Thing 17: Crowdsourcing


“ARE YOU READY… to meticulously annotate 12,000 documents of 18th-century French paleography?!”
Photo by Marc-Olivier Paquin on Unsplash


Researchers in a variety of disciplines are tapping into the possibilities of crowdsourcing and what it can accomplish. This can mean crowdfunding to support scholarly projects (such as a recent campaign to pay for the restoration of Mary of Gelderland’s unique medieval prayer book), or a relatively passive involvement, such as SETI@Home, in which participants allow the project to harness the power of their computer processors when not in use.

More active varieties extend to ‘citizen science’– calling on the public to volunteer their involvement. At one end of the scale, this may mean massive data projects like those Zooniverse hosts, which have attracted over 1 million registered users thus far. Participants are given tasks which are relatively low-skill, but which humans perform better than computers – such as identifying animals from captured video. Current Zooniverse projects range from providing checks on census data to mapping vibration patterns on steelpan drums. Crowdsourcing isn’t limited to the sciences; the Transcribe Bentham project at UCL asked members of the public to transcribe Jeremy Bentham’s letters, and proved wildly successful.

We have our own projects here at Surrey, such as asking the local community to provide measurements of air quality in their daily lives. Previous projects have used the public and an app to help track endangered species in the wild. Citizen science is increasingly seen as an important part of public engagement, since beyond its immediate results it helps to explain the (sometimes mysterious) work of the university to the public, and demonstrates how what we do has real impact.

There are valid concerns about crowdsourcing as a research method – it’s not a magic wand. It is important to factor in the reliability of the information collected, the ethical relations between the project and ‘crowd-taskers’; crowdsourcing may not be a good fit with some research methods. Undeniably, however, the network of technology and massed individuals can make possible projects that would have been impossible (or at least unaffordable) for researchers to carry out directly. The most obvious examples would be with big data, where simple analysis needs to be carried out over massive information sets.

But it isn’t always about big data. The work that the Run CoCo Project in Oxford does, for example, is about inviting members of a community to contribute unique knowledge and expertise to a project – for example, by adding historical information, identifying people or locations in a photo, or even asking communities to upload their own content. You might also take a look at other types of projects which ask participants to add contextual information, such as HistoryPin, or to identify locations in paintings, such as the Tate Britain ArtMaps project.

It also need not always be about very prominent or grand projects. Several (mostly commercial) options exist that allow projects to set small-scale or short-term tasks that are then farmed out to participants. For example, Amazon Mechanical Turk pays participants a small sum to do simple tasks, such as annotating image content. This can then be used to create training data for machine learning, for example.

Despite claims made about its efficiency, it’s important to remember that while a crowdsourcing may involve the public or those interested in your discipline, it doesn’t usually mean less work. Instead, it’s a matter of switching to the roles of designer, publicist, manager and supervisor. You need to engage those who may be interested – and keep them engaged. Studies have shown that in many projects, a small minority do the majority of the work. You also may need to monitor or quality control output.


“Here is my data. Please sort.”
Photo by Jordon Conner on Unsplash

This Thing certainly doesn’t ask you to undertake a crowdsourcing project, but it does ask you to consider whether crowdsourcing has or might have any role in your work. Take a look at some of the types of projects on Zooniverse and Run CoCo. Run CoCo offers a variety of resources for those interested in a project; take a look at their guides and reports. You can also find guides at Jisc or THIS Institute.

You may also be interested in crowdsourcing as a learning activity. If you’re involved in teaching, think about the role that participation in these projects might play in the classroom.

Further reading



Photo by Tom Plouff on Unsplash

Author: rdpsurrey

Providing personal and professional development opportunities to researchers at the University of Surrey.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s