1970 HUNGARIAN CENSUS OF POPULATION Information on the data collection and processing (1977)

I. THE COMPREHENSIVE REWIEV OF THE DATA COLLECTION AND PROCESSING - 6. Sampling techniques applied in the census

pairs of elements of the systematic sample. In the case of the 1 per cent representativity sample, applying 100 as multiplier, it may be supposed that not significant correlation exists between the units to be selected. The systematic selection has, at the same time, a stratifying effect on the procedure, since the selection takes place from a population arranged according to the regional principle. Consequently, several parts of the region (counties, districts and communities) maybe regarded as strata. Thus the automatic stratification originating from the systematic selection inceases the accuracy of the sample, and results in more reliable samples displaying several parts of the region (counties). th In the course of the practical execution, the selection was begun by the 100 sheet of the Register of dwellings in each district or town, including towns having county rank, and in Budapest. The initial number could hardly influence the reliability of the estimation, but lightened the technical performance of the work. On account of the use of 100 as the initial serial number, it may have occurred that one flat less in the districts and five flats less in the counties have been included to the sample. It can, to some extent, counteract the effect resulting from the fact that the selection was carried out on the basis of the total number of private dwellings and institutional households equally. Consequently, if the 100

t

h element of the sample was an institutional household, the next one in the series had to be included to the sample. For sake of simplicity and perspicacity of the selection, these negligible deviations had been taken into consideration in advance, and they could be eliminated in the course of a correction process to be expounded later. Corrections had to be performed anyway since the selection has taken place by dwellings and on account of the different number of inhabitants, the number of persons to be include became a probability variable. As opposed to 1960, the practice was different in the respect that while the-shaping of the sample took place centrally ten years ago, the task was now performed by the cooperation of the county Directorates or the Statistical Directorate in Budapest. Nevertheless, the testing of the sample survey has been carried out centrally. The substance of the sample survey consisted of the Registration forms of dwelling and the Personal inquiry sheets which have been copied by the county (or Budapest) Directorates. The copies contained the whole material recorded on the general questionari.es (A and C formulas), but did not refer to the supplementary questions shown in the questionari.es of 25 per cent representativity (B and D formulas)

. The correction of the sample The correction procedure consisted of the.inclusion of a set of suitable flats to the sample as well as of the removal of another set of flats of appropriate size. The correction has been performed according to the preliminary number of the flats and inhabitants as published in the first volume of the series, namely adjusted for the "present" population figures. The final data to be published later as the results of the full-scale data processing, are of course, different, on account of the subsequent registration of persons omitted on the one hand, and of the corrections made during the supervision, on the other. 4. Népszámlálás 49