The editing process also includes another type of correction, namely the assignment of a full set of characteristics for a person or a housing unit. The assignment of the full set of housing characteristics occurs when there is no housing information available. If the housing unit is determined to be occupied, the housing characteristics are assigned from the previously processed occupied units. If the housing unit is vacant, the housing characteristics are assigned from the previously processed vacant unit.
The assignment of a full set of characteristics for a person in a household containing no information for all or most of the people, although persons were known to be present, is done in the following manner. A previously processed household is selected as a substitute and the full set of population characteristics for each substituted person is duplicated. These duplications fall into two classes: (1) "persons substituted due to noninterview, e.g., a housing unit indicated as occupied but the occupants are not listed on the questionnaire, and (2) "persons substituted due to mechanical failure, e.g., where the questionnaire page on which persons are listed was not properly microfilmed.
The editing process ensures that the distribution of characteristics for persons and housing units assigned by the computer corresponds closely to the distribution of characteristics for persons and housing units actually reported in the census. Before the advent of the computer, this type of process was not feasible and the distribution of characteristics to be used for assignment purposes was drawn from the preceding census or another source.
Specific tolerances are established for the number of computer allocations and substitutions that are permitted. If the number of corrections is beyond tolerance, the questionnaires in which the errors occurred are clerically reviewed. If it is found that the errors resulted from damaged questionnaires, from improper microfilming, from faulty reading by FOSDIC1 of undamaged questionnaires', or from other types of machine failure, the questionnaires are reprocessed.
STF 1 includes four tables (56, 57, 58, 59) that allow users to determine the extent of allocation and substitution for a given geographic area. For example, if a user wants to determine the extent of substitution in a particular county, table 56 provides the total number of substituted persons in creating tabulations for that county. Similarly, table 58 provides a count of housing units substituted. In addition, tables 57 and 59 provide data on the extent of allocation. For example, a user interested in the number of persons with sex or marital status allocated for a particular geographic area would want to know the value in the appropriate cells (3 and 7) of table 57. Similarly, the 9 cells of table 59 provide data on the number of housing units for which particular characteristics were allocated.
1Film Optical Sensing Device for Input to Computers.