Assisting Geophysicists in Data Management: Perceptions and Opportunities

Data sharing is becoming an increasingly important part of the research lifecycle. The number of scientific fields has grown to accommodate the exponential increase in data production, but many scientists are having trouble keeping up. It serves the scientific community well to better understand the incentives that are driving scientific research and the barriers that prevent scientists from sharing data. We performed a survey of scientists contacted through the American Geophysical Union (AGU) asking about their data sharing attitudes and behaviors. The sample of 1,372 researchers are largely from the earth sciences. Respondents to this survey came from 116 countries, were largely from academia, and are spread across all career stages. The barriers that prevent data sharing are largely related to fears that data will be misused and a lack of awareness that better data sharing tools and policies are available.

Results demonstrate that scientists have a positive attitude towards data sharing in general but are more skeptical of the practice for their own data. Most respondents indicated that they would be willing to share at least some of their data provided that they were guaranteed to receive proper acknowledgment and citation for their work. However, fewer scientists indicated that they would actually share their own research data and expressed concern that their data might be misinterpreted. Researchers indicated that they were largely satisfied with their current data sharing practices despite lack of training and ignorance of important data management terms such as provenance and metadata. Researchers also tended to try to solve data related problems themselves or within their research groups. As such, not much scientific data was shared on public repositories and was often spread between servers, personal storage devices, and institutional repositories.

This study indicates that the barriers to data sharing are related to the view that researchers can handle data management themselves despite lack of training, ignorance of who to ask for help, and ignorance of common data management tools and best data practices. These attitudes and behaviors are not likely to change unless researchers are incentivized to share their data. This could possibly be done by associating data sharing with scientific reputation.