Rethinking Wiki engagement in medical research: insights from a residency at NIHR

Wikimedia residencies most often take place in GLAM institutions with the goal of enriching Commons. These are vital for free culture. However, residencies focusing on  Wikipedia have received far less attention. As a result, there is less community knowledge, fewer tools, and fewer conversations about best practices. 

In this article, I would like to present a case study of my residency at the National Institute for Health and Care Research (NIHR), a major research funder in the UK. Rather than presenting a checklist of accomplishments¹, I want to share some of the insights into the challenges faced, lessons learned, and reflect on what worked (and what didn’t). My hope is that some aspects might resonate with the experiences of others, and some ideas prove to be helpful and applicable for current and future host institutions and their residents. 

Health information on Wikipedia

To begin with, let’s look at how Wikipedia measures in terms of information about illness and health. The English language Wikipedia is one of the most widely used resources for health information. Each month entries on illnesses, medicines, and associated topics are read more than 100 million times.² When a pandemic breaks out or a celebrity is diagnosed with a condition, we can clearly see how a great number of people turn to Wikipedia to understand what’s going on. It is crucial that what they find there is reliable, reflects our most current knowledge and that is written in a language that allows them to understand the most important things to know. Although health information might be a smaller piece among the factors influencing people’s behaviour around health, it is nonetheless important and it is a piece we have the tools to improve.

With regard to the reliability of medical and health-related articles, Wikipedia performs well, perhaps even better than in many other fields. The information in these entries is generally accurate and based on reliable sources. A major test of Wikipedia’s reliability in this area came during the COVID pandemic, when the community successfully fought off misinformation while dynamically updating articles as we learned more and more about the virus.³ This is a feat that should not be underestimated and likely contributed to saving lives. But even before the pandemic, studies have shown that health information on Wikipedia is generally reliable.⁴

However, when it comes to other aspects of these articles, there is less to be proud of. Wikipedia is intended for a general, lay audience but, to put it bluntly, most of its readers won’t have a clue about what they’re reading in these entries. In any given country, a significant proportion of the population has low levels of health literacy which means they struggle with understanding and acting on health information.⁵ Despite this, Wikipedia entries on health and medicine are filled with heavy jargon, overly complex and technical language. Multiple studies assessing the readability of these articles have found that they are difficult to understand and require a relatively high level of literacy.⁶

Wikipedia is generally reliable, but that doesn’t necessarily mean it’s up-to-date. Many articles are based on sources published 5-10 years ago or even older which can easily contain outdated information. A common issue affecting article quality and readability is the way new information is added: rather than replacing outdated content, they are often simply appended to the end of a paragraph or section without removing what has been superseded. Accumulating over the years, this resulted in some articles mutating into something like Frankenstein’s monster, hastily stitched together from disjointed and sometimes contradictory fragments. In the end, they read more like a chronology of research papers than  a clear, coherent overview.

Researchers as Wikipedia editors

Given how crucial reliable, accessible health information is and Wikipedia’s central role, it makes sense for researchers and their institutions to engage with the encyclopedia. The National Institute for Health and Care Research was greatly situated to improve medical and health information on Wikipedia, having research on a wide range of issues from breast cancer through digital health to elderly care and gambling addiction. As NIHR’s Wikipedian in residence, my role was to act as a bridge between the worlds of research and Wikipedia, facilitating the process to ensure that contributions were accessible, ethical, and impactful. 

Initially, the residency was set up with the typical goal of encouraging people to become Wikipedia editors by teaching the necessary skills, organising edit-a-thons and then providing ongoing support for their editing. However, early on, we realised it was necessary to rethink whether this was the most suitable approach for the residency.

Contrary to my expectations, the biggest obstacle to engaging researchers with Wikipedia wasn’t their preconceptions about the encyclopedia. Although these concerns are still going strong, my experience showed that when experts were presented with data about the number of readers each month and given a clear explanation of Wikipedia’s checks and balances, their anxieties tended to fade away comfortably.

The challenges of getting researchers to edit Wikipedia were more practical. Most are busy, overwhelmed, and under constant pressure to publish and climb the academic ladder. Even though communicating research findings (dissemination) beyond traditional venues is now an established part of doing research, Wikipedia editing is still not widely known, accepted, or rewarded by the systems that provide funding or employment. As a result, there is little extrinsic motivation for Wikipedia editing, and it remains mostly an altruistic endeavour. People struggled to find a 2-3 hour slot in their schedules needed to properly learn how to edit. Given these circumstances, no matter how open or enthusiastic they were about Wikipedia editing, only a fraction of people could actively participate in the end.

Our editing workshops were not successful. When they were shorter, covering the most basic skills with practicing and editing assigned as “homework”, typically only a few participants went on to try editing in their sandbox. When events were longer to include time to practice together and work on their edits, we ended up with a couple of sentences at best, which sometimes further increased the mentioned fragmentation of the articles. Despite our efforts, editing didn’t develop into a longer-term habit.

