The Creative Commons has released Recommendations for Independent Scholarly Publication of Data Sets. This is a working paper.
Here's an excerpt:
In an ideal world, any data collected by a research study would be available to anyone interested in validating or building on that data, just as is the documentation describing the study itself. Some data has value that goes beyond the study for which it is generated, and getting the data to those who can use it for reanalysis, meta-analysis, and other applications unimagined by the study authors is to everyone's benefit. Data reuse failure is receiving growing recognition as a problem for the research community and the general public. The road to reuse is perilous, involving as it does a series of difficult steps:
- The author must be professionally motivated to publish the data
- The effort and economic burden of publication must be acceptable
- The data must become accessible to potential users
- The data must remain accessible over time
- The data must be discoverable by potential users
- The users use of the data must be permitted
- The user must be able to understand what was measured and how (materials and methods)
- The user must be able to understand all computations that were applied and their inputs
- The user must be able to apply standard tools to all file formats
- The user must be able to understand the data in detail (units, symbols)
This report considers how the genre of the data paper, suitably construed, might be used to help a data set survive these trials.