An Experimental Study of Learning in an Online Citizen Science Project: Insights into Study Design and Waitlist Controls

Abstract

The field of citizen science needs experimental studies to develop a better understanding of the connection between citizen science participation and learning. We conducted an online experiment to test whether social interaction and the primary learning activity, mapping, led to increased content-learning in the citizen-science project, YardMap. Participants were randomly assigned to three treatment/control groups: the full, socially-networked mapping project, the mapping project with its social tools disabled, and a waitlist-control group, whose members took the pre-test and post-tests along with the active participants, but were blocked from participating in the project for a two-month interval between the tests. Based on general linear models, post-minus-pre-test scores (learning gains) did not differ between the two treatments, nor between the treatment and control groups. Individual level of activity in the project did not affect learning gains, but the learning gains were negatively associated with pre-test score for all three groups, indicating that learning occurred, not only in the treatment groups, but also in the control group. Retrospective analysis of participant completion rates, effort, and responses focused on developing possible explanations for this outcome. This analysis uncovered factors that reduced the ability to detect learning, which may be common to many Citizen Science research studies. Of particular interest were factors related to the use of a waitlist control design in online settings. This research concludes with new recommendations for the design of controlled studies of informal science learning, including new controls to explore the mechanisms by which waitlist control participants learned as much as participants in the treatment groups.

Keywords

• citizen science
• social learning
• waitlist control
• online learning