Czynniki powodujące tendencyjność w próbie udziałowej uzyskanej poprzez ankieterów społecznych

Abstract

The following hypothesis was formulated: the behavior of interviewers of the Section for Study and Evaluation of the Program at the Polish Radio brings about a bias in results of surveys on the declared tastes and preferences concerning the literary radio broadcasts among manual workers and peasants. Two kinds of biasing factors were hypothesized as coming into play: 1° – those coming from the selective choice of persons for interviewing (even within the compass of demographic traits determining the quotas, as when an interviewer selects persons more intelligent and familiar with the topics dealt with in the survey); 2° – those connected with the influence of a person of an interviewer in the course of interviewing, when his/her behavior betrays his/her being a member of intelligentsia and, in consequence, associates itself in the mind of a worker or peasant with a higher educational level as well as with more sophisticated cultural tastes. The influence of such a “cultural suggestion” would make interviewees declare their literary tastes on a higher level than they actually are. To test this hypothesis a secondary analysis of answers to five questions in the 1963 survey on listening to literary broadcasts was made. Analysed were answers of 438 radio listeners, workers and peasants (men and women separately), by various cultural-educational levels of interviewers as well as by various ways they selected the persons for interviewing (These characteristics of interviewers were obtained by means a separate questionnaire sent to them at the same time). The results of analysis corroborate the hypothesis, making also possible its quantitative refinement. It turned out that the factor of selective choice works fire strongly than that of a “cultural suggestion” in the direction of over-representation in the sample of the declared higher level literary tastes. The influence of “cultural suggestion” manifests itself more markedly answers to the closed or pre-coded questions (and particularly so when the latter contain additionally a semantic suggestion) than to the open-ended questions – or even to closed ones but worded more correctly. It was also discovered that the higher the cultural-educational level of an interviewer, the higher the level of the declared school education of an interviewee. In addition, it follows from the analysis that the influence of “cultural suggestion” appears only among workers and peasants declaring their educational level higher than the elementary one. This seems evidence that it is the higher level of their cultural aspirations that makes them over-estimate their literary tastes with regard to Radio broadcasts. The analysis of answers of 446 interviewers to the open-ended question the most frequent motives of refusals brought the following results: lack of time (26%); lack of interest in the survey (24%); feeling that surveys are useless as the Radio does not take into account the tastes its audience (15%); fear (15%). The last two motives were mentioned more often by interviewers being manual workers or peasants themselves % and 19%, respectively) than by those being white collar workers (13% and 13%). Other motives, like shyness, resignation from expressing y opinion, critical attitude toward a scientific value of survey results, negative reaction to the violation of a private character of one’s opinions, were mentioned but rarely. The final conclusion of the analysis is that quota sample surveys bring results that considerably over-estimate the actual level of literary tastes among more cultivated manual workers and peasants.

83-110