Henry Sauermann

Abstract

Scientific research has for a long time been performed by professional scientists in a distinct institutional context. Recently, however, scientists have started to cross institutional boundaries by involving the general public (the crowd) directly in research. This crowd involvement tends to be confined to empirical stages of research process (e.g., data collection and processing) and it is not clear whether the crowd can also contribute to conceptual stages of the research process, in particular the formulation of research questions (RQs). Drawing on prior literature on problem solving, we first develop a framework that ties dimensions of research question quality to different types of prior knowledge. We also discuss potential benefits and challenges of involving the crowd in RQ formulation and theorize how knowledge interventions affect RQ quality. We then use data from a field experiment in the medical sciences to explore features of research questions generated by the crowd and to test the effectiveness of knowledge interventions. Our results show that the crowd can generate high-quality research questions that differ in fundamental ways from those typically produced by professional scientists. More importantly, we show that the quality of crowd-generated RQs can be improved through simple interventions that provide crowd members with scientific field and process knowledge they are otherwise lacking. We discuss contributions to the literatures on the organization of science, distributed knowledge production, and crowdsourcing.