Hi future

In the past three weeks, Georgia Tech received nearly twice as many applications for a new low-cost online master’s program as its comparable residential program receives in a year. The degree—which uses Massive Open Online Course technology—is the first of its kind, and its popularity suggests a growing demand for online learning.

There is more here, by Douglas Belkin at The Wall Street Journal.


Price elasticity of demand? $6000 versus $44,000 (plus living expenses) is a compelling argument.

Why is there so much greater demand from Americans? Do foreigners place more value on the signaling component of the residential degree? I can imagine if you're a foreigner that getting through the admissions office represents a certain vetting that's not so important to an American, but the extent of the difference is astounding. Does GT have some reason to prefer foreigners in its residential program?

You get a student visa with a residential degree.
You can get a better chance at an H1B visa with a US advanced degree -- i don't know if an online degree counts the same way for visas or not.

That's a good candidate explanation. I'd bet the online degree does not offer visas. I should have thought of that.

There's got to be a way to figure out how much people are paying for the visa component of residential degrees... It would make an interesting paper. Here it looks like GT is picking up $20 bills that US citizens ordinarily just throw away. GT is scalping immigration slots.

I keep arguing that if you want a pro-liberty immigration policy, open borders is not the answer, visa prices are the answer. The people who make citizenships and other immigration slots ought to seek compensation for doing so.


My theory is:

(a) Americans are in general more open to trying a new fad. I'd be curious to know the demographics for the online enrollment. How many are fairly well settled people just doing this on a whim since MOOC's seem the in-thing to do. I also wonder how stringent were GT's pre-requisites for enrolling.

(b) If a foreign student wants a "fake" American degree I think there might be some cheaper options & a lot of the native employers who would be fooled by a fake American degree anyways wouldn't know the difference between GT versus the cheaper degree-shop.

You would have to disambiguate 1) added human capital; 2) consumption value (which might also be higher for a foreigner); 3) signalling (which also might be higher for a foreigner who otherwise finds it hard to signal credibly); 4) the short-term visa and the greatly increased prospects for a longer term visa.

It's an interesting problem. Simplistically, it looks like foreigners are valuing whatever they get that the US students don't get at $38,000 plus whatever the school makes on other costs.

FWIW, I had never heard of Georgia Tech until I came to the US.

Alternatively, foreigners are more astute buyers. Or Americans are early adopters.

That's a lot of money to pay for nebulous conservatism.

I suppose it's a candidate explanation, but I doubt it.

Conservativism? You know what I mean...

When are you going to offer a MOOC on chess TC? Playing here in the Philippines, against strong masters and experts, I was able to raise by rating by at least 200 points in less than a year. Now I'm clearly Class A and going for expert.

Now, I hear you doubters saying: well, since nowadays you can play chess software such as Fritz, which is rated at least a grandmaster, why do you need a MOOC with human instructors? Have you not read "Average is Over"?

No, I've not read it yet (but have ordered it on Kindle), but the answer is: humans respond better to feedback from other humans. Sure, I can set the chess software to "human mode" and "blunder mode" (they have these settings in computer chess) so it plays an occasional mistake, like a human, but there's nothing like beating a human in chess, and the reward of knowing you are not playing against a soulless machine. It's hard to articulate better than that, sorry, but online chess players go to great lengths to detect cheaters who use a PC to help them move--for that very same reason. Then how does a MOOC, which is online, introduce the human element? Well, good question. I admit that face to face instruction (such as what I'm getting) at a local chess club is best, perhaps akin to being in the front row seat at a classroom in Harvard, but a MOOC, with fellow humans would be second to that experience and beats having to play your PC alone.

Are the two GT degrees completely indistinguishable from one another?

Even if GT did try to make 'em indistinguishable it'd be rare that a recruiter wouldn't ask. Especially once this sort of scam gets more popular.

Well ... I share your excitement, but I am still cautious as I learn more about this and having given several lectures live in Karachi recently and then followed up by Skype.

I am still not convinced that it is possible to address the cheating issue in online education. Sure, we can get students into supervised conditions during exams. But homework is where the cheating is likely to occur. In an in-person class, even one that is relatively large, professors (and TAs) can quickly come to conclusions about cheating on homework by looking for consistencies between in class tests and in class discussions on the one hand and homework performance on the other hand. Without being able to place faces to names and without eye contact, this becomes very difficult.

What happens when an online student does miserably in tests but aces the HW? Distance creates lots of temptations.

I can only come up with three possibilities for the purpose of homework. Admittedly, I do not work in education.

1. Signal task completion to future employers. If so, what does cheating mean in the this context? The task is still completed.

2. Assist students in learning concepts. If so, cheating on homework results in a lower test score. So long as homework is weighted appropriately, where's the problem?

3. Busywork. Who cares?

Culturally, Georgia Tech places a high value on academic rigor. And particularly in Computer Science (I have a BS in Computer Science from Georgia Tech in 1999), the academic approach is both theoretical and pragmatic - the teaching tends to be theoretical, but evaluations are in code, with very little in the way of direct instruction. It's not about mastering syntax, it's about showing you understand the theory, and that you can figure out how to apply it on your own. My professors were happy to let me "cheat," if by "cheat" you mean go outside of their instruction to find a way to solve the problem on my own. There are standard checks in place to make sure I'm not stealing work from a classmate, but other than that I'm free to solve the problem however I can.

I'll say also that the value of a Tech degree is tied closely to that approach. Attrition rates tend to be high, and grade inflation is not much of an issue. Sink or swim and all that. I expect that approach not to differ in a MOOC offering.

So it remains to be seen how well this works in practice. Based on cultural (academic approach) identity, I suspect it may work better at Georgia Tech, in this specific discipline, than in many other universities.

From a strictly applied perspective I'd think of Comp. Sci. as a very unlikely target for a MOOC. Wonder why GT selected that to go MOOC.

Rahul, I have conjecture on this.

I've taken a couple of online classes via coursera. One of them was a programming class, and the other a songwriting class. The programming class was much easier to do via peer evaluation because it was clear whether a program was functioning as requested. The program either met a specific requirement or it didn't - and the well defined requirements pointed me like an arrow toward the solutions I needed. Also the discussion boards were active and useful. In contrast, the songwriting class evaluation was so fuzzy that I found peer evaluations next to useless (class content excellent, but I got little out of peer review giving insight into my efforts).

I guess all that is to say that the programming class worked very well in a MOOC format. fwiw.

Makes sense but what I mean is often one hears that Programming is not Comp. Sci.

My impression of Comp. Sci. Masters is more of Compilers, Architecture, Algorithms, Computational Complexity & that sort of more theoretical thing.

Comments for this post are closed