Slashdot Mirror
Interview: Ask CEO Anant Agarwal About edX and the Future of Online Education
Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the CEO of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects. As of 2014 edX had more than 4 million students taking more than 500 courses online. The organization has developed open-source software called Open edX that powers edX courses and is freely available online. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.
Is it time dump / rework the degree system?
A big issues with the Education system is that there is way to much push on getting a degree and if stuff was re engineered to better fit today's world. Maybe the trade / tech schools would be in better shape and or the USA will have something like the German apprenticeships split schooling system.
http://articles.chicagotribune...
Coursera is the platform i'm spent time on. Up until six months ago, there was a huge variety of offerings, as a one or two course format. Lately, there seems to be paid 'specializations', and the offers have decreased. My questions is: "will MOOCs be consolidating to a few paid offers in the future ?"
and do you have a study prep course for kobayashi maru
A couple of years ago edX got a bunch of investment money and was being run as a business, with hopes of making money from the course offerings, despite having no clear business plan or strategy for doing so.
(I believe originally the plan was to have companies pay to get lists of high-scoring graduates for potential employees, which didn't work out, and last I talked to [edX chief scientist] Piotr, he said you had something going with Pearson but couldn't elaborate because of NDA.)
What is your business plan and what strategy do you have for making money?
Do you foresee MIT and other prominent colleges offering a wide range of affordable online-only degrees at any point in the future? Right now there are a few accredited universities offering online-only degrees in a number of fields (most notably, Western Governors University). But most fields of study at most major universities still require old-fashioned physical class attendance. Aside from a few token classes and maybe one or two fields of study, most universities still do classes the same way they've been done for hundreds of years (show up to this classroom at 10:30 a.m., good luck finding parking).
Do you ever see a future where a student could get a real degree from MIT in a mainstream major without ever setting foot on campus?
SJW's don't eliminate discrimination. They just expropriate it for themselves.
I'll bite. If Republicans did get a chance to develop education for the masses, it would probably look a lot like this. No wasted overhead (teachers unions and administrators), optimized for home schoolers (so they can build their choice of religious education into the day), and no socially-motivated mandates on curriculum (e.g., no health class, no gym class, etc.),
If you doubt me, check out the demographics of who is enrolling in "open enrollment" online course at the K-12 level today...
What about the quality of the course? Not just the content, but how it is organized, how it is taught/facilitated, etc? I know that QualityMatters is out there, but is "certifying" a course as being QM compliant enough?
Don't blame me, I voted for Kodos
>> As more people, who previously had very limited opportunities in education, are getting help from the likes of you
Is this a serious question? Do you realize that there are books available for free from your local library that can provide much more of an education that any bunch of videos can? Or do you feel threatened by that too?
(Long story short, if you need to watch a video to understand a topic...you're probably not really competing for my job.)
Pearson are a bunch of price gouging douchebuckets.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
These look like post-secondary school courses. So, bzzzt, thanks for playing.
>> post-secondary
I'll bite again then, using the same structure. If Republicans did get a chance to develop post-secondary education, it would look a lot like this. No wasted overhead (tenured professors, administrators, sports or clubs), optimized for people already working (so they can avoid the SJWs who hang around campuses and coffeeshops), and no socially-motivated mandates on curriculum (e.g., no SJW themed composition courses, etc.)
If you doubt me, check out the demographics of the people who enroll in rural two-year and "commuter" four-year colleges today. :)
So far I've completed quite a few online courses in edX and Coursera. One thing that find troubling is that there is no way beforehand to know if a course really is worth the time and effort to commit to complete all the activities or not. Honestly, some courses are *pretty* bad while others you couldn't find any better even paying big bucks. And not always is obvious to idetify which is which without a significat effort.
Do you think that could be useful to solve this problem:
- Require every course to have a detailed syllabus in the suscription page?
- Publish a detailed previous course rating to help students to save their time instead of being forced to invest and waste it sampling them?
- Or like Amazon do with their product, publish previous participants opinions (grouped in some maningful way, by obtained grade, courses completed, age, etc.)?
Maybe by the time (if any) that you make it out of high school you'll understand that the delivery method and the content are orthogonal things.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
My point was that to discuss that these courses are for homeschooled K-12 kids is just ridiculous. They are college courses.
