n.a portion, within a body of records, selected for permanent retention through a process that employs a preset pattern of selection to produce a subset of the whole; a systematic random sampleHindus, Hammett, and Hobson 1979, 13A systematic sample of an unnumbered records series might require manual counting of all cases in order to pick out, for example, every tenth case.Ham 1981b, 361The report’s recommendations, which the court made operational in December 1980, call for selecting a basic corpus of records from the counties by a systematic sample (using docket numbers).Boles 1981, 127In a typical systematic sample, every hundredth element of the population would be selected for the sample.Hull 1981, 13As the name implies, systematic sampling depends upon the establishment of a particular pattern of selection in that either every nth. file is preserved, or else all the papers for a particular month, year or other chronological unit. It used to be considered that, especially in the former case, the result was sufficiently haphazard for statistical analysis, but this is not necessarily the case and researchers have complained in the past of the wrong papers being retained by this method. In the chronological case, there has been a tendency to retain census year records, or for records to be kept for every fifth year.Kepley 1984, 240The Koln Archives in West Germany, for example, retains all uniform case files on individuals whose surnames begin with the letter “H.” This systematic sample is obviously easier to administer than trying to select randomly from the entire series, since one would first have to number each file.
Systematic sampling replaces true randomness with a systematized pattern of selection that is easier to implement and relatively, though not actually, random. Systematic sampling is conducted in cases, for instance, where random numbers cannot be generated to match case numbers because no case numbers exist. In such cases, a simple pattern, such as choosing each tenth file for retention, is used to imitate randomness.