Information sharing is essential to scientific progress. In principle, unconditional sharing of knowledge among academics is enforced by the priority-based scientific reward system in which the first person to discover a result gets credit for the discovery. There is, however, a tension between communal sharing and the competitive incentives for researchers during the research process itself. A scientist who shares results provides stepping stones for discovery by others who may not acknowledge the contribution. This tension, as well as commercial potential for academic work has led to concerns over misappropriation of scientific research and increased reluctance to share information ad materials.
Intellectual Merit: The project involves theory and a survey to support empirical research on what drives academic researchers to share information. The theory considers three contexts in which researchers share: one-on-one situations in which one researcher is asked by another to share specific data or materials; public sharing, such as conference participation where researchers present work that is neither published nor patented at the time of presentation; and target sharing in which they share certain types of information with trusted colleagues prior to public dissemination.
Preliminary results suggest that information sharing depends on the conditions in which the research process is embedded, and these conditions themselves depend upon dimensions of scientific policy (e.g., journal and federal funding agency policies). The models also provide hypotheses about the extent to which researchers share (and with whom) as a function of individual characteristics including type of research, age, and rank as well as other environmental factors. The theory provides the context for a survey of academics across a wide array of fields to include engineering, social sciences, biological and medical sciences, mathematics, physics and statistics.
Broader Impact: Combined with econometric analysis, the research provides a new framework for understanding of the ways in which open science operates or not across a broad spectrum of academia.
Information sharing is critical for scientific progress, so much so that communalism is considered an ideal. However, a number of studies suggest that commercial incentives for academic patenting, as well as competitive pressures, compromise information sharing. This grant focused on the incentives for university scientists to disclose or share information about their research. While prior studies focused on disclosure incentives in terms of concerns that competitors would scoop patentable or publishable results, our work has highlighted the importance of disclosure in attracting resources as well as enhancing ones reputation regardless of the impact on publication. We constructed a theoretical framework to examine disclosure and competitive incentives among scientific researchers in these different environments. We considered situations in which a researcher who has made a discovery of partial value in solving a problem in her field decides whether to keep it secret until completion of the project or to disclose it to her colleagues. We first model disclosure that risks unwanted competition despite the fact that it gives credit for discovery and may attract collaborators with skills and/or resources to augment the project and it also the potential to remove competitors. We then modeled the contrasting situation where the researcher wants to attract others to work independently on her line of research. In this model, the disclosing scientist gains a reputational benefit if she convinces other researchers to pursue her line of investigation—potentially creating a hot new area of research. We refer to the first model as competition/collaboration and the latter as the research leader/hot area. The two variants gave a rich set of results for disclosure as a function of the resources and/or quality of potential competitors relative to the focal researcher. In the competition/collaboration model, where we introduce a distinction between trusted and general colleagues, we find that an increase in the quality of one’s trusted colleagues, all else equal, increases the likelihood of disclosing to general audiences. An important factor behind this result is the fact that the focal researcher can improve the quality of her project if she attracts a higher quality trusted colleague to collaborate, which acts as a deterrent to potential general competitors. Our results highlight the different faces of competition. In the competition/collaboration case, the marginal impact of an increase in the quality of general colleagues is negative, but it is the greatest in magnitude when the focal researcher is more capable than her general colleagues. This result occurs because collaboration rarely occurs in equilibrium in this case, so that the dominant factor in the focal researcher’s expected payoff is the reduction in her probability of being first to solve the complete problem. In the research leader/hot area model, the marginal impact of an increase in the quality of general colleagues on entry and the stage of disclosure are both positive. With higher quality general colleagues, the focal researcher’s probability of being first to solve her problem declines, but if she can attract others to this area, she benefits from having introduced the problem. The higher quality colleagues are more likely to be first to complete the problem. The focal researcher discloses at a later stage, however, because this increases the deterministic portion of her expected payoff. This face of competition, though quite intuitive in the context of the research leader/hot area case, is not present in the traditional view of competition. Based on the theory we constructed a survey regarding the willingness of academics to share preliminary research results with other researchers. As well we queried respondents regarding their demographics, their research profile and their views of the nature and intensity of competition. The survey was emailed to almost 82,000 academic researchers of whom about 78% are based in the United States. There are 9,492 respondents (11.6% response rate). Respondents came from 121 institutions. Multiple departments are represented among respondents: 31% in medical school, 13% in the biological sciences, 17% in physical sciences, 16% in engineering, 12% in mathematical sciences, 6% in agricultural sciences, and the remainder are in the social sciences. Just over 70% of respondents are male. Only 3.9% responded that they never or rarely shared with trusted colleagues whereas 15.9% never or rarely shared with general audiences. However, when they share results are often withheld: 75.9% (53.1%) claimed that they never or rarely withheld from trusted colleagues (general audiences). When asked about the stage of development of an idea when it was disclosed, 46.6% claimed to disclose to trusted colleagues at the conceptual stage whereas only 10.5% claimed to disclose at the conceptual stage to general audiences. When asked the importance of disclosing preliminary results as a means of attracting collaborators, 36.6% of respondents indicated that this was either important or very important when disclosing to trusted colleagues. The comparable number for general audiences was 32.7%.
