More On The MOOC Thing

After completing my first MOOC (I reported on it here), I decided to go try out the major MOOC platforms out there. The dataviz MOOC I took with the Knight Center on Journalism in the Americas was offered in-house, so, I decided to go look up the big MOOC providers: Coursera, Udacity and EdX. I have signed up for a Coursera course but the start date is TBD. I have also signed up for a course in EdX scheduled to start on 2/20. I also registered for Introduction to Statistics over at Udacity, taught by one of Udacity founders, Sebastian Thrun. I have now completed this course and just started another statistics one there.

So, how is a Udacity course different from the MOOC I have taken with the Knight Center? First of all, it runs continuously. It is self-paced, so, students can jump in and out at any time, take as much time to complete as they want. The course itself is a series of very short Youtube videos (the longest ones are around 2-minutes long) where students are introduced to a series of statistical concepts and have to apply them and answer questions that are instantly marked (correct or not). These bite-sized lectures are a big difference with the Coursera model of putting entire existing college courses online.

At regular intervals, mainly, the end of a logical unit, students have to complete a problem set as well as as optional programming units using Python to program the statistical concepts introduced in the immediately preceding unit.

The lessons themselves are a bit uneven. Some are really easy and then, all of a sudden, you get hit over the head with something brutal. Also beware. A lot of stuff is not standard. The other downside is that there is a gap between the lessons and the problem sets. It is very possible, I think, to complete the lessons without difficulty but get stuck on the problem sets. I got stumped a couple of times simply because the styles of the lecturers (Thrun does the lessons and Adam Sherwin does the problem sets and the final exam) is very different and I could not figure out what I was being asked to do in the problem sets. I also gave up on the programming units (they are optional anyway) after the first few because they required some knowledge of Python that I don’t have. And besides, I was in the middle of learning R and I did not want to confuse myself with learning another computer language at the same time.

Also, I must say that I very much doubt that anyone taking this course for the first time would be able to pass the statistics course in our program. It is not statistics 101. It is a lower level stats course. If you want a real take-down of the course from a statistics expert, read this blog post from someone who also took the course. The population v. sample issue is the one that most jumped at me. Every so often, I would go back to my stats texts for the proper formula and written explanations.

As with all other Udacity courses, there is a wiki to go with it, although there isn’t much there. And then, there are the discussion forums where students can post questions and other students are expected to contribute answers. The course assistant popped in in some to resolve issues with the units themselves.

At the end of the course, there is a final exam consisting of 16 questions, you need 8 correct answers to pass “with proficiency”, but you can take and submit your answers as many times as you want.

As I mentioned, the course runs continuously and is self-paced, so, compared to my previous MOOC, it was a pretty lonely experience because, by definition, you work on your own. Also, it looked like the course had been offered for the first time last Spring/Summer, so, the bulk of the forum activity dated back from that time. As I was going through the course (it took me about a month, total), there was not much activity in the forums. Not only that but, as I perused the forums just to see the kind of questions asked, I realize that the bulk of them were not so much questions as people who already knew the subject and offered feedback on the pedagogy (a better way to understand Bayes theorem!). So, I don’t know if Udacity discloses user statistics but it looked to me that the main population of students (from the forums) was, as in my previous MOOC, composed, not really of college students, but of more advanced (older?) people who already knew statistics and were, like me, getting a taste of the platform and the pedagogy.

One last thing on this. Udacity upgraded its platform in the middle of my taking the course. This was a nice improvement as I initially kept getting error messages and had to constantly refresh the page.

The million-dollar question is this: how often will the course be updated? Some students in the initial version noted a few errors and corrections were placed in the video notes / comments but the videos themselves were not changed. Are these videos recorded once and for all, never to be updated? If that is the case, then, it is pretty ghastly. And though the course does not require a textbook, I would recommend to get one anyway, even something like Statistics for Dummies.

The Udacity course I am taking now is in its first offering but has me chomping at the bit because, even though it is self-paced, they instructors only post one lecture / problem set per week (for 16 weeks, as I understand it) but these first ones, at least, are pretty easy so  I usually complete them on the day they get posted and then, I have to wait a week to get the next one, which seems to be a waste of time for me. Maybe once we are past the initial offering of the whole course, future students will have it truly self-paced.

Maybe this once-a-week format is because it is the first offering of the course is simply because the entire course is not yet ready. If that is the case, then fine. After all, one would not want a repeat of this Coursera fiasco:

“When word spread this weekend that a massive open online course about online education had to be suspended due to technology problems that left many students angry, officials from Coursera and the Georgia Institute of Technology were not available for comment. In interviews Monday, however, officials of both Coursera and Georgia Tech confirmed that the major issue concerned the ability of the 41,000 students to discuss topics in small groups, and that the technology for that feature indeed was not working. The officials also said that they were confident that fixes would be made in a short time period, and that the course would then continue.”

