Insights for Informal Science Institutions from Citizen Science Projects
Today we have a guest post from Dr. Ryan Cook, Citizen Science Learning Researcher at the Adler Planetarium. Ryan earned his PhD in socio-cultural anthropology from the University of Chicago. His research interests include ethnographic investigations in Mexico and the US on the intersection of science and religion.
It has been my pleasure to be a researcher for Zooniverse, based at the Adler Planetarium in Chicago, since May 2012. This position has exercised my anthropologist’s curiosity about how people understand and engage with science, taking it in an interesting and very productive new direction that I plan to continue. Thus I am pleased to have a chance to share my work on this blog.
At this writing I am close to completing my portion of a federally-funded project studying Zooniverse volunteers. I have benefitted greatly in this research from the assistance of your esteemed edu-bloggers, Kelly and Laura, as well as my former Adler colleague Jason Reed and former supervisor Karen Carney. Specifically, we tried to determine whether and how much volunteers’ conceptions of and attitudes towards science changed through their participation in virtual citizen science projects.
This week, I presented some of our findings at the Visitor Studies Association’s annual conference in the town that beer built: Milwaukee, Wisconsin. Outfitted with a snazzy poster and a pile of official Zooniverse postcards and stickers, I argued for the relevance of our studies of Zoo volunteers to museums and science centers that want visitors to their websites to learn about science.
To know what could possibly be learned about science in Zooniverse, Karen, Kelly, and I put together a model of understanding science to guide us. We based the model’s criteria on what scholars who theorize, research, and teach science claimed as central characteristics of the sciences — for instance, relying on sense experience, proceeding methodically or logically, and revising knowledge in light of new evidence.
I then spent several months combing through Zooniverse databases and Google Analytics tables, trying to create a quantitative picture of how volunteers engaged with the tasks, blogs, and forums making up each Zoo. Figure 1 shows an example of the data by which we quantified and compared engagement among Zoos.
Following the lead of some preliminary statistics, Kelly and I applied our model to mapping out opportunities for learning about science in a subset of mature Zoos (i.e., those launched before the shift to an all-in-one-page design strategy). The Zoos were chosen in pairs with similar tasks but different levels of volunteer engagement:
- classifying – Galaxy Zoo Hubble and Supernovae
- marking – Milky Way Project and Moon Zoo
- transcribing – Old Weather and Ancient Lives
[*Since the Supernovae Zoo was retired during the course of our project, it was included in the engagement variables but left out of subsequent research stages.]
Upon matching these engagement statistics to the range and type of learning opportunities we identified, three main patterns emerged:
- Opportunities for science learning were unevenly distributed within and across Zoos’ webpages. Talk and the Forums, for instance, allow a wide range of volunteers to engage in rich communication with each other and with moderators, administrators, and the science teams regarding the scientific import of the Zoos.
- The parts of the Zoos where volunteers went in the greatest numbers and spent the most time were typically those with the fewest, most limited, and least obvious learning opportunities. High-traffic, low-opportunity pages included the classification, marking, and transcription tasks at the core of each Zoo, as we can see in Figure 2.
1. Of the more than 700,000 volunteers to visit these Zoos at the time of our analysis, only a small percentage stayed long enough or reached enough pages to encounter many of the learning opportunities we identified.
Each of these findings makes sense if we bear in mind that Zooniverse did not start out as a platform for volunteers to learn about science, but rather as a tool for scientists to carry out certain kinds of data-intensive research.
I contended in my VSA presentation that this mismatch offered museums and science centers some guidance in how to (re)design their websites to improve the chances that visitors would encounter opportunities to learn what the institutions decided was important. Laura, Kelly, and the Zooniverse team have been testing out ways to design more learning opportunities into the “stickiest” parts of the Zoos.
And as for me, I have followed up this quantitative work with a series of in-depth interviews of heavily involved volunteers. By coding their responses based on an extended version of our science learning model, I aim to find out what they feel they learned from their Zooniverse engagement and how it helps us to determine how one segment of volunteers engaged with the science learning opportunities we identified. This interview material will appear along with the engagement data and the science learning model in my report, which should be completed by late September. Stay tuned: you will hear about it first!