Skip to main content
Menu

Book

A physicist reveals how citizen science is reshaping research

The Crowd and the Cosmos: Adventures in the Zooniverse

Chris Lintott
Oxford University Press
2020
288 pp.
Purchase this item now

To understand how star formation varies in different galaxies, Chris Lintott and Kevin Schawinski needed to sort elliptical galaxies from spirals. But image recognition software can be costly to produce and needs training data to generate accurate results. The repository they had in mind—the Sloan Digital Sky Survey—included photometric observations of nearly one billion objects and spectra from more than four million objects. Inspired by Stardust@home, a 2006 project at the University of California–Berkeley that enlisted thousands of volunteers to search for interstellar dust particles in millions of images, the pair launched Galaxy Zoo in 2007.

During the first year, 150,000 volunteers contributed 50 million high-quality data classifications. With this training dataset in hand, the duo wrote an algorithm to identify star-forming blue ellipticals. The results were surprising: other elliptical galaxies substantially exceeded the star formation rates of the Milky Way.

Galaxy Zoo’s success led to the creation of Zooniverse, a more expansive platform—cofounded by Lintott, Lucy Fortson, Pamela Gay, Stuart Lynn, Arfon Smith, and others—that allows researchers to apply the same approach to other scientific challenges. In The Crowd and the Cosmos, Lintott joins a growing cadre of authors focusing on the contributions of volunteers to the scientific enterprise (1–5), documenting the lessons learned from these projects, as well as trends in astronomy that opened the doors for new ways of doing science—from making data more open and remotely accessible to analyzing big data with machine learning. He also reflects with passion and curiosity on the detective work needed to explain the mysteries of the Universe.

Zooniverse and projects therein have enabled more than a million volunteers to contribute their skills remotely to scientific research—no specialized background or expertise required. The platform also features a community forum that allows citizen scientists’ research questions to emerge. While participating in Galaxy Zoo, for example, Dutch schoolteacher Hanny van Arkel found a blue blob in galaxy IC 2497 and asked a simple question on the forum—what is it? The object, dubbed “Hanny’s Voorwerp” (“Hanny’s thingy”), turned out to be a rare quasar ionization echo.

Lintott uses Hanny’s Voorwerp to highlight the problem of “inattentional bias,” which can occur when one is so intently looking for a specific object that one misses the unexpected. Professionals can be prone to this bias, Lintott notes, whereas newer volunteers “are likely to be conscious of all sorts of things in the data, some interesting and novel and some not.”

Zooniverse expanded the perceived value of citizen science, shifting it from educational to a tool for making meaningful contributions to scientific research. More than 200 peer-reviewed publications have arisen from Zooniverse projects, 63 from Galaxy Zoo alone.

Citizen science projects have had to overcome a hierarchical view of science that originated in the Victorian era and persists today. This view assigns status and greater value to the contributions of those who direct or fund a scientific endeavor as compared with the contributions of those who do the work. Lintott pushes back: “Rather than…viewing the Eminent Scientist as apart from the Crowd—someone who assigns tasks and reserves only for themselves the right to analyze the results—it instinctively seemed to me that it is impossible to draw a clear line between where the supposedly menial tasks of data gathering, classification, and exploration stop and some sort of Proper Science starts. It’s all just science.”

Forty-two meta-studies explored the sociotechnical opportunities and challenges presented by Zooniverse, including the need to incorporate design thinking into projects; the need to understand computer–human interactions as well as what motivates volunteers to participate; the need to develop robust data quality assurance protocols; and the need to rethink who should be included as a publication author. By addressing these challenges, the Zooniverse helped to revolutionize the practice of citizen science.

Although convolutional neural networks can be trained to classify galaxies or to differentiate galaxies from stars, they still may overlook unusual and important objects. By paying close attention to each system individually, Lintott asserts, citizen science may help us discover something fundamentally new.

References and Notes
1. M. Nielsen, Reinventing Discovery (Princeton Univ. Press, 2013).
2. C. Cooper, Citizen Science (Harry N. Abrams, 2016).
3. M. E. Hannibal, Citizen Scientist (The Experiment, 2016).
4. D. Cavalier, E. B. Kennedy, Eds., The Rightful Place of Science (Consortium for Science, Policy and Outcomes, 2016).
5. A. H. Kimura, A. Kinchy, Science by the People (Rutgers Univ. Press, 2019).

About the author

The reviewer is at the Nelson Institute for Environmental Studies, University of Wisconsin–Madison, Madison, WI 53706, USA.