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 gross error of the total distribution of educational level (3. 49 percent) and net error (0, 72 percent) are the lowest in addition to those of the age among all the criteria under survey. Though there are larger differences in some types of schools, mainly in the rubrics of vocational secondary schools and the graduates of the training colleges for kindergarten teachers, percentage distribution is adequate here too. Table 5. INDEX-NUMBERS OF THE RELIABILITY OF THE DATA OF EDUCATIONAL LEVEL OF THE CENSUS Percentage dist. Difference between percentages Identically grouped items Gross Net Educational level P. C. Survey Difference between percentages Identically grouped items errors Index of inconsist. Educational level Census P. C. Survey Difference between percentages in the percentage of the census returns Index of inconsist. No certificate of final exam. in. sec. school 88, 54 88, 66 - 0,12 99, 59 0,95 - 0,13 0, 083 Cert, of final ex. from: Grammar-school 6,10 5, 88 + 0,22 90,69 14,91 + 3, 71 0,159 Secondary techn. sch. 3, 76 3, 71 + 0, 05 88, 67 21,33 + 1, 33 0, 222 Vocational sec. sch. 0, 39 0, 33 + 0, 06 62,38 60,40 +14,85 0, 606 Teachers' train. coll. 0, 60 0, 59 + 0, 01 88,46 20, 51 + 2, 56 0,206 Kindergarten teachers' tr. col. 0, 09 0,10 - 0,01 95,45 27, 27 - 18,18 0, 273 Technical 0,09 0,11 - 0, 02 69, 57 86,96 - 26,09 0,870 Other 0, 31 0, 51 - 0, 20 72,84 119,75 - 65,43 1, 201 Higher ed. level 0,12 0,11 + 0, 01 53,13 84,38 + 9, 38 0,845 Total 100, 00 100, 00 98, 26 3,49 a/o, 72 b/o,ioo a/ Sum of the absolute numbers of net errors, compared with the absolute numbers of persos. - b/ Weighted aritmetical mean. Economic activity In connection with activity (Table 6) some uncertainty may be observed in the judgement of active wage-earners and dependents: this is shown by the differences of -0, 63 as well as +0, 59 between percentage distributions (summarizing the numbers of active wage-earners broken down into two items). The gross error is larger in the smaller groups, thus in the groups of persons on maternity leave and especially temporarily occupied active wage-earners. Evidently this latter concept was more difficult to interpret during the survey. As a result of all this the gross error of activity distribution is 10,15 percent. However, it is reassuring that the net error is only 1, 28 percent, namely they are due merely to carelessness and misunderstanding and not to bias - they mostly cancel each other in the larger groups at least. Thus the net error affecting the results published is only one eighth of the gross error. 104