Lifespan of Data Measured in Dog Years

December 19, 2013 | A paper by a team of Canadian researchers, led by Timothy Vines of the University of British Columbia, suggests that scientific data rarely survives as long as twenty years after publication. In an attempt to recover data from over 500 published studies from the period of 1991-2011, the team found a steady decrease in the percentage of raw data that could be retrieved over time, with four out of five data sets inaccessible after twenty years. The authors, whose work appears in today’s issue of Current Biology, recommend that scientific journals take more active steps to retain data themselves, rather than leaving it to researchers. Cell Press