1970 HUNGARIAN CENSUS OF POPULATION Information on the data collection and processing (1977)
I. THE COMPREHENSIVE REWIEV OF THE DATA COLLECTION AND PROCESSING - 11. The control of the reliability of the census
The errors that can be observed in either positive or negative direction cancel each other, thus the inaccuracy of the data to be found in the statistical tables is considerably less than the gross errors showing the total wrong rubrics. The net error fluctuates between 0, 05 and 0, 82 by age groups while the lowest value of the proportion of gross errors is 0,53 and the maximum is 2,88 p. c. The latter value can be found in the age group 65-69. The index-number of the net errors of the total distribution by age can be obtained if we add these errors irrespective of their signs. In the five-year age groups the proportion of the absolute sum of net errors to the number of the registered persons is not more than 0, 24 p. c. and in large groups it is not more than 0,10 p. c. The comparison of the results with the results of the census of 1960, studied by Kálmán a/ Tekse, is very meaningful. ' Table 4. ERRORS OF THE DECLARATION OF AGE IN A FIVE-YEAR AGE GROUP (1960, 1968, 1970) Number of persons under survey Gross errors Net errors Source Number of persons under survey No. perc. proportion No. perc. proportion Census of 1960. Members of worker-employee households 58 603 448 0,76 76 0,15 Members of not worker-employee households 3 952 84 2,12 27 0,68 Microcensus of 1968 19 419 238 1,22 52 0,27 Census of 1970 26 163 386 1,48 64 0,24 In view of the two censuses the data of the whole population of 19 70 falls between the percentage error of the two occupation groups of 1960. The degrees of accuracy of the microcensus of 1968 and the census of 1970 are close to each other. The detailed data of Table 3 also show that no age group contains biassed distortion of the age reduced or raised to make a round figure. According to other sources a tendency to give an age higher may be observed at very old ages, as shown by a survey of the persons above 90. Educational level The basis census returns showing the education of the population have proved reliable to a considerable extent. In percentage distribution there are differences of decimal and in some cases - data of smaller order - centesimal percentage between the data of the census and the post-enumeration control. (Cf. table 5.) a/ Ibid. pp. 24-25, 34-35. 103