Friday, March 7, 2014

The Diversity of MOOCs and the Dangers of Over-Generalization

A lot of academic criticism of MOOCs derives from the fear that small, private online courses (SPOCs) open only to fee paying students, but based on MOOCs offered by “superstar” online professors, will replace local faculty, especially in lower tier institutions. Low-paid contingent faculty and teaching assistants (or worse, software programs and auto-graders) will then handle the heuristic teaching and grading, leaving fee paying students without expert human guidance. This is the "MOOCs will Destroy Higher Education as we Know It" argument. 
If you carry the hierarchical, every-student-for-themselves assumptions inherent in an xMOOC into the future you will never escape the reasons why so many caring educators oppose MOOCs in the first place. . . The problem with MOOCs isn’t the name. It’s not even the components of the acronym. The problem with MOOCs is that they’re being designed to create low-quality, hierarchical courses that can be championed by unscrupulous administrators to fire caring professors and leave unsuspecting students to fend entirely for themselves.
 Rees appears to be arguing here against SPOCs, not MOOCs. But let’s take up that part of his argument directed nominally against MOOCs. Rees says:
Cosmetic changes will not solve (MOOC quality) problems. Only re-thinking the entire xMOOC experience from the ground up will have even the slightest chance of creating something worthwhile.
The initial bother here is Rees’s false choice between merely “cosmetic” changes vs. “rethinking MOOCs from the ground up”. Rees’s dualism neglects that middle ground of actual or potential positive developments within the existing x-MOOC framework. Rees would almost certainly know about these if he took a break from general critique to explore particular MOOCs.  I’ll turn to this task below.
We can dismiss Rees’ unsupported claim about what MOOC designers intend.  Have x-MOOCs been “designed to create low quality courses”. On the contrary, their designers seek to create high quality courses, as they understand them, with instruction better than they perceive to be today’s norm. Indeed Anant Agarwal has made it the highest priority of edX to use the huge data sets available in computerized courses with tens or hundreds of thousands of students to improve instruction and learning. MOOCs may nonetheless be low quality courses, of course, but that is another question and would have to be addressed on the basis of relevant criteria of value.   
Have MOOCs been “designed to suit the needs of unscrupulous administrators?”  This ids another unsupported claim. Agarwal, Koller, Ng and the other platform executive have had quite different aims – to scale up instruction and make quality higher education globally available. While this may also result in substitution of technology for labor in the university, it hardly follows that that was the intended result, and in any case is not a critique of MOOC instruction itself. It is no critique of a hammer that an imbecile can use it to bash in someone’s head. It might nonetheless be a very good hammer.
Evaluating Particular MOOCs
So let’s consider the tasks of instruction in actual MOOCs. 
Instruction in courses typically involves three inter-acting dimensions: didactic, discursive and heuristic. The didactic dimension conveys knowledge through lectures and textbook readings; the discursive facilitates understanding through discussion and critical feedback; the heuristic shapes skills through drill and practice.  
In a geometry unit, for example, the didactic component may present basic concepts such as line, plane, axiom or proof; the discursive component may provide opportunities for discussion of e.g., proof strategies; the heuristic component may provide multiple opportunities to e.g., determine the area or perimeter of figures, to invent proofs, or to construct bi-sections of line segments.  In a literature unit, the didactic may convey background knowledge on the author, genre or period, the discursive may provide opportunities for guided interpretive and critical discussions of a selected story, and the heuristic may provide opportunities to use strategies of analysis, e.g., to locate the protagonist, the central conflict, and the dénouement.  
Good instruction, whether in a classroom or a MOOC, provides for the development and integration of knowledge, understanding and skill. It aims to engage students into worthwhile activities and practices by melding “I know,” and “I understand,” with “I can do.” When we think in general terms about the evaluation of particular MOOCs, we want to ask how well these instructional dimensions can be handled through computer software. 
Rees and other academic critics have condemned MOOCs as in effect using the best of 21st century technology to deliver the worst of 19th century instruction. The lectures, they claim, consist of videotaped talking heads. The discussion boards are useless and riddled with spam and flame wars. The problem sets are repetitive, trivial, and exhausting. In short, MOOCs are poor tools for didactic, discursive, and heuristic dimensions of instruction.
But is this critique justified? I will consider each of the three dimensions of instruction, with specific examples from well-known MOOCs.
(1) Didactic Instruction. While MOOC critics may wish to portray live classroom instruction as akin to Mark Hopkins dialoguing with a few students on a log, the reality is generally quite different: faculty members serving up uninspiring lectures to halls full of disengaged students. If talking heads are the problem, academic critics are in no position to pin the blame on MOOCs.
MOOC presentations, however, are not in general talking heads. On the contrary, MOOC presentations are an entirely new form of pedagogical experience. The large data sets and the rich data trails of computerized instruction have enabled researchers to explore new questions, such as the optimum length for lecture segments, and the most effective way of articulating them with readings and problem sets. Research suggests that 6 minutes is optimal. MOOCs now typically break up brief lecture segments with problem sets and free-response questions. Lecture segments are supplemented with interviews, clips from panels at research conferences, interviews with other scholars, research presentations by graduate students, and on-site videos of knowledge use in ‘real-world’ settings. While a single instructor inevitably gives his course subject matters a personal twist, MOOCs offer unique opportunities for presenting multiple points of view.
