As reported in a special report in The Signpost last month ("WikiEd course leads to Twitter harassment"), things went very wrong in December 2021 in a class supported by Wiki Education's Wikipedia Student Program. I came away from the ensuing discussion understanding that there's a gap between what we do and the community's understanding of what we do. Our program is the single largest outreach effort in the Wikimedia community. In 2021 we brought over 12,000 student editors to Wikipedia, who added over 9.6 million words to 12,000 existing articles and created 1,000 new articles.

It's human nature to make judgements based on extreme cases, both good and bad. And once we’ve found these patterns, confirmation bias does the rest. If you only come across outstanding student work — the student who revamps a mess into a high quality summary, or the student who creates a new article — you come away thinking students' work is exceptional. And when you read about the kind of incident stemming from our program that was highlighted in January's Signpost, you can get the impression that the results of student editing are (as one editor said in the discussion) "decisively mixed". My experience, as a Wikipedian since 2004 who supports the 12,000+ student editors writing on Wikipedia each year through Wiki Education’s programs, is different. I see almost everything students do.

Who we are

I'm writing this as User:Guettarda, a longtime contributor to English Wikipedia. But I'm also writing this as User:Ian (Wiki Ed), the Senior Wikipedia Expert at Wiki Education. Wiki Education is a US-based nonprofit bridging the gap between Wikipedia and academia. Our biggest program, the Wikipedia Student Program, supports 12,000+ student editors a year, all of whom attend a higher education institution in the United States or Canada.

Our organization has its origins in an effort to increase the pool of Wikipedia contributors, and to improve article quality. This project has grown from an initial 18-month pilot in 2010, involving 24 universities and over 800 students, into our current program: in Fall 2021 we supported 5,972 student editors, who contributed 4.7 million words to over 6,400 articles. Since 2010, more than 102,000 students have added 85 million words to 116,000 articles. Our program is the single largest outreach effort in the Wikimedia community by count of new editors recruited, articles edited, and content added. Moreover, almost all of this content passes Wikimedia editorial review and remains in Wikipedia.

Both in terms of bringing in new editors, and in terms of bringing in high quality content, Wiki Education has been very successful.

The content students edit

In the course of normal editing, you encounter the work of our student editors quite regularly. They create new articles about minerals. They expand existing articles about environmental science. They write about music of the world, and African archaeology, and poetry. They write about deafness in various countries. Wiki Education's courses cover nearly every discipline taught in higher education. Our program for university partnerships is one of the Wikimedia movement's biggest success stories for a way to systematically improve content areas.

Students add a lot of scholarly citations to Wikipedia. A 2020 paper by researchers Jiro Kikkawa, Masao Takaku, and Fuyuki Yoshikane reported that in 2016 15.5% of the editors who added scholarly citations to English Wikipedia articles were participants in Wiki Education-supported programs. Citations matter because they're the way Wikipedia articles achieve credibility despite being written by pseudonymous editors. And thanks to university libraries, students have easy access to a wealth of scholarly work that might be difficult for most readers (and many Wikipedia editors) to get access to.

In 2016, students in George Waldbusser's Biogeochemical Earth class transformed a redirect into an article about the Boring Billion, an approximately billion-year period in the Earth's history where very little happened. In 2020, another Wikipedian took the raw material that the students had created and converted it into the Good Article that exists today. It's impossible to say whether they would have created this article from scratch if they didn't have the student work to improve.

But the idea that student work creates an impetus for article improvement is consistent with what Kai Zhu, Dylan Walker and Lev Muchnik found in their 2020 study of the impacts of student editing. Using data from Wiki Education's Dashboard, they were able to trace the fate of 3,300 articles that were edited by students in Fall 2016, and compare them with a control group that students hadn't touched. It turned out that after the students had finished editing, the articles they worked on had 12% more page views than the control set. This also translated into more page views downstream, in the articles linked from these ones, and more edits from other Wikipedians. As I wrote in a 2020 blog post:

At the start of 2016 the article about Kalief Browder — an African American boy who was imprisoned in Riker's Island for three years without trial for the crime of allegedly stealing a backpack — was about 380 words long. In spring term that year, a student editor in Fabian Neuner's Black Lives and Deaths class worked on Browder's biography and expanded it to more than 5,500 words. The difference in this case is that the article has continued to attract edits since the class ended. In the remainder of 2016, after the class ended, the article received 135 edits, almost precisely the same number of edits (133) as it had received in the previous two years. Not only that, the article has continued to develop as other editors have tweaked and modified the article. At the end of the class, 82% of the article was the work of one student editor; today it's only 53%. The article about the short life and tragic death of Kalief Browder hasn't gotten much longer, but its overall quality has continued to improve.

Because students tend not to stick around beyond the duration of their class, it's easy to think of their impacts as one-off. But instructors do tend to stick around, and over the course of many classes, an instructor can make a real impact on a topic area.

Since 2013, Erik Herzog has included the Wikipedia assignment in his chronobiology class, which runs every other Spring. When chronobiologists Jeffrey C. Hall, Michael Rosbash, and Michael W. Young were awarded the 2017 Nobel Prize in Physiology or Medicine, the world rushed to Wikipedia to discover who these people were. And Wikipedia's biographies of these three scientists existed, for the most part, because of the work that was done by students in these classes.

