The future of MIT education looks more global, modular, and flexible | Hacker News Reader

The future of MIT education looks more global, modular, and flexible

newsoffice.mit.edu

88 pointsilamont22 comments

>'The report also suggests that MIT consider offering different levels of certification through its online-learning ventures, MITx and edX, and recommends that the Institute redouble its commitment to access and affordability — possibly by increasing MIT’s undergraduate population, which has remained stable for decades despite increasing demand, or by providing flexibility to allow students to complete a traditional undergraduate degree in less than four years.'

Personally, I'd hope to see 'differing levels of certification' expanded to the point of offering credit courses for distance learning. Go ahead and differentiate it further from the residential brand than MITx if necessary.

I've watched hours and hours of OCW, and completed several courses on MITx with a handful more scheduled for the fall. As a rule, these courses are far more challenging and just plain better than what I pay quite a bit to my state University for.

I would jump at the opportunity to pay MIT many multiples more than the pittance they charge for ID Verified courses on EdX in order to take MIT born credit courses.

>'The report recommends that MIT explore more modular approaches to teaching — both online and on-campus. Data from the first 17 MITx and HarvardX courses indicate that only about 5 percent of registrants earn a certificate of completion. This and other edX data, the report says, suggest that “students are focused more on learning certain elements of a class and less on completing what has traditionally been considered a module or unit of learning.”'

I can't say I agree with drawing or even supporting that conclusion from the poor completion rates. I think the overall completion rates are effectively meaningless - signing up is trivial and instant.

I did some crunching of the older stats EdX released a while back and from my foggy recollection even the tiniest measures of intent correlate with dramatically higher completion.

More modular courses aren't going to help the 35% of people who sign up but never even look at the courseware - it's already modular for all they know.

>'These edX results are in keeping with surveys of MIT’s own faculty and students: In a 2013 survey, 25 percent of MIT faculty, and 40 percent of MIT students, indicated that some of the Institute’s course offerings might benefit from being broken up into smaller modules.'

>'The unbundling of classes also reflects a larger trend in society — a number of other media offerings have become available in modules, whether it is a song from an album, an article from a newspaper, or a chapter from a textbook.”'

I can't say I agree with the idea of 'unbundling' courses much at all.

Personally, getting a 'holistic' understanding of the subject is precisely why I'm pursuing a degree and why I chose to work through books or MOOC courses as opposed browsing tutorials or StackOverflow posts as-needed.

I didn't always feel this way. I self taught myself into a career well above my apparent station in life and felt supremely confident that I could go on picking and choosing, learning whatever I needed to know next.

Eventually, I realized that the gaps in what I knew did just hold missing facts or techniques, they constituted missing questions too. Things I didn't even know to ask.

That's the value of a 'complete' course and education - the foundation and supporting framework that ensures you can always keep building.

Interestingly, the median age of people taking courses on EdX is 23-30 and the median age of on-campus students is surely below that. I didn't reach the above epiphany until I was 28 or 29, well into a respectable tower of a career.

It would not surprise me if the people who might think they want for $0.99 singles of education are following the same course.

I would wonder if they'd look at this and worry that there wouldn't be enough interest:

A University's Offer of Credit for a MOOC Gets No Takers http://chronicle.com/article/A-Universitys-Offer-of-Credit/1...

>'A University's Offer of Credit for a MOOC Gets No Takers'

At glance, most coverage of this is a bit misleading as it leaves out a few key points.

* It's not offering CSU credit so much as transfer credits to be applied toward CSU. That's great if I am already attending or plan to attend a school in the CSU system, but it's roundabout at best and useless (in terms of credit) at worst if I'm not.

* You're required to complete a proctored test via Pearson after completing the MOOC for a certification in order to receive those credits - another complication.

* I wondered why I don't recall hearing about this - CSU announced this option in September 2012, through Udacity. Udacity was only a bit over a year old at that point while EdX and Coursera were still in their infancy, 4 and 5 months respectively. There's much greater awareness of MOOCs today and moving forward.

It's purely anecdotal, but my estimation is that a majority of serious (make an effort to complete and pass the course) MITx students would have interest in a credit option.

Thanks for pointing out the requirement for the Pearson certification - I didn't see that mentioned in the article at all!

Also, an interesting point about not hearing about it. It doesn't seem like it's in the best interest of the school - the article mentions that 'The council has not yet advertised its services directly to MOOC students'.

>'The council has not yet advertised its services directly to MOOC students'.

Thank you for pointing this out.

While I'm all about Hanlon's Razor some of this just feels so shady.

In the case of the CSU offering there no direct advertisement, but a slew of articles pointing out that no one signed up.

Looking a bit further I found that my own local University system where I'm registered as a CS major started offering roughly the same deal last fall [1].

There was no announcement and even now I can't find a single thing about the specifics on either site. This is while there's an ongoing study by the Gates foundation on the effectiveness of this sort of program.

Could it be that these institutions want to give the impression that no one wants the $100 MOOC alternatives to their $800-1500 courses?

1: http://technical.ly/baltimore/2013/09/05/umuc-to-offer-credi...

Thanks for pointing these out. The perception in the MOOC community is really not much more nuanced than the headline, so it might be scaring others off from making a big credit push.

That's an interesting article; thanks for sharing! I doubt it would have much of an impact on their decisions because of the limited scope of the offer:

  "The offer applied to only a single MOOC, in computer science, and the credits might
  be useful only to students who intended to finish their degrees at Global Campus."

The idea of breaking down courses is really interesting to me, because I think a "course" is a great way to organize content in a field. Let an expert guide you through a series of perspectives. But completing them often takes more dedication than most busy people can muster.

