answer to this, but this is something quite different than the “everyone learn to code!!11” phenomenon even though it’s likely to be conflated with it. The announcement talks about a “professional degree” — which makes me go “hmmmm,” particularly with the alignment with the corporate sponsor, AT&T.
3) Related: so many of those currently in MOOCs already have their Bachelors Degrees (based on statistics I’ve seen). This means that, contrary to the hype about “access,” we might be talking about a new market to monetize degrees for lifelong learners. Profitable for Udacity and GA Tech (maybe), but not really addressing any “problem” with higher edu as it current stands or costs.
Greetings, Audrey. Good thoughts.
1) What’s so bad about CS pedagogy, apart from xMOOCs? I’ve seen both good (constructivist, project-based) and bad (too much lecturing, opposition to creativity).
2) I’m not sure about the CS MA, but there seems to be a big demand for computing skills at various levels.
3) Ouch. Maybe some of the content will be open, like the lectures. Otherwise one of the MOOC movement’s most powerful supports falls to the ground.
Great point on #3, Audrey. This isn’t quite disruptive in the same way. I don’t know enough about the market for CS MAs, but it seems like here more than anywhere it will depend on the students being well-trained when they come out of the program. This is likely a field that is easier than most to certify – there being resolute answers to many questions – but I imagine it is also new enough that you don’t really need an MA to get a job if you have the skills. This recalls the difference between what Adam Smith called demand and effective demand. Right now they are teaching 300 students in the program all of whom have found a way to pay the $40k (out of state) for the program. This is the current effective demand – i.e. the people who both want to take it and can pay for it. Is there any evidence that there are 10,000 students just itching to take a CS Masters (from a computer), but who can only swing $7k? Where do they get these numbers? Or is it just another example of Say’s law – if you create the supply, demand will emerge to meet it.
I am also curious about the access software and hardware this provides. The average CS student in a conventional Master’s program would likely be supported by site licenses, campus labs, and reduced fees. I imagine MOOCers will not be allowed to MOOCh off these resources in quite the same way. This means they would need to contract for those separately (along, very likely, with any library or textual materials they would need.) This isn’t included in the $7k, but would be in the $40k. I can’t help but imagine it might be a significant cost.
And finally, on your coda, inflection point: what is frustrating is that, with this program and the other credit bearing MOOCs, we are now entering a bit of a tunnel. We won’t really know how these students fare until they are done with the program, which will take three years. This is true for all of the fully online programs that may or may not start flourishing in coming months: they will get a lot of press early on, flash in the pan stuff. Then there is the slog of actually teaching, learning, reflecting and assessing – and then seeing what kind of skills and employment the students can get. Whatever happens in terms of this as an inflection point, it will be a long transition period.
Subtle points, Sean.
I wonder about the license fees: how many are required, and if virtualization can solve that problem.
Bryan, I’ve just finished teaching a graduate computer science course, so I’ve got some perspective on this.
First of all, while the cost is low, it doesn’t seem strikingly low from my perspective in California –
the total cost for our students for an MS in CS is about $16,000. So $7k is attractive, but it’s more like a factor of 2 rather than 5 to 1. Of course, Georgia Tech is a more prestigious institution than my local Cal State, but on the other hand I am certain that my students will be suspicious that a “Georgia Tech MOOC MS” will not be viewed as equivalent to a “Georgia Tech MS”.
Second, I’ve completed one of Udacity’s CS courses and I thought the content was quite well organized and presented. But on the other hand, it was material I already knew pretty well so it’s hard for me to judge how well it works for someone new to the topic.
Third, my students (who of course self-selected for a blended program) tell me that they dislike fully online courses and prefer having some f2f contact with their instructors. Some of this is probably because, based on what they tell me, the quality of the online courses they’ve had has generally been pretty low.
I don’t think most of these students will be that interested in the Georgia Tech program, at least until if and when it establishes a reputation for quality and value. The cost alone made them suspicious of what they would actually get. I think it’s a very interesting experiment, and I know that Zvi Galil is a smart guy, so I wish them well. But I’d be surprised if they meet their enrollment targets, and it will be interesting to see if they can hold that price point if they have, say, 500 or 1,000 students instead of 10,000.
-MB
Thank you for that perspective, Michael.
Sounds like Georgia Tech will have to overcome those value hurdles to win enough students to justify the program. Perhaps Udacity will be able to present some off-campus star power.
Re: cost, are your students paying in-state tuition?
Because this graduate program is self-supporting, i.e. not subsidized by state tax dollars, the rate is the same for everyone. If it were state supported it would run closer to $10,000 in-state I think.
I don’t see huge value in on-line star power. Getting to know a local instructor who works in the field and can help you get an internship seems to me better value than a famous professor who you watch on-line.
I agree on the practical value of a useful rather than famous instructor (setting aside overlaps between the two), but universities still rely heavily on the stars.
