Research Data Management

Reproducibility and replicability are concepts that have received increased scrutiny over the past decade. Although their meanings can vary depending on discipline and research community, they both point to the fact that advances in the scientific enterprise depend on the credibility of previous work. Improving reproducibility and replicability are goals supported by major funders and scientific organizations. 

The authors of Reproducibility and Replicability in Science (National Acadamies Press, 2019) define them broadly in this way:

• ...reproducibility [means] computational reproducibility—obtaining consistent computational results using the same input data, computational steps, methods, code, and conditions of analysis
   
• ...replicability [means] obtaining consistent results across studies aimed at answering the same scientific question, each of which has obtained its own data
   
• ...reproducibility involves the original data and code; replicability involves new data collection and similar methods used by previous studies

Other researchers and organizations have come up with different definitions, summarized at Curating Reproducibility (CURE). A more comprehensive treatment of reproducibility can be found in the Stanford Encyclopedia of Philosophy, Reproducibility of Scientific Results.

Best practices in research data management, like those articulated in this guide, were developed to address growing concerns over reproducibility and replicability. Developing a data management plan, creating file naming conventions and data workflows, carefully recording protocols, and cleaning data and making calculations using open source software, among other things, can help to ensure that your work can be reproduced or replicated by others. Refer to the contents of this guide and to the Further Reference on this page.