JonathanHaber, who constructed a complete college education in the humanities– in asingle year - from MOOCs, and thus knows as much about actual MOOCs as any commentator, states:  
MOOCs are often criticized for just transferring a “sage on stage” pedagogy from the lectern to the computer screen, scaling up the worst aspects of oversized lecture classes. But as my year of MOOCs went on, I saw a new visual language developing, as single talking heads were supplemented (or replaced entirely) with conversations among colleagues (the visual style of one of my favorite courses: HarvardX’s   "The Ancient Greek Hero") interviews with experts, on-location shots, and even on-screen performances. Such creativity helped to make lectures one of the most engaging and, ironically, intimate components of massive online courses, while also raising the bar for all other forms of online learning (most of it far duller than your average MOOC).
At their best, MOOC presentations are as engaging as the best television news magazine programs. And this is not surprising, as the universities providing the best MOOCs are investing in cutting-edge production studios and hiring media professionals. The best MOOC presentations are more like segments of “60 minutes” than talking heads.  We may question whether producing “edutainment” of this sort is the best use of scarce institutional funds, but that is another question entirely, and one that completely undercuts the “talking head” line of attack.
(2) Discussion and Feedback. Let’s acknowledge that the MOOC discussion boards are far from adequate substitutes for live discussion.  Even some of the most supportive MOOC commentators have found them to have little value. The good news is that the discursive segments of many MOOCs have moved well beyond the discussion boards to live conversations in physical or virtual spaces.  
Coursera has sponsored “meet-ups” – where students in their MOOCs can get together for course discussion - in dozens of cities. Coursera has also teamed up with the U. S. State Department to run MOOC camps for college age students in many countries. The city of Boston has teamed up with edX to initiate Boston-x, a project providing Internet computers and meeting places for MOOC learners. Other cities have emulated Boston-x, dedicating space in libraries and other public buildings for MOOC study. These efforts parallel those of the Library 2.0 movement; librarians around the world are now re-thinking optimal uses for their brick and mortar building spaces in the age of web 2.0 technologies and digital books, and one answer is meeting spaces for online learners.  C-MOOCs, and now many x-MOOCs well, make extensive use of video-conferencing – e.g., via Skype - to connect group members. MOOCs offering nothing more than discussion boards are simply behind the MOOC curve.
(3) Heuristic Instruction. Academic critics claim that the “can do” element of MOOCs is restricted to monitoring auto-or- peer graded problem sets.
Many actual MOOCs, however, have moved way beyond such mechanized problem sets, making project-based authentic learning central to the MOOC experience. Here are two examples:
(1) At the University of Virginia’s Darden School of Business, Professor Michael Lenox has offered several iterations of his “Foundations of Business Strategy” MOOC on the Coursera platform. Lenox uses “Coursolve,” a crowdsourcing software program, to connect his course with partner organizations where students work to solve real-life challenges. He says:
 “Entrepreneurs don’t always have the resources to hire external support to address their needs, but we’ve seen firsthand that students are hungry for the chance to apply their knowledge to real-world problems,” Lenox says.  “By collaborating with organizations, students can strengthen their skills development while potentially providing businesses and nonprofits with valuable insights.”
                In the final course project, Lenox invited students to undertake complete a strategic analysis of one of the partner organizations. More than 400 students completed analyses in partnership with 100 different organizations, including established businesses, startups, resource-strapped nonprofits and social enterprises. 78% of those students ended up participating directly with senior managers in the strategic decision-making of their organizations. Reporting on the second iteration of the MOOC, Lenox added,Hundreds of in-person study groups formed in over 50 countries. Students included young entrepreneurs and mature small business owners; non-profit organizers; a study group of Mongolia students led by a Peace Corps volunteer; a group of  unemployed women in Ohio looking to improve their job prospects; a group of students ijn Bolivia led under a program from the U.S. State Department; and a group of Arab and Israeli students participating through the YaLa Young Leaders program building détente through education.”
(2) Cathy Davidson of Duke University is now offering a Coursera MOOC on 'thehistory and future of higher education'. Davidson has been a national leader in pushing the x-MOOC format in creative directions. Because her MOOC has many thousands of students, the student group can take on projects not possible within a classroom context. In one project, her students are collectively creating a rich, multi-media trans-national timeline of higher education since 1800. Each student is contributing reports on significant historical events in higher education in their geographic locations - countries, states, municipalities. The students are learning historical research methods and skills in reporting historical events. Many higher education institutions previously neglected by historians of education, including those long closed, are in this way being entered for the first time in an accessible historical record. The student group is now collaborating on editing and coordinating the information and producing the final online product. [5]
One does not need to love either of these particular skill development efforts to recognize that they go way beyond auto-graded problem sets.
6. Conclusion
My point in the above remarks has been to show that academic critics of MOOCs have relied on a stick-figure caricature. Real MOOCs, even x-MOOCs, are diverse, and many MOOC leaders have addressed the didactic, discursive and heuristic dimensions of instruction in creative ways. The best MOOCs replace the ‘sage on the stage’ stalking head with presentations employing a “new visual language” of instruction; they build in group experiences, whether physical or virtual, with opportunities for interpersonal student dialogue; they make project-based learning the centerpiece of the educational experience.
Of course, not all MOOCs do this. Some do, some don’t. And that is precisely the point I am making. If we want to assess the pedagogical value of MOOCs, we will have to turn our attention to particular courses to see how they handle instructional tasks. The most important dimensions of instruction, the didactic, discursive and heuristic, provide useful pegs upon which to hang such particular evaluative judgments.