More so, in fact, as we keep increasing the proportion of instructional staff that is adjuncts. That structural piece alone could keep the star system in place.
1. As someone on the marketing side instead of the tech side, it’s still not clear to me how competitive MOOC classes and programs (or MOCs, as the case seems to be) will be marketed. It’s great to be first because of the free PR, but when dozens or hundreds of programs are offering CS master’s degrees in MOOC format, each with the capacity to teach 10,000 students, the potential student pool becomes remarkably finite and the marketing costs become remarkably high–especially when you’re paying Udacity 40 percent of your total revenue.
2. Elite universities will always have brand reputation to shore up their MOOC value proposition, and a school like GA Tech will be simply outmuscled on a global stage if it has to compete against MIT, Stanford, Lucerne, Tokyo and dozens of other schools offering similarly priced or similarly structured products.
3. I very much agree with Audrey’s third point. MOOCs are a way to extend education, and I would take it a step further: MOOCs are a way to transfer workplace training costs from employers to employees. Why else would AT&T stake out ground in this arena? As a for-profit business, lowering costs is a focus, and this is a way to do it.
Great thoughts, Jens. Let me try a response to each.
1: Perhaps there will be a cost-cutting competition, with MOC (I like it) prices driving downwards.
Alternatively, a fierce competition could lead to a massive die-off, with a few programs ruling the space. Which bring me to…
2: Elites could rebuild their status in this new space. So now’s the chance – a very short chance – for others to claim a shot through innovation.
3: “MOOCs are a way to transfer workplace training costs from employers to employees” – that may well be the case. Many employers I speak with complain that schools are doing an insufficient job of preparation, and worry about their own costs in (re)training. That might be where this is coming from.
There’s one other detail buried in some of the news reporting that I thought was significant. They discussed a number of different admissions processes, one of which seems to let students bypass the G.R.E. or other traditional admissions criteria and simply prove their readiness to benefit from the program by succeeding in their first two MOOCs. It was unclear if those students would be paying for those classes or if anyone who took the free versions form Udacity could then use those to argue for admissions. If the later, I hope it’s a practice that gets replicated at the undergraduate level. I would love to see colleges considering successful completion of certain MOOCs as an indication of college readiness rather than SAT and ACT scores.
Whoa, good catch, Robert. Are we seeing the expansion of a MOOC ecosystem to internalize prereqs?
If MOOC providers gobbled up the role and the market that SAT-type tests now have, that might be trading one bad penny for another. And I’d hate to see completing some MOOCs added on top of the expectations high school students have on them now to compete for top colleges. But, in theory, activity in MOOCs could be a better indicator of college readiness than what the admissions process currently relies on. If I was an admissions’s officer and an applicant told me they had skipped the SAT and were submitting a few Coursera certificates instead, I’d sit up and notice.
What I’m really hopeful of would touch more on non-competitive colleges with larger numbers of students placed into remedial or developmental courses, often via standardized placement exams. In the U.S., those developmental classes are a de facto probationary status and a giant ripoff. Students pay for them but don’t get credit toward their degrees for them, living as second-class citizens within their entering class. And those students are the biggest share of the college non-completion rate. MOOCs offer a way to correct several problems with that system. UW – LaCrosse has a MOOC experiment right now and also a small-scale online class for incoming freshman designed to help them in the spring and summer before entering so they can avoid the developmental class and place into college algebra. The next step in the evolution in theory could be to use that instead of the developmental course system. Perhaps the MOOC remains non-credit, but it could be free.
That idea of using MOOC completion as college entrance data is very appealing.
I’m not sure about the use of MOOCs for remedial work, though. Didn’t that Columbia study find MOOCs did best for overachieving, self-motivated learners, the opposite of folks ending up in low-end classes?
I haven’t heard of that study, and I’m curious about it, so I’ll go hunt for it. With or without some solid data, there is a narrative taking shape like that, and I’m a little bit suspicious of it. First, because there are some examples emerging of developmental level MOOCs that seem to counter that narrative. (UW – La Crosse MathMOOC is the one I’m most familiar with.) Second, my feeling is that all education only works for motivated students. We can graduate students from traditional classrooms who haven’t found their motivation yet, but whether they’re learning in those cases is questionable. The distinction I would make is, regardless of motivation level, if students need a lot of support or not. Many MOOCs at the moment might not work for students who need a lot of support and therefore might not work for students who genuinely require developmental courses. But, again, I don’t think that’s necessarily in the DNA of MOOCs. The massiveness of them can mean a person is on their own and unsupported, but the massiveness can be used to exactly the opposite effect.
Here’s an intro to that study, Robert: http://www.insidehighered.com/news/2013/04/26/online-courses-are-second-choice-community-college-students-some-subject-areas
Perhaps institutions can supplement MOOCs with support mechanisms. That might be one way to bring the two together.
Thank you.