Friday, March 23, 2018

Higher ed's grudge against MOOCs -- Part 3

Last update: Saturday 3/24/18
This third post concludes the discussion I began in Part 1 and continued in Part 2. It summarizes the ten online degrees offered by Coursera's partners. Then it offers a plausible scenario wherein the courses in Coursera's online degree programs could contribute to the disruption of the lion's share American higher ed that still needs to be disrupted for the benefit of most American students.
H. Coursera's online degree programs 
The following table displays information about the online degree programs offered by Coursera's four partners that are American universities:
  • All of these programs assume that students are enrolled for less than full-time (9 or more credits per semester); so most students will take at least two years (four semesters) to complete the programs
     
  • Data for Coursera's programs came from pages on Coursera's Website; data for Georgia Tech's computer science masters came from a page Georgia Tech's Website. (The names of the programs are hyperlinked to the referenced pages)
     
  • Four American universities account for seven of Coursera's ten programs.
     
  • Tuition and other data was only available for the University of Illinois (four degrees) and the Arizona State University (one degree)
     
  • Tuition was not available for the University of Michigan (two degrees)
     
  • Coursera did not provide data about the fees students would pay in addition to tuition. One must therefore assume that students would pay at least $300 in fees for each of the four of more semesters they are enrolled. Therefor the real cost to students will be the tuition that appears in the table plus at least $1200 in fees
     
  • Georgia Tech's Website provides information about tuition ($5,100) and fees ($301 per semester). Assuming most students took 7 semesters, the total cost would be about $7,200 -- the figure cited most often in the media 

University
Degree
Tuition



Georgia Institute of Technology
$5,100



University of Illinois at Urbana-Champaign
$19,200



University of Illinois at Urbana-Champaign
$22,000



University of Illinois at Urbana-Champaign
$30,000



University of Illinois at Urbana-Champaign
$19,200



Arizona State University
$15,000



University of Michigan
n/a



University of Michigan
n/a

The tuition charged by Coursera's partners for their online degrees may be higher or lower than what they charge for their on-campus programs. No matter. As per the preceding table, the tuition charged by the five American programs for which data is currently available is high. Compared to what? Compared to the $5100 tuition charged by the Georgia Institute of Technology for its high quality masters degree in computer science that was developed by its faculty and Udacity. Georgia Tech set the gold standard for affordability, and its affordability has enabled its enrollment to attain impressive size and diversity.

To those who are still asking, hopefully, whether other top U.S. universities will follow Georgia Tech's pioneering trail by offering their own affordable online programs, I say good luck with that. Georgia Tech's program has been in operation since 2014 and its success has been well publicized within the academic community. If other top universities were inclined to follow its lead  they have had four years to do so. They haven't. So the question that should be asked asked is, "Why not?"

I. Disrupt what really needs disrupting
I wanted Coursera to pick up the low tuition pennant that Udacity had cast aside. Not getting what I wanted, my first reaction was disappointment. My mistake. 
  • Somehow I had forgotten that most elite universities, Coursera's preferred partners, are loathe to reduce their tuition because they know that most Americans associate high quality with high prices. If a university charges low prices, it must not be providing high quality education, right? On the other hand, most of the students who attend high priced elite universities can afford the high prices. So the status quo works for those for whom it was designed to serve; so it ain't broke; so it don't need fixing, right?
     
  • However it's also undeniably true that, aside from the top five percent of America's universities that serve the most affluent five percent of America's students, the rest of higher education is becoming increasingly unaffordable for the less affluent 95 percent of America's students. 
After taking a long second look at what Coursera created, my disappointment vanished when it suddenly occurred to me that Coursera's pivot may have repositioned it to become a primary facilitator of the disruption of the 95 percent of U.S. higher education that really needs disrupting. Talk about irony!!! Please consider the following scenario ... :-)

J. Licensing course packages and electives
Udacity was the first MOOC provider to implement the obvious insight that the fundamentals of any significant set of job-related skills cannot be covered by a single course. So Udacity produced certificate programs that were packages of related courses. The courses in a package covered the basic skills that students seeking entry level positions in an employment sector were expected to possess. Then Coursera, edX, DataCamp, and other MOOC providers produced their own packages.

This insight was obvious because it's the basis for traditional curriculum design. Identify a set of fundamental concepts or skills, then allocate their presentation across a carefully sequenced set of core courses. Thereafter students can take related electives, i.e., courses in specialized topics. Here's a couple of familiar examples:
  • American History is often presented in two semesters: colonial times to the Civil War, then Reconstruction to modern times.
     