Similarly, since 2012 Joan Strassmann's behavioral ecology classes have added well over 1.8 million words to Wikipedia in articles about bees, wasps, spiders, and flies. Her sister Diana Strassmann's classes have added over 1.3 million words to Wikipedia, primarily in the areas of poverty, justice and human capability. Last term they created two new articles: Medical racism in the United States and Anti-Apartheid movement in the United States. Other students successfully contributed to articles like Discrimination based on skin color and Racial disparities in the COVID-19 pandemic in the United States. Past classes have made edits like this one to the Discrimination against drug addicts article and this one to the Mental health and immigration detention article. Both classes have had students stick around after the end of the class to take their work through the Good Article process.

Who are the student editors?

As of 2020, Wiki Education brings in 19% of new active editors on the English Wikipedia. Student editors in programs supported by Wiki Education make up about 3% of all active editors on the English Wikipedia. In other words, the interactions you had with a student editor, or the work you read (good or bad), isn't necessarily representative of the whole, just as an interaction with any editor isn’t representative of the community as a whole. The scale at which our program operates is massive.

The community has long been aware of the issue of systemic bias in Wikipedia, and the fact that this is a problem that is exacerbated by the composition of the community. According to WMF's survey data, only 0.5% of Wikipedia editors in the US are Black or African American, while 8% of our student editors identify as Black or African American. Only 5.2% of editors in the US are Hispanic, while 12% of our students are. And while only 22% of Wikipedia contributors are women, 67% of our student editors identify as women (and another 3% identify as non-binary or other).

The relatively narrow demographic base of our contributors is part of the reason why large areas of content are under-represented. The issue of the coverage of women on Wikipedia is quite well known, but this is also true for members of minoritized groups and for areas in the Global South. Even within the United States, the contributions of students from an Appalachian State University class on The History of Coal were eye-opening to me: it was apparent from their writing (and from the emails I exchanged with some of them) that they had a familiarity with this topic that most editors lack. In particular, their creation of the Broad form deed article filled an important lacuna in Wikipedia's coverage, but one that may not jump out at you if you don’t live in an area where strip-mining regularly occurred on land occupied by people who only owned the surface rights.

Adding contributors who are outside of the normal demographics of Wikipedia contributors may not be able to fix the problem of systemic bias that stems from what McDowell and Vetter have called Wikipedia’s "logocentric reliance on written knowledge", but it is a valuable part of the community’s overall toolkit to improve this problem.

So how do we do it?

How do you support 6,000 students at once without creating a massive free-for-all where students are struggling, failing, and generally taxing the resources of the community? And, it’s worth asking, are we doing that?

Our model involves supporting instructors in higher education in the United States and Canada who take their students through the "Wikipedia assignment". After going through an online orientation, our Assignment Design Wizard takes instructors through a process that uses their input to create a timeline for the class. The instructor then submits their course page for approval. My colleague Helaine Blumenthal goes through each submission, and makes sure that they are suitable. Since instructors can customize the timelines, Helaine ensures that the key components are all there, and that the class is following our best practices developed since 2010.

Students create a Wikipedia account and sign up on the Wiki Education Dashboard After this, students are prompted to take several training modules and go through a series of exercises. After this, students begin the process of drafting an article. This involves selecting a few candidate articles (students are encouraged to select stub- or start-class articles, and stay away from GAs and FAs). Once they've narrowed their choices down to one topic, the Dashboard takes them through a step-by-step process in which they build a bibliography, draft their contribution, peer-review one-another’s work, and finally move their work to mainspace.

People being people, problems occur. The Dashboard monitors a lot of what students do, and sends a whole range of alerts. If students are behind on the trainings, the Dashboard will email the instructor and ask them to remind their students to get up to date. If students assign themselves FAs or GAs, the Dashboard emails them and their instructors and strongly encourages them to pick a more suitable article. If they assign themselves an article subject to discretionary sanctions, I am notified. If they edit GAs or FAs, I am notified. If they move their sandbox to mainspace, I am notified. And most importantly, if their work is flagged as a potential copyvio, I get notified (as do their instructors). These are just some of the things we do; we’re constantly looking for new and better ways to monitor student work, and to catch problems before they become problems. Also, the help of community members who flag me when things go wrong is immensely helpful.

Beyond this, students are encouraged to get in touch if they have questions or run into problems. We provide multiple ways for them to get in touch, either on-wiki or through the Dashboard. We also encourage instructors to get in touch with us, to ask questions and to relay student questions. In addition, the Dashboard lets me monitor classes and check in on what students in any class are up to. We have things set up to pay additional attention to classes working in areas where they may run into problems. This system enables us to head off potential problems, intervene early where we can, and be notified when something goes wrong. Of course, things do go wrong sometimes — and these are usually the ones you hear about — but for the vast majority of the 12,000+ students editing every year, things go well, and Wikipedia gets high-quality information added.

As an individual, I can't imagine anything I do would have a bigger impact on the world than my contributions to Wikipedia. Health permitting, I might be able to contribute to Wikipedia for another 30 years. But every term working with student editors I make an impact that’s orders of magnitude greater than what I can do on my own. Working with student editors allows me to help add content that would otherwise take decades to be added. This is why what we're doing matters to me.