Making a case for open research: Implications for reproducibility and transparency

Scientific work is time-consuming and expensive. However, after data has been collected and reports have been written, science becomes an information problem. There are thousands of journals and funding agencies across specialized disciplines with different requirements for data transparency. Science-focused domains substantiate research findings through defined research methods and data analysis techniques, and emerging practices around data storage and code publishing now allow other scientists to see exactly how researchers come to their conclusions. These new practices not only increase transparency of the scientific process, they also enable replication and derivative research. Despite these emerging practices, the work of research transparency does not always easily enable reproducibility. At face value, the requirements (e.g., America COMPETES Act; G8 Open Data Charter, etc.) for data sharing seem straightforward: a scientist receives funding and therefore is required to share her data with other scientists. However, the physical sciences and social sciences produce different kinds of data that are more or less easily stored. Additionally, in some cases, researchers can (and should) put restrictions on who can view their data, because of ethical concerns. This panel will begin with current reasons for and methods of sharing data, including national and global initiatives, then will move into problems with open data, sharing data and reproducibility in both quantitative and qualitative research.