If you want an in-depth, first-person account of what happened in this course, you should read this blog post:

“There are three key factors contributing to this course calamity and all link to the group assignment. The first, a ‘technical glitch’ was big enough to cause one of Google’s servers to crash. Another, causing considerable distress to students is the lack of instructions for the assignments and the group activity—there was no clarity provided on the objective or purpose of the groups.”

Who could have guessed that to have thousands of people edit Google docs at the same time would not work out so well. And for once, you can break the Prime Directive of the Internet and read the comments from other students. They are very interesting.

There is a bit of schadenfreude about this when it started bouncing about the Internet. This was the failure we had been waiting for, hadn’t we? The one that would finally get us past the hype and get us a bit more realistic about the format and its possibilities.

But we’re not there yet. MOOCs are still riding high for now and Coursera’s fiasco may be to Udacity’s benefit as it is the platform that seems to have the most wind in its sails right now, especially with recent California deal:

“Now California state universities are set to begin enrolling students in MOOCs for credit. Earlier this month, the president of San Jose State University, Mo Qayoumi, announced that his institution will commence a pilot program: 300 students will receive course credit for online classes in remedial algebra, college algebra and statistics. Qayoumi was joined at the press conference by California Governor Jerry Brown and Sebastian Thrun, the controversial ex-Stanford prof and co-founder of Udacity, which will supply classes for the program at the cost of $150 per customer, er, student.

“This is the single cheapest way in the country to earn college credit,” Thrun “quipped.”

It’s not quite free, as early MOOC proponents began by promising. It is worth mentioning, too, that Udacity is a venture-funded startup, that classes will be supervised not by tenured profs but by Udacity employees, and that Thrun declined to tell the Times how much public money his company will be raking in for this pilot—or what more may have been promised should the pilot prove “successful.”

Okay, fine, but let’s get this straight: public money has been mercilessly hacked from California’s education budget for decades, so now we are to give public money, taxpayer money, to private, for-profit companies to take up the slack? Because that is exactly what is happening. Wouldn’t it make more sense to just fund education to the levels we had back when it was working?

Emphases mine, and good question at the end.

Because, let’s face it, the format is far from being the perfect model for education. First off,  again, from my limited experience, I see a lot more people in there for professional development than strict college education, and yes, a lot of people from developing countries. Also, the completion rates are still atrocious. Isn’t it insane to turn over a lot of money for a format that looks like it has potential but is far, far from proven to be effective.

It also seems that a lot of the course offerings are in maths, computer sciences, STEM more broadly. But there is little outside of the technical fields. Is it because these are easier to automate, with instant, automated grading? When I took the dataviz MOOC with Alberto Cairo, I don’t know how much time he and his assistant spent patrolling the message boards and reviewing projects, but they seemed very hands-on. Take a Udacity course, and you will be likely if you bump into an instructor in the forums. I am sure there are more Humanities / Literature courses out there, I would be curious to see if they just rely on peer-grading and discussion boards.

What bothers me, and that is why I highlighted it in the quote above, is that one could argue that the problems with MOOCs don’t matter because the courses are free (and therefore, you get what you pay for) and they don’t give credits… well, now they might. And the possible trend of pushing undergraduate education online through MOOCs is problematic to me on several levels:

  • it seems that then states abdicate their commitment (financial or otherwise) to public education.
  • It creates an additional form of inequalities: those with the means to do so get themselves an on-campus education and those who cannot just get what they can online (and I am willing to bet that quality control will be problematic because the point will be to save money).
  • MOOCs are, by definition, one-size-fit-all. This does not work for everybody. There is value to interactive education that the Udacity model cannot capture. MOOCs may represent another form of standardization rather than an innovative model. It is actually a very passive way of learning.
  • And again, what becomes of the latest obsession with retention / completion with MOOCs failing so badly at both?

But, in times of administrative bloat, one can see how the model would be attractive to administrators in search of cost savings.

And ultimately, when investors dump $15 million into Udacity, they will want something in return and that “something” will not be some fuzzy, idealistic, “free college education for all”.

Overall, I think there is still a lot that needs to happen to MOOCs for the format to be the real revolution that it is being touted to be. It is not. At least not right now. And I would not be so quick to bury the “old” university model. Every new technological innovation was once touted to be the death of the university from the early correspondence courses in early 20th century to online. None of this has happened. The real, serious fear would be for  short-sighted politicians and clueless administrators to use this as the obvious cost-cutter it seemed to be, but that would be at the expense of the mission of public education. That has to be fought at all costs.

One thought on “More On The MOOC Thing

  1. Pingback: Weekend Omnibus » Duck of Minerva

Leave a Reply

Your email address will not be published. Required fields are marked *