  • Foreign languages are often taught in two or three one-year sequences, e.g., French I (Fall and Spring), II (Fall and Spring), and (maybe) III. By the end of the last year students are expected to be able to read French literature and converse fluently in French.
When MOOCs were first evangelized back in 2012, a number of observers (including yours truly) suggested that elite universities should produce MOOCs that other universities could license, the same way that the faculty of elite universities often authored the textbooks that other universities adopted for their courses. Twenty-twenty hindsight quickly revealed that this was a "bad" idea. More optimistically, a "good" idea whose time had not yet come. There were many reasons. Here's three:
  • In 2012/13, elite universities were producing one-offs, one MOOC over here about this topic, another over there about another topic; whereas curriculum committees needed packages of core courses and related electives.
     
  • MOOCs were new enough that a department chair could not determine the commitment of the authors of a MOOC. Would the authors maintain the MOOC for several years, correcting errors asap, updating content from time to time, etc, etc, etc? Or would this courseware become "abandonware"?
     
  • Then there was the mistaken notion that MOOCs would replace local instructors. The concept of flipped courses was beginning to catch on, but not enough for enough department chairs to see MOOCs as providing the videos and texts that students could be tasked to watch and read before attending interactive classroom sessions the next day.
An increasing number of Coursera's partners are now offering online degrees because an increasing number have developed packages of core courses and related electives. No more one-offs. It goes without saying that the developer universities will maintain their courses, i.e., correcting errors and updating content as required. Nowadays the concept of flipped courses is widely accepted (albeit not universally). Therefore the pool of potential clients who might license Coursera's core courses and electives is large and growing.

Coursera's "app store"
It seems natural for Coursera to host its partners' courses in Coursera's cloud, then broker the licensing and collection of license fees in a manner akin to the way that Apple and Google broker the licensing and collection of fees for apps in the App Store and Google Play, respectively. Students would access the courses via secure logins. Prospective subscriber colleges and universities could try the core courses and electives for a limited number of their students at first, then expand to volume licensing (with appropriate discounts) if the trials produced satisfactory results.

Coursera would distribute most of the fees to the universities that developed the courses, keeping an agreed upon percentage for itself. These licensing fees could generate substantial income for Coursera's partners and for Coursera.

Lifelong learning that's affordable for more non-traditional students
Why should other colleges and universities license the courses developed by Coursera's partners? I just mentioned flipped courses and implied their use in the on-campus courses that are mostly for traditional college students. 

But traditional students are becoming a smaller and smaller segment of the academic market. As machine learning and other components of AI disrupt more and more careers, more and more employees will have to go back to school from time to time throughout their careers. These older, non-traditional students have become the primary market that most colleges and universities will have to serve ... or close their doors if they fail to serve.

The vast majority of non-traditional students will not be able to take courses on campus; they will require high quality online courses ... but high quality online courses are expensive to develop and maintain ... but more affordable if licensed and therefore more affordable for students enrolled in the colleges and universities that become Coursera's subscribers.

Caveat: It must always be remembered that the licensed online videos and texts  will only provide part of a subscribing institution's student learning experience. Their own instructors will be needed to assess students' term papers and other projects, to facilitate online discussions among students about course materials, to answer questions about points in the online videos and texts that students did not understand, to be available for occasional one-on-one online chats with students about their personal interests in a course, and to determine the most appropriate grades, based on the subscribing institution's grading metrics, for students who complete a course.

Learning science research
Finally, I suggest that Coursera set up an affiliated, but independent non-profit institute that would be charged with conducting research on the massive stores of data that will be generated by students enrolled in courses offered by Coursera's partners and by its licensed subscribers. This institute should conduct studies designed to advance learning science using anonymized versions of the raw student data collected via Coursera's platform. Its findings would enhance the learning experiences of all students everywhere. 

I suggest this kind of independent non-profit entity for two reasons: first, a non-profit institute would be eligible to receive more grants from foundations and government agencies; and second, we are now living in the darkening shadows of the scandalous misuses of the data collected by Facebook, Google, Twitter, and other profit-oriented social media about their users. 

Readers may recall that early on edX announced its intentions to conduct similar research. Being a non-profit entity should assure that edX will be less tempted to misuse student data than a profit-oriented entity like Coursera. However, as I understand it, few if any of the for-credit courses in the programs developed by edX partners are licensed for use in for-credit programs offered by other institutions at this time. By licensing its courses to non-partners, Coursera could generate far more data than edX. More data should trigger keener insights into the learning processes of different kinds of students -- male vs. female, minority vs. majority, older vs. younger --  when subjected to the powerful, but data hungry algorithms of machine learning.


Roy L Beasley, PhD
DLL Editor

P.S. Coursera's pivot to becoming an online program manager (OPM) for online degrees does not imply that it will stop offering non-credit, low cost certificate programs. To the contrary, these programs will continue to enable faculty at elite universities to try their hands at online teaching in low key environments. Those whose experience is satisfactory will be more willing to participate in Coursera's online degree programs, programs from which the lion's share of Coursera's future earnings will come.