Data Sharing Manifesto

This document is also available as
.pdf or .md.

Data Sharing Manifesto

  • Databrary's primary purpose is to foster open data-sharing among a community of researchers in developmental science in particular, and in psychological and neural science more generally.
  • Open data-sharing is entirely voluntary. Investigators choose whether to share, when to share, and what to share. Databrary hopes that most Investigators will choose to share, will share early on in the research process, and will share most or all of the information they have permission to share. But, there is no requirement to do so.
  • Some models for when to share are as follows:

    • At the end of the formal grant period of a federally funded project.
    • Within one year of the date an article is accepted for publication or appears (online or in print).
    • When an article is accepted for publication.
    • At the time of submission of an article for peer review.
  • The more investigators who share and the more information they share, the more the community of researchers will benefit, the more scientists will discover, and the more the public will benefit.

  • Investigators who use information shared by others should follow ethical principles.
    • They should cite resources contributed by other Investigators. Citation is an important currency of scholarly activity.
    • They should contribute data and analyses of their own to Databrary in a timely way. Not everyone has the resources to collect highly valuable data sets. But, secondary analyses of data sets should be shared and linked to their original sources.
    • Investigators who use datasets contributed by others should consider contacting the data contributor when embarking on a new analysis of a contributed data set. It may also be appropriate to invite the data contributor to collaborate.
  • Databrary encourages data from unpublished studies to be shared. In fact, Databrary provides an excellent home for them.

This document is also available as
.pdf or .md.