Another complication was that each participant brought their own article to work on which made it harder to check and fix all the different edits and keep track of subsequent changes. This also created more potential points of conflict with other editors. Digesting the fundamental differences between Wikipedia and academia is not easy. It requires a different style of writing, and follows different rules about sourcing. First-time academic editors often bring their particular set of “bad edits”: writing in jargon-heavy language, favoring their own research papers, citing primary studies instead of secondary sources.

Even when the issues raised by other editors were justified, having your first edit removed, sometimes accompanied by a harsh comment, does not inspire anyone to participate in further editing. Plus, new editors were not equipped to reply to talk page messages since there was no time to teach this in workshops, especially as it can’t be done using the visual editor. In the end, participants gained a better understanding of Wikipedia but most didn’t get to experience the achievement and joy of successful editing and did not become regular contributors. Struggling with workshops forced us to rethink our approach. Do researchers actually need to edit in order to contribute?

Testing a different format

To make engagement with Wikipedia more streamlined and to ask less from researchers, we adopted a different format. The new workshops were shorter (60-90 minutes) and thematic. Instead of participants making small edits across a range of articles, we focused on a single topic and the comprehensive improvement of one corresponding Wikipedia entry. Participants were linked by their expertise in that topic, often part of the same research group or members of several groups who used the workshop as an opportunity to collaborate. Ahead of the session, they were asked to read the article, consider holistically what was missing, outdated, or needed improvement. When time allowed, we also divided who works on which section.

The workshop itself completely skipped teaching how to edit. We began with a 10-15 minute introduction to Wikipedia’s key policies and the kind of plain, accessible style of prose we aim for. The bulk of the event was spent collaboratively editing a Google document version of the article, with everyone working on the same file. If questions or ideas came up, there was space to discuss and draw on each other’s expertise. 

A lot can be done during this short time but participants weren’t expected to finish everything in one sitting. At the end of the workshop, now familiar with the task at hand, we agreed on a deadline by which they could work on the article independently. During this time, I could add annotations to the text, highlighting unclear point or missing references. I also did a final copyedit to ensure that the language used was accessible. Once the text was finalised, the actual editing on Wikipedia was done by myself, indicating the group effort in the edit summary. I also managed any feedback or objections coming from other editors.

Impactful engagement without editing

Not needing to learn how to edit, participants could focus fully on the text as a whole. While I emphasised that the article they were working with was the product of community effort and we should have good reasons when we change anything, I encouraged them to be bold in editing: cut repetition and outdated information, rephrase overcomplicated, long sentences for accessibility. The resulting articles were significantly improved, stylistically more coherent, easier to read, and based on the latest scientific evidence. 

Even though focusing on a single article per workshop might seem to limit the overall impact on Wikipedia, this is not necessarily the case. The new parts written during these workshops are ideally well-rounded so they are easier to be reused in other relevant articles. People can come across a topic through various articles. For example, material created in our workshop on head and neck cancers was also incorporated into articles on oral cancer, alcohol, dysgeusia, tobacco smoking, smokeless tobacco, betel nut chewing, salted fish, and many others. Finding and editing these articles requires extra time from the Wikipedian but with relatively little work the results of the workshop can be woven deeper into the fabric of Wikipedia. That’s something I’d definitely encourage others to do as part of their editing efforts.

The joys and responsibility of contributing to Wikipedia

Workshops like this come with their own challenges, and much depends on group dynamics and facilitation. Still, based on my time at the NIHR, this format could offer a lot to Wikipedia. When successful, it might also give participants a sense of having contributed to something meaningful. It might not result in new editors, but people gain a deeper understanding of how Wikipedia works. And for those who are interested, the option to learn editing still remains open.

I encourage researchers and organisations to consider hosting a Wikipedian in residence or engage with the site in other ways. There are many practical arguments for doing so, and ways they can benefit from it. However, I believe there is a strong moral argument that also needs to be made: after a worldwide pandemic, in an era of post-truth, misinformation and cynicism, science has a greater responsibility than before but also less credibility in the eyes of many. To rebuild trust and ensure access to reliable information, science needs to meet people where they are and communicate in ways that are accessible and beneficial. Wikipedia is a central space for doing so, and we can’t afford to overlook it.

1. For this, see my report on Wikipedia: https://en.wikipedia.org/wiki/Wikipedia:NIHR ↩︎
2. https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine/Popular_pages ↩︎
3. https://academic.oup.com/gigascience/article/doi/10.1093/gigascience/giab095/6505121 ↩︎
4. https://en.wikipedia.org/wiki/Wikipedia:Academic_studies_of_health_information_on_Wikipedia ↩︎
5. https://evidence.nihr.ac.uk/collection/health-information-are-you-getting-your-message-across/ ↩︎
6. Multiple studies pointed out that these articles are reliable but they typically require a university level reading comprehension. ↩︎