Importance of Preserving Raw Data

Our effort to make the Apollo Lunar Surface Experiment Package (ALSEP) data, acquired from 1969 to 1977, readily available to current and future researchers, as reported in this session last year, has made us realize some potentially serious problems that may be common to legacy scientific data. One such problem was the loss of key information contained in the original data. This happened (a) when the original investigators thought it was enough to archive only the data of interest to contemporary researchers; (b) when the data were reformatted to make them readily decipherable to current researchers by someone who did not fully understand the details of the data, thus introducing incorrectly translated information; (c) when errors associated with reading of the original data were not corrected; and (d) simply due to a programming error in reformatting. The only sure way to avoid such a problem is to archive the raw experimental data in addition to the processed data, even though they may not be readily decipherable to most current researchers. For such raw data to be useful, it is also important that complete metadata necessary to interpret the raw data be preserved.