By breaking courses down to a more granular level of components, there is a huge chance to recognize people for the pieces they are able to complete and let them string those together in new ways that the traditional course designers hadn't thought of. But there are huge challenges around figuring out how different modules could be fit together into new ways. For instance, the Open Educational Resource repositories have had really slow adoption because of it, even though they have in total a ton of great educational content (and quantities of chaff).

I wrote a little bit more about this today, and how digital badges could help a provider like edX crowdsource good organization of many distributed modules. http://remediatingassessment.blogspot.com/2014/08/mit-report...

I think it's really all about switching to a different metaphor for organizing content and examining the entailments of that metaphor to make sure their internal logic leads you to the values you want to promote.

> The report recommends that MIT explore more modular approaches to teaching — both online and on-campus. Data from the first 17 MITx and HarvardX courses indicate that only about 5 percent of registrants earn a certificate of completion.

I am suspicious of this claim because I recently did an analysis of this data: http://minimaxir.com/2014/07/online-class-charts/

The average amount of students who complete classes that they've registered for is about 3.4%, much lower than they indicate: http://minimaxir.com/img/online-class-charts/class-attendanc...

I'm one of those people who registers for several online classes but never finish them, or many times never participate at all.

It would be helpful if they offered a "degree" of some kind for completing groups of classes. Mixing it with the concept of gamification like Khan Academy does might also help with participation.

There's at least one MOOC offering "nanodegrees" for completing a series of classes, but charging $150 a month to participate is a bit on the high side for something that most employers won't care about. There's a bit of a divide between the benefit of education in general and whether that education will have a meaningful impact on your financial bottom line in the long run.

EdX offers the "x-series" - https://www.edx.org/xseries - and Coursera offers "specialisations" - https://www.coursera.org/specializations - they're charged per course. According to them, when you pay the $50+ per course, your completion rate goes up dramatically, to something like 70% - probably due to the fact you've paid something for it.

>'It would be helpful if they offered a "degree" of some kind for completing groups of classes.'

They do.

EdX - xSeries [1]

Coursera - Specializations [2]

Udacity - Nanodegrees [3]

Obviously, no one agrees on what to call these groups. Worse, the length, cost and schedules are equally disparate which isn't a problem specific to groups, but the state of MOOCs in general.

One group offering might be 7 classes at $50 each, with each class covering 8 weeks with an estimated effort of 8-10 hours per week. Another is 9 classes at $50 each, with each class covering 4 weeks with an estimated effort of as little as 3-4 hours per week. One group can be completed in as little as three months, another requires a minimum of 3 years.

1: https://www.edx.org/xseries

2: https://www.coursera.org/specializations?utm_medium=topnav

3: https://www.udacity.com/nanodegrees

Whats worse is the poor quality of the course content compared to the real thing. MOOCs are something else, they should be in the same category as a YouTube video.

Roger Grosse and I (Colorado Reed) were just chatting about this article: we created http://www.metacademy.org with this goal in mind.

Recently, we've been discussing the future of Metacademy and how we can better work towards building an open, flexible, and modular educational system. I'd love to hear any ideas (if you're in Berkeley, I'm also open for a coffee-chat on this topic!)

This is great. I was wondering whether it's possible to have a more formal way of scheduling studying and marking progress (beyond the checkmark, I mean) -- like the ability to create a prioritized list of concepts, set some dates and track completion. And maybe maintain my own little summary notes and maybe even check others' publicly shared notes?

Great point - the checkmark is a trivial attempt: quizzes would be better, and as you mentioned, a review schedule. And the "notes" idea is fantastic (or maybe even flashcards with an Anki-like review schedule); we discussed it several months ago but it fell off the radar. Thanks for the thoughtful feedback!

Very nice. I'll be using this.

* At a glance, some sort of automated progress tracker would be my first request. I'm certain that I'll end up missing sections and/or losing my place and drifting away as long as it's manual.

* Offering an estimate time to completion is great. Leveraging those estimates to give me an option to select 'something that will fit in 5/15/60 minutes' from my current topics would be even better.

* The site snaps, that's great.

* Creating a TOC of the off-site content (video, books) in the overview of each roadmap would be nice. If I decide that I'm going to commit to follow one of these roads I'm going to be looking to source the books and download the videos so that I can watch them on the subway.

* Interactive 'exercises' where possible. This is a major plus of the EdX courses I've done - submitting results or complete Python code and getting instant feedback.

* Suggested background / prerequisites for a roadmap would be nice as well.

Great feedback, thank you. I agree that better progress tracking would be quite helpful. Also, we hadn't thought of giving people learning material based on the amount of time that they have to study -- really interesting idea.

Any plans on implementing business lessons as well? Searching 'marketing','finance', and 'accounting' came up with nothing. As much as I love studying the hard sciences, I feel that at least some degree of familiarity with the domain of business could significantly help the scientist or engineer when they are out looking for jobs, working within a large company, or trying to start their own thing.

A good business curriculum (or even just a series of classes) is something that I find lacking online, as well as within the STEM majors at my university.

We're currently discussing whether to aim for breadth or depth with metacademy, i.e. whether we want to work towards becoming the "ultimate AI resource" or a differently structured wikipedia. If you, or someone you know, has the necessary knowledge/interest to contribute business content, we'd love to work with them to add it to metacademy.

this is very progressive, but in the MOOC era, schools have also moved much of their content to be accessible only to their students. Education is becoming more open, but also more closed.

Reply on news.ycombinator.com