And your point is what? Not everyone can afford 40K/year for "The College Experience." Not everyone feels that tying yourself down with 80K in student loans before you're 25. As for tenured professors, you do realize that most universities are not hiring tenure track positions, right? They're relying almost completely on adjuncts. And what's the political orientation of universities? (Hint, it isn't Republican or Libertarian). So - GASP - these bastions of left wingedness are exploiting professors now! Administrators. Well, any organization needs some degree of administration, but if it is a school without walls, then, yes, it'll need less of it. As for sports and clubs, well, heck, if I went to a college for sports and clubs rather than education, see my 40K/year statement above. What you're saying is that the current unsustainable tuition/boarding costs just need to keep rising. You must be a left wing limo liberal elitist.
How does edX think of the issue of measurement at scale, in ways that go beyond multiple-choice or other constrained response models? In particular for courses in subjects where responses are not structured: liberal arts, social sciences, art...
Traditionally the essay is the preferred mode of measuring student progress. For good reason. The trouble is that essays need to be graded. NLP-based tools for scoring essays can help a bit -- at least for determining holistic scores. But the state of NLP is such that reliable and accurate tools for determining whether the content of student responses is correct are still lacking. And in any case such tools can be gamed.
The default in MOOCs seems to be: peer grading. How well does that work?
Are there other ways of solving the measurement problem?
I'll word them in the form of questions?
Can you improve the search capabilities of the Discussions to make them less write-only? It would be especially good if while writing a new comment, the student's keywords and sentiments towards the keywords could be evaluated in a high-dimensional space so that 'nearby' (cosine distance metric?) comments and threads would be brought to the student's attention. Pie-in-the-sky to support smoothly merging new comments into existing discussions... Perhaps deferring unrelated parts or helping to make those parts into bridging comments to other threads?
Could a professor invert the presentation using EdX? Ask questions first, and use the wrong answers to guide students to study only the parts they need to study? Then when the student had worked all the way through the course, the student would already have faced the exam questions and be ready to answer them successfully, while also spending the minimal necessary time studying the material. For students like me, the most effective teaching videos would be fairly short, with the professor explaining and discussing the material one-on-one with a student who shares my specific points of ignorance.
As a financial model, have you considered rigorously proctored exams? The price of the exam would be set to cover the costs of administering the proctored exam, plus a percentage to EdX to support the development of course material and the operation of the website...
Those are my top 3 suggestions, and if you don't like them, I have others (plus apologies to Groucho). Over the last few years I've taken quite a number of courses from EdX, but it mostly signifies having too much time on my hands, and most of the courses were on a scale from entertaining to shallow. I definitely feel that the weakest part of EdX is the Discussions, however. Lots of teaching talent, but the EdX tools strike me as weak or immature.
Freedom = (Meaningful - Coerced) Choice != (Speech | Beer^2), and sad sock puppets' bad mods avail them naught.
Coursera and EdX primarily source their instructors from college professors. Udacity is more open to bringing in experts from technical fields. In my experience, college professors in the computing fields are often people who chose to get a PhD straight out of college (perhaps because they were mostly interested in research), and as such may not have much (if any) industry experience. Why has EdX chosen a model that focuses limits the ability for technical experts to provide classes?
EdX came out of MIT. MIT used to be a strong proponent of OpenCourseware. OpenCourseware classes were both open access *and* open source, so that other instructors could use the material, though admittedly many classes (at least in the computer security domain) never posted videos. EdX courses are open access, but rarely (if ever?) open source. Do you think dropping the requirement to be open source has helped EdX succeed where OpenCourseware failed?
To my mind, a platform like yours ought to be well suited to offer a very wide range of courses in niche subjects. I tried, just for the heck of it, to search for things like Inuit Language, Bobbin Lace and K-Theory, none of which turned up results (as expected). I understand that it takes time, of course, to make these things, but my question is - will there ever be that kind of courses which may only attract small audiences, or is this going to yet another 'profit first' educator, like so many others?
Not at all. Previous posts, which I were commenting on, were discussing homeschooling of K-12 kids. My comment was absolutely appropriate.
Professor Anant ex students from 6002x are eagerly waiting for a next course from you as you are probably also missing the contact with the students . A topic about Multiprocessors would be nice even if it's about the 1990 MIT Alewife Machine. Is there a chance you'll give another earth moving course on Edx?