Research data management (RDM) is a set of practices that make it easier for researchers to find and use data for their research projects when they need it. RDM is like brushing your teeth and flossing to prevent cavities.
Because we have high expectations about the integrity, validity, and reliability of research results, the daily data management practices are very important to the outcomes of research. RDM includes all activities performed while handling data generated or gathered as part of a research project. While good data management practices supports good research outcomes, poor data management practices can make data unusable and lead to the failure of a project. RDM includes decisions and practices relating to:
While specific strategies for managing data differ between fields of research, like research design, methods, and expectations for research integrity can vary, there are some good data practices that apply across all types of research. These include the following:
Other good research practices that effect data practices (usually the responsibility of the Lead/Principal Investigator)
Data is a key piece of the scholarly record. This means that the way you manage your research data has an impact on the accuracy and integrity of the research record (i.e., scholarly literature including journal articles, conference posters & presentations, abstracts, etc.). Since the scholarly literature is used to inform new areas for research, whether data are managed well can effect the potential for data curation, sharing, and reuse or secondary analysis. This is recognized by the Office of Research Integrity, the National Academies of Science, federal funding agencies requiring data management plans, and initiatives like FORCE11. Kenneth Pimple describes data management as “the neglected, but essential, twin to the ‘scientific method.’”
Source: Coates, H. (2014). Ensuring research integrity: The role of data management in current crises. College & Research Libraries News, 75(11), 598-601. https://doi.org/10.5860/crln.75.11.9224
Data management is an important part of the conversations about challenges in reproducing and/or replicating published results. Many areas are facing these challenges, such as psychology, cancer research, cell biology and more, though the specific issues differ by the field of research. Advancing our understanding of the world requires an accumulation of evidence from more than one study. Another way of saying it is that one study does not prove a theory. Many studies producing consistent data and results are necessary for a theory to be accepted as the most likely explanation.
Comparing, aggregating, and analyzing data across multiple studies requires that data are accessible, interoperable, defined, well-documented, and citable. One approach to describing and making these ideas practical are the FAIR Guiding Principles, which are: Findable, Accessible, Interoperable, and Reusable.
