what if we had Math 101 online, and what if it was well regarded and you got fully accredited and certified? What if we knew that we were going to have a million students per semester? And what if we knew that they were going to be paying $100 per student, right? What if we knew that we’d have $100 million of revenue from that course per semester? What production budget would we be willing to field in order to have that course?
Pointer from Tyler Cowen.
I might push back and say that the same course is unlikely to work well for all one million students. As I have said before, the key is not so much in presenting content. It’s giving students feedback in a way that maximizes their rate of progress.
And you couldn’t come up with a system that worked for different students for $100,000,000?
People (like you, Dr. Kling) who have actually taught online (I don’t consider MOOCs to be teaching) know that there is much more to teaching than just presenting course material.
Otherwise, we’ve had this MOOC for a very long time, called a … library.
I am glad that we have the means to distribute course content to hundreds of thousands of students at once. Now we just have to teach them (i.e., interact and provide feedback). That is the difficult part, and the costly part too, because it’s not very scalable. Precisely because not scalable, maybe good fodder for a startup, per Paul Graham’s suggestion.
I haven’t used Rosetta Stone. Have they solved the problem of providing appropriate feedback for mass users?
I agree that feedback is key to learning. Maybe the question is how many types of learning models are there – and how many corresponding feedback types would one need – to be effective for the vast majority of students. Seems like a handful (i.e. 4-7) could suffice (though it likely depends on the subject).
So the company that (in my dreams) I’ve wanted to found is The Calculus Company. It would teach calculus a zillion different ways: for math majors, for science students, for business majors, for future engineers, for poets, for remedial students, for community colleges, etc., etc., etc. Math is articulate knowledge, i.e., can be totally expressed in words & symbols. That makes it ideal for on-line learning. Tacit knowledge, on the other hand, is harder to teach that way.
I posted an article about this very subject on my own blog, here: http://trotskyschildren.blogspot.com/2014/10/the-demise-of-academy.html
Arnold, I see your point about why the online course, as excellent as it may be, probably isn’t scalable to 1 million students. That said, and haven’t been a student myself, I would draw a distinction between a class such as math and many types of science, where a student’s progress is dependent upon a solid understanding of more basic concepts, and other subjects. For example, I believe a subject such as history could effectively be taught to 1 million students online. This is because the various concepts one can learn in history are, for the most part, independent of other concepts. It is not necessary to learn one subject before progressing to another. As an example of how history could be effectively taught online to large scale, I recall an excellent series….I want to say it was produced in the 1990s, by Eugene Weber called the Western Tradition (you can find it here if you’ve not seen it before: http://www.learner.org/resources/series58.html). I believe a large proportion of high school or college age students who would view the Western Tradition would learn a lot of new information.
To summarize, I would categorize subjects, from those that are highly dependent upon a solid foundation of prior concepts, to those that aren’t. From those subjects most highly to least dependent, the list would look something like this:
Math, including advanced math
Physics
Medicine, biology
Engineering & computer sciences
Chemistry
Econ
Languages
Finance
Accounting
English
History
Online courses would be most scalable for the subjects at the bottom of this list
Andreesen, like most people pushing online education, can’t conceive of people who don’t want to learn or can’t easily learn. Online education works much better for smart people. And even for them, it’s imperfect.