Rozmiary i kierunki tendencyjności w ankiecie prasowej

Abstract

The present article is a comparative analysis of results of two research works on the popularity of „Głos Szczeciński“ (The Voice of Szczecin) daily paper. One of them was based on questionnaires attached to the paper and returned by riders by mail, and the other on interviews with questionnaires with residents of Szczecin selected at random. The questions put in the two questionnaires were the same: they concerned, among other items, bow often paper was bought, opinions as regards the paper, and preferences as regards various columns and articles. Afterwards, the results obtained in the two investigations were compared in order to ascertain the differences in answers to particular questions and thus to verify following hypotheses; among there who answered the questionnaire a) men, b) people of good education, c) long-established residents of Szczecin, d) the middle-aged, e) regular readers, f) members of political parties, g) individuals with clearly established opinions as regards the content of the daily – will be over-represented. The comparative analysis did indeed confirm most of these hypotheses. No tendentiousness in answering the questionnaire was established as arising from age structure and- the duration of residence in Szczecin. The structure of bias indicated above is explained by the mechanism of the shaping of such phenomena. They are a result of the cumulative effect of objective and -situational factors (the more often the reader purchases the particular daily, the more likely is he to come across the questionnaire), and factors of autoselection (readers having certain social and demographic characteristics who come in contact with the questionnaire are more willing to complete and mail it); the correlation between, the frequency rot reading (or buying) the daily, and similar features account for the very marked over-representation among those who answered the questionnaire, of party, members, men, and people having a higher standard of education. On the basis of a series of similar investigations concerning the direction and range of bias in self-returning questionnaires, it has been possible to establish, a coefficient of tendentiousness in relation, to each of the questions according to the formula: [SEE summary_Gostkowski.pdf]

where a = the percentage of answers to the questionnaire, p = the percentage of answers obtained in interviews with the representative sample of readers. Knowing the coefficient Wa, it is possible – on the basis of data from the questionnaire – to establish the percentage of answers to be found in the representative research according to the following formula: [SEE summary_Gostkowski.pdf]