UNIVERSITY POLICY AS A BARRIER TO SHARING SOFTWARE IN SCIENCE

Background: From data collection to analysis and the presentation of research results, scientific research relies on software and an increasing number of researchers not only use, but create software as a part of their work. While there is increasing interest in sharing computer programs in an effort to reduce duplicative effort and improve the reproducibility of science, researchers have noted that university policies related to software sharing can be confusing and hard to navigate. In an effort to better understand social and cultural barriers to software sharing in the academic environment, this study benchmarks current academic policy with respect to research software ownership and sharing. Methods: This study gathered intellectual property, copyright, and invention policies from 119 highly ranked American research universities. Each of these polices was examined to determine whether software was explicitly considered as a faculty research product and if so how software products were treated by the university policy. Results: Less than 10% of institutions did not explicitly identify software in their intellectual property policies. Of the university policies that do address software, 53% recognized it primarily as a technological innovation, while 36% recognized a scholarly role, and the remaining institutions did not clearly identify how it should be considered. Conclusions: While software is broadly recognized, the standards and practices governing software in the academic environment remain diffuse. Policy clarification enhances sharing of research software and improves the efficient diffusion of modern science.