This horrifying graph comes from a paper by Timothy Vines and colleagues at the University of British Columbia, who wanted to get an idea of how the accessibility of scientific data erodes over time.

They could only confirm the existence of 23% of the data sets from more than 500 articles published between 1991 and 2011 on the subject of plant or animal morphology. The odds of a data set remaining accessible fell by 17% per year.  Two-thirds of the data sets from papers published in 1991 were lost or  deemed inaccessible because of antiquated storage media.

Vines and colleagues acknowledge that they might have recovered more data if the authors they contacted had more incentive to spend time digging it up. Still, it’s maddening to think about the careless dissolution of a resource that is largely produced by public funding. Leaving the preservation of data to individual scientists “denies future researchers any chance of reusing them,” Vines and co-authors assert.

They say  funding agencies and journal publishers should require scientists to upload data sets to public archives where it can “be preserved in perpetuity and can no longer be withheld or lost by authors.”

Source: The Availability of Research Data Declines Rapidly with Article Age by Timothy H. Vines, Arianne Y.K. Albert, Rose L. Andrew, Florence Débarre, Dan G. Bock, Michelle T. Franklin, Kimberly J. Gilbert, Jean-Sébastien Moore, Sébastien Renaut, Diana J. Rennison; Current Biology , 6 January 2014 (Vol. 24, Issue 1, pp. 94-97)