Good practices for sharing analysis-ready data in mammalogy and biodiversity research

Abstract

As both producers and consumers of data, scientists play an important role in defining how accessible their research outputs are to others. First by deciding to share, but also through the choice of file formats and data structures used to share data. Steps taken by authors, editors, and typesetters to format and store data often complicate the ability of future users to work with these data. At late stages of the scientific workflow, making analysis-ready versions of the data takes relatively little time and effort in exchange for a significant increase in usability and, potentially, other well-known benefits of data sharing such as more citations and potential collaborations. Well-structured and analysis-ready data also reduces the risk of unintended alterations introduced while cleaning and rearranging published data. We wish to reconcile what is easy to read and intuitive with machine-readable data that does not need extensive processing or advanced programming skills for inclusion in new analyses. For those who use and report biodiversity data and the results of specimen-based research, we wish to create awareness of the major differences in structure between data at the analysis stage compared with data arranged and formatted for reporting. We hope that the reader might apply these practices when sharing data with other scientists and with the public.