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

The size of the sample (the number of the elements) The rate of selection of the sample - similarly to the census of 1960 - was 1 per cent. The size of the sample was fixed according to the following considerations: The size (the number of the elements) influences, to a large extent, the accuracy of the results obtained by representative techniques. The application of a large number of elements should be aimed at, on the one hand, in order to get more elaborate results, while the aim at the quick data processing calls for a rather small-scale sample, on the other. At the same time, the requirement should be considered that the procedures of the full-scale data processing be not delayed or hindered by the operations connected with the sample survey. In order to weigh the pros and cons of these contradictory aspects, a sample size seemed to be justified which ensured an acceptable accuracy of the estimation of as many nation-wide data as possible, as well as the breakdown of the county-wide data, according to as many criteria as possible. The breakdown according to smaller territorial units had to be abandoned. In considering the practical requirements of the sample data, under the term acceptable accuracy" of the data is meant that the relative standard error of the majority of the data included to the smallest sample does not exceed 5 per cent on the 95 per cent probability level. In the case of the number of elements included to a 1 per cent sample, this requirement is satisfied if the data occur with 1, 5 per cent of relative frequency at least. It means that the sample enables us to estimate - within the confidence intervals as above - nation-wide data (projected data) concerning 150 000 persons, i.e. a relative frequency of a sample including 1 500 persons observed. The unit of the sample survey Complying with the method of the registration, the selected unit was the flat. This kind of selection was more favourable from practical points of view than the individual selection which would have increased the time spent in registration work and supervision. Besides, from the aspect of the housing, household and family statistics, it is more advantageous if the questionaries of the persons belonging to the same family or household are not taken apart. Individual selection has been applied in the case of persons belonging to the institutional households (nurseries, children's homes, students' and workers' hostels, hospitals, hotels, resthouses etc.) It means that every lOO**

1 person was included to the sample. The sampling technique and its practical performance When shaping the representative sample, a "systematic" sampling technique was used according to which, taking the material ordered in the sequence of the counties and districts (towns), then the communities ordered in alphabetical order, finally taking in turn the districts, streets and th street-numbers - the registered material belonging to each 100 flat became included to the sample. The accuracy of the sample might be affected by the mutual dependency of the elements of the population to be included to the sample. In such cases, a correlation exists between the single 48