Machine-readable data files in the social sciences: An anthropologist and a librarian look at the issues

The problems in making machine-readable data files (MRDFs) accessible as well as in using them are basically the same for all social scientists. There are two kinds of problem: the quality of the data in the MRDFs themselves and accessibility. The social scientists’ concern is quality control; the librarians’ is accessibility. “Clean” datasets make it possible to ask questions that could not be asked in the past. In analyzing large amounts of data, social scientists can detect the “weak signals” of important relationships. Three problems confronting librarians are: availability of bibliographic control; the quality of documentation; and level of user skills.