Consensus Development in the Sciences: The Role of Collaborative Information Behavior in Emerging Knowledge

This dissertation examines the application of consensus development methods by the scientific community, which influences policy, shifts paradigms, and impacts societies. Consensus methods are used as a means to solve problems that are too complex for individuals to solve alone. Such controversial topics require urgent attention due to uncertainty regarding the available research and the potential scale of impact of expert deliberations. Separate and multiple group deliberations on identical topics have produced different and, at times, competing recommendations despite assurances of rigorous fact-checking and sufficient evidence. This raises questions about decision-making processes and participant selection. In other words, how expert groups reach a consensus and why certain group members were chosen has yet to be subjected to a thorough examination. Specifically, this study asks: How do experts bring multiple, occasionally conflicting viewpoints in a group setting into agreement? What is the role of collaborative information behavior in expert consensus development? Moreover, how do experts understand the process of building consensus? The study begins with some scaffolding and provides a collaborative information-behavior lens through which readers can understand how research has arrived at this social-epistemological turn when people are seeking information. The study then presents a definition of consensus development before launching into a discussion of scientific consensus, providing some background on the three organizations (the National Institutes of Health, the National Academies of Sciences, Engineering, and Medicine, and RAND Corporation) whose experts participated in the consensus study. Using an ethnographic methodology, interviews and document analysis were conducted, and analytic coding was used to explain the findings. The analysis of these findings is first broken down based on expert committee onboarding and group deliberations, which include the sociology of professional development and knowledge negotiations, respectively. Then, the study discusses group negotiations regarding scientific and technological knowledge and describes how consensus-forming groups reach agreement in small, collaborative, and (often) closed-off settings. The research identified three knowledge groups (experts, knowledge brokers, and engaged public members); three types of knowledge (expertise, knowledge brokerage, and public knowledge); and nine key collaborative information behaviors for consensus development (homogeneity, adaptation, heterogeneity, collaborative filtering, invisible labor, suggestion, compromise, interpersonal influence, and advocacy) that are needed for group agreement. The study concludes with a discussion and next steps for future research that emphasize the diversity of individual backgrounds and contributions of those involved in consensus development (e.g., inclusive practices for public members), as well as the use of assistive technologies when conducting research (e.g., AI-assisted searches and recommendations on databases) and when meeting (e.g., Zoom).