Les recherches sur la varacite des responses obtenues au cours du recensement national

The author draws attention to the fact that there is a lack in Poland of systematic research and analyses specially designed to check the veracity of data obtained during the National Censuses. As an example she discusses the results of this type of research done in Yugoslavia and United States. Then she presents the results of her own empirical research on the veracity of data obtained during the last National Census of 1960 in a small town where, two years later, interviews with a random sample of 1004 inhabitants were done, by a group of students. A comparative analysis, based on 690 pairs: interview-census form, was made with relation to sex, age, marital status, number of children, school education and number of rooms occupied. It was discovered that the contradictions between interview data and the answer given in the census form concerned most frequently the education (153 cases) and the number of occupied rooms (123 cases). The former contradictions consisted most often in a difference of one school class and in 130 cases appeared, in persons with a primary education. It was more often in the Census that such respondents declared a greater number of classes (75 cases). Among 23 persons with a higher educational level who gave contradictory answers, it was found that they declared more classes in the interview than they did in the Census form. An analysis concerning discrepancies in the declared number of occupied rooms was supplemented by checking on the data from the files of the local Housing Bureau. It was found that: 1°a – the number of rooms declared in interviews was greater than that given in the Census forms; 2° – the number of rooms as indicated in the Housing Bureau registers (i.e. the actual state of things) was more often accordant with that obtained in the interviews than with the figures found in Census forms. The fact that the respondents declared a greater number of occupied rooms when interviewed by students, such declarations being more often accordant with reality, was hypothetically explained as follows: under the conditions of a permanent shortage of flats, accompanied by the state control over occupied rooms, there is a general tendency to paint one’s housing conditions in as dark colours as possible. The Census situation in which answers are obtained is highly official and forced, whereas during interviews performed by students, in a more spontaneous and informal way, the respondents in a lesser degree adopted such a defensive attitude and as a result they felt freer to declare a higher and truer number of rooms