Assessing Citizen Science Participation Skill for Altruism or University Course Credit: A Case Study Analysis Using Cyclone Center. (4th June 2018)

• Record Type: Journal Article
• Authors: Phillips, Christopher; Walshe, Dylan; O'Regan, Karen; Strong, Ken; Hennon, Christopher; Knapp, Ken; Murphy, Conor; Thorne, Peter
• Is Part Of: Citizen science. Volume 3:Number 1(2018)
• Publication Date: 2018-06-04
• Subjects: participation -- academia -- credit -- altruism -- climate -- volunteers -- tertiary education; Science -- Citizen participation -- Periodicals; Volunteer workers in science -- Periodicals; 507.2
• DOI: 10.5334/cstp.111
• Languages: English
• ISSNs: 2057-4991

Abstract

A common challenge in citizen science projects is gaining and retaining participants. At the same time, the tertiary education sector is constantly being challenged to provide more meaningful and practical work for students. Can participation in citizen science projects be used as coursework with real practical experiential-learning benefits, without affecting the citizen science project outcomes? We seek to begin to answer this question via a case study analysis with Cyclone Center (CC), which asks participants to classify tropical cyclone characteristics through analysis of infrared satellite imagery. Skill of individual users has previously been shown to be obtainable once classifiers have looked at approximately 200 images using an expectation-maximisation likelihood approach. We use skill scores to determine if participation for course credit or altruism influenced skill for volunteers and students from two universities under three increasingly complex categories of classifications (eye or no eye; stronger, weaker, or the same; and which of six fundamental storm types). A bootstrap resampling approach was used to account for discrepancies between sample sizes. Overall, there is limited evidence for substantive differences in classification performance between credit awarded and altruistic participants, with only one finding of significance at <p = 0.05 (Maynooth University showing lower mean agreement with the volunteer consensus on eye vs. no-eye). There is evidence that integrating participation into a larger assessment that requires the students to show understanding of the project may reduce a low-skill student tail. Furthermore, students' perceptions of the coursework compared to more traditional assignments were overall favourable. These findings, if replicated for other citizen science projects, open up possible avenues to more generally increasing participation in, and exploitation of, citizen science projects in the academic sector.