Unintended effects of open data policy in online behavioral research: An experimental investigation of participants’ privacy concerns and research validity
Bingjie Liu, Lewen Wei
Abstract
Research outlets are increasingly adopting the Open Data policy—requiring or encouraging researchers to release data publicly. However, public data sharing in the digital age may threaten participants' privacy and thereby discourage participants to disclose. We tested this possibility and explored solutions with a between-subjects online experiment. Participants from Amazon's Mechanical Turk (N = 294) were randomly assigned to one of five conditions with data sharing policies varying on the level of publicness or not mentioned. Participants in the public-access condition reported greater privacy concerns and fewer unethical conducts than those in the private condition. Public data sharing also indirectly decreased participants' amount of open-ended disclosure via privacy concerns. When asked, participants reported privacy violation as their primary concern regarding data sharing. Findings suggest sharing data with researchers' gatekeeping, rather than indiscriminately public, may be a solution that better serves the goals of Open Science—ethical and valid science.