Data are the lifeblood of science. But all those carefully recorded observations may be in danger. Because a new study shows that data from the recent past are being lost at an alarming rate. The journal Current Biology has the data to prove it. [Timothy H. Vines et al., The Availability of Research Data Declines Rapidly with Article Age]

Scientific studies build on research that came before. And scientists turn to the facts and figures from previous work to aid in their own analyses or confirm that the earlier results still hold up. But how often can they access the older data they need?

To find out, researchers selected 500 studies published between 1991 and 2011. And they sent the authors a request for the studies’ raw data. Twenty years after publication, 80 percent of the data was unavailable.

In some cases, the authors couldn’t be reached. When they did respond, many reported that the data were simply not accessible, buried in an attic or saved on a now unreadable floppy disk.

This loss of information is an impediment to ongoing research and a waste of funding. Perhaps scientific data should be recognized as an endangered species, and efforts made to keep data around for future generations.

—Karen Hopkin

[The above text is a transcript of this podcast.]