Pages Navigation Menu

Extracurriculars: Do Massive Courses Make Digital Sharecropping More Efficient?

Ever since Colorado State University – Pueblo Professor Jonathan Rees brought up the term “digital sharecropping” in a Twitter response to my earlier discussion of peer assessment in MOOCs, I’ve been watching how student work is being used outside the MOOC. While the term digital sharecropping is relatively new, the practice of using students to do real work has been around for ages. So, it makes sense to illustrate the changing nature of employment and how MOOC students are engaging in work while in class.

MOOC crowdsourcing

James Cridland via Flickr

Let’s start by looking at who works for internet companies. Mary Meeker of the VC firm Kleiner Perkins Caufield & Byers puts out yearly slide decks that serve as benchmarks for internet trends. On slide #6 of the 2013 deck she lists the leading consumer internet properties by how many individuals use these services on a monthly basis. The numbers mirror what is happening with MOOCs: they show international reach and massive size. She points out that while there are 244 million internet users in the U.S., Google alone receives over 1 billion visitors each month.

That can be seen as a form of digital sharecropping. Although Google officially employees some 50,000 workers, this is misleading. Professor Rees and I would agree that Google actually employs the billion monthly visitors to make their service operate with amazing efficiency. The short queries we all use when searching Google is the content we provide about ourselves. Google uses this content along with our location, our interests and our social interactions to deliver a short list of the most useful items to display back to us. Google then sees which item is chosen from each short list and uses that action to understand each user’s preference.

Taken in aggregate, these billions of searches each month provide Google with free labor to generate information it can sell to advertisers, who themselves provide a lot of effort and information that Google benefits from.

Google doesn’t pay most of their employee/users, but in return for their contributions, most Google employee/users receive rapid search results, accurate translations and maps that highlight popular locations. This example of digital sharecropping applies to each of the massive websites listed by Meeker: Microsoft, Facebook, Yahoo and Amazon all rely on user behaviors and content, which they monetize through ads or transaction fees.

The “customer as employee” has many earlier examples. McDonald’s and Starbucks have trained customers to help them eliminate staff by ordering while standing in line, picking up their orders, delivering them to their tables and cleaning up after themselves. In similar circumstances all over the world, customers have shown they often prefer self-service and the benefits of quick and predictable results.

In academics, the use of students has long been a mainstay for faculty research. At many universities, student participation in research projects can generate course credits that count toward their major. Most scientific journals accept work done by students as valid, even to the point of generating the term “student guinea pig.” In return, faculty promotion within the university is tied to the research funding they bring to their department.

With this as background, it’s not much of a leap to imagine employment of the massive online student population.

So, I was intrigued to receive a notice near the end of Dan Ariely’s business psychology MOOC, A Beginner’s Guide to Irrational Behavior from Coursera, proposing a crowdsourced essay:

To take advantage of the thousands of bright minds in this class, we propose a collaborative essay that will show how the concepts studied in this class could inform the student experience.

*Note that this project is about applying the themes of behavioral economics to your observations and should be directed at a general audience that is unfamiliar with the specifics of this class. (In other words, we’re not looking for suggestions about how to improve the course in this particular project.)

From this modest starting point, a process was developed using Google Docs (more data for Google), working groups were formed and student-contributed content is being generated. I applaud Ariely for making his process as open-ended as possible, even if the result yields something of limited viability. While there is no indication this project is sponsored research, his experiment has value in many domains. How many people would participate and at what levels? What methods or objectives would produce the best work and what goals would students come up with for the essay? It’s a process that deserves close inspection.

An interested observer would note how much time was taken to encourage and then gather student brainstorms. Course forum posts show a wide range of approaches suggested by students and a need to review them all. This process was not automated and yielded multiple calls for volunteers to help as organizers, editors and writers. Eventually, students surveyed their cohorts, gathered the brainstormed contributions and developed a plan. A small group of engaged students has given themselves the tasks of compiling all ideas and sifting them against the purpose of the essay to discover an essay structure. An outline will be developed and evaluated to then set expectations and develop instructions for writing and editing.

While the Ariely experiment might seem rudimentary and manual, I have never before seen a project like it done at this scale. Just as it took McDonald’s and Starbucks years to refine the customer engagement practices they use in their stores, Coursera and other MOOCs are developing an idea of how students want to be recognized beyond “taking the course” alone.

This new-found MOOC student-labor pool is also the focus of other experiments. What if students could engage in projects for outside organizations from within the course structure? A look at Coursolve, for instance, points to two other Coursera classes. Foundations of Business Strategy, in addition to the customary case studies of such classes, assigned students to projects tackling real-world problem in partnership with outside organizations, including businesses and nonprofits. Introduction to Data Science contains an “Optional Real-World Project” page for students to work on real external data sets.

Some have called the kind of work we do for Twitter, Facebook and Google “play labor”. The potential for learning environments like MOOCs to coordinate student efforts might be the tipping point when “sharing goes pro”.

John Duhring (7 Posts)

John Duhring has been s a founding team member at nine startups, including Supermac Software and Bitmenu. During his career he has also applied technology to learning at large companies such as Prentice-Hall, Apple and AOL. Follow him on Twitter @duhring.


One Comment

  1. The following interview with Duolingo founder presents another angle on mass crowdsourced services focused through online courses.

    “One of Duolingo’s most interesting aspects is its business model. As part of their learning process, users are given the option to collaborate with each other to produce crowdsourced translations. Duolingo claims that their quality can rival with the work of professional translators, and and made no mystery that it was planning to sell them to third parties.”

    http://thenextweb.com/la/2013/06/09/duolingo-founder-and-captcha-creator-luis-von-ahn-talks-android-crowdsourcing-and-ab-testing-interview/