Scholarly publications and scientific claims are the descriptions of a research work and its conclusions, but by themselves do not provide full disclosure on how the research has been done. A comprehensive research publication or claim should be accompanied with the underlying data and analysis used to formulate the results. This transparency of the research process, including sharing the data and analysis, allows us to understand and validate the conclusions.
It is not any longer the case that advance in science and knowledge is achieved in isolation by an individual or a sole scientific field. Collaboration is a critical piece of research progress, and increasingly includes multiple disciplines and skills. Sharing data openly allows researchers in fields outside yours to find your data and foster new collaborations across fields.
Making your data easily accessible enables other researchers to reuse and analyze the data from other perspectives, with the potential of offering new insights from the original work. As shown in the history of science where making data available from previous experiments or observations was essential to new discoveries, openness helps to accelerate the pace of science.
Scientific truths must be verifiable and reproducible. However, some scientific fields are suffering from what it is now refer as a reproducibility crisis. Whether we are living a reproducibility crisis or we are simply now more aware than ever that reproducibility might get lost in the abundance of scientific publications in the 21st century, is up for analysis and discussion. But we can all agree that a critical step towards enabling and improving reproducibility is to share the data and analysis accompanying a claim.
As we have seen, sharing data openly is good for science and research. But it can also be good for you. When you share data in a public trusted repository (such as the Harvard Dataverse), you get automatically a formal data citation for your data set, with a persistent url (DOI) and proper attribution. As part of scholarly etiquette, when others use your data set, they cite it, and your number of data citations increase.