by David Gries, JT Huang, AJ Renold
In Commons-based Peer Production and Virtue, Benkler and Nissenbaum note how the internet has given birth to an era of decentralized collaboration “among large groups of individuals, sometimes in the order of tens or even hundreds of thousands.” These groups cooperate effectively to, “provide information, knowledge, or cultural goods.” In this blog entry, we briefly examine four case studies of peer production: 1) Mechanical Turk 2) Wikipedia 3) HealthTap 4) Khan Academy. We found it useful to conceptualize these collaborative communities along two axes: 1) the nature of the work and 2) motivation of the contributors. Drawing from Haythornwaite, we distinguish the nature of the work as “Lightweight” vs. “Heavyweight”; we characterize the motivation of contributors as market-driven vs. other more nuanced social/individual motivators. (1)
Amazon Mechanical Turk more or less defines Haythornwaite’s concept of “Lightweight” peer production. Turkers are relatively anonymous and independent contributors supporting the goals of a project organizer. The work of Turkers is highly modularized, and the job of coordination and delegation is at the level of the project organizer. Essentially, Turkers are not much more than ‘cogs in the wheel’. Their work is usually limited to tasks that slightly outstrip the ability of a computer, such as recognizing pornographic images.
In terms of the motivation, it is clear that Turkers are mostly market-driven actors. Turkers do not seem to care much whether they are adding contributing to overall societal welfare–so long as they are getting paid. Panos Ipeirotis, an NYU professor, noted as many as 40% of delegated tasks on MTurk are to help generate spam that can stay one step ahead of major mail client spam monitors (2)
Wikipedia, in contrast to MTurk, largely defines Haythornwaite’s concept of “Heavyweight” peer production.There is no centralized delegating authority on Wikipedia. Rather, nothing happens without consensus. In addition, its contributors must possess a great level of domain knowledge, procedural knowledge, and expertise in order to contribute effectively. Further, Wikipedians spend relatively greater amounts of time and energy to create their product than people on MTurk.
Wikipedians volunteer their time and knowledge for no direct monetary reward, so their reasons for contributing must necessarily be rooted in other, more nuanced motivations. The paper “What motivates Wikipedians?” lists several motivations that Wikipedians may have: Protective (eg. “By writing/editing in Wikipedia I feel less lonely.”), Values (eg. “I feel it is important to help others.”), Social (eg. “People I’m close to want me to write/edit in Wikipedia.”), Understanding (eg. “Writing/editing in Wikipedia allows me to gain a new perspective on things.”), Enhancement (eg. “Writing/editing in Wikipedia makes me feel needed.”), Fun (eg. “Writing/editing in Wikipedia is fun.”), Ideology (eg. “I think information should be free.”) (3)
Healthtap is a peer produced source of medical information authored by doctors. Doctors publicly author articles and vote for consensus on articles relating to their area of medical expertise (4). Like Wikipedia, the nature of the work is inherently heavyweight. Consensus is crucial, as misinformation is literally a matter of life and death. However, Healthap is lightweight insofar as its questions are modularized to the extent that they can be answered by a single doctor.
Like Wikipedia, doctors are not paid for their contributions. Yet unlike Wikipedia, it is not anonymous. Granted, Wikipedia is not completely anonymous either, but contributor identities are not mapped to their real identities. Therefore, the consequences for bungling one’s job on Wikipedia do not neccesarily translate to one’s livelihood in real life. If a doctor errs on HealthTap, it can directly harm the doctor’s real life reputation. This hints at one of the reasons for doctors to contribute to HealthTap; doctors gain (or lose) reputation on the site, which translates into real world reputation and can help potential patients discover (or avoid) their practice.
Khan Academy (KA) is best known as an online educational platform. (5) Often overlooked is the fact that its mathematical exercises are the result of a peer production process. (6) Creating exercises is a unique combination of heavy and light weight peer production. On the one hand, work is delegated to contributors by an authoritative body that identifies exercises that are needed or requested by KA users. This type of workflow is characteristic of lightweight peer production, similar to that of MTurk.
At the same time, KA contributors must possess a high level of skill to successfully contribute to the site. KA contributors are not simply sorting through porn or writing spam messages. Rather, their work demands multi-faceted domain knowledge that is more characteristic of heavyweight peer production like one would find on Wikipedia. KA contributors must learn the programming framework used by KA to build exercises. And on top of that, they must also possess significant subject matter expertise and pedagogical know-how to create questions that possess pedagogical value.
In terms of motivation, KA contributors align more closely to the wikipedia model. That said, some contributors are motivated by contribution visibility. Interestingly, KA runs its exercise creation process via Github. Thus, even as contributors commit to KA’s repository, they also increase their activity visibility on Github as a whole. In the tech sector, a person’s Github activity is often used to signal technical skill. (7) Thus, some KA contributors might be taking the long view hoping to pad their resume in the hopes of eventually landing a job. With that said, it is certainly arguable that KA users are more socially minded than their spam-producing Turker counterparts.
These four examples illustrate how different types of peer production–lightweight and heavyweight–draw contributors across a spectrum of motives. The interaction between the nature of the work and the motivation of contributors creates fascinating dynamics in each instance. Any one of them merits further study, which may be helpful to keep in mind as we continue to work on our final papers.
(1) Haythornthwaite, C. (2009) Crowds and Communities: Light and Heavyweight Models of Peer Production. In Proceedings of the 42nd Hawaii International Conference on System Sciences. Los Alamitos, CA: IEEE Computer Society.
(3) “What Motivates Wikipedians?” http://pensivepuffin.com/dwmcphd/syllabi/info447_wi12/readings/wk02-IntroToWikipedia/nov.WikipediaMotivations.CACM.pdf
(7) “Kristinas branch” – https://github.com/Khan/khan-exercises/pull/43449