Próba porównania odpowiedzi na pytania prekategoryzowane i otwarte w badaniach nad konsumpcją kulturalną
MetadataShow full item record
Surveys on “cultural consumption” (e.g. reading of books and newspapers, listening to the radio, watching the TV) pose a difficult problem of validity and reliability of research techniques applied. In view of a widespread tendency of respondents to overstate their cultural consumption (due to the influence of the positive model of a “cultured citizen”), it is necessary to formulate the questions in such a way as to minimize the bias. As a rule the researcher has at his disposal two alternative manners of formulation: he may choose either closed (pre-categorized) or open-ended questions. Very frequently the decision depends on researcher’s intuition. To ascertain the differences in answers obtained by means of both kinds of questions 50 respondents selected from among a sample of 298 persons were re-interviewed. Chosen were respondents with low educational level, interviewed by interviewers who did over 3 interviews daily, for it was suspected that in these cases the bias was most likely. In the first interviews which related to reading of books and newspapers, radio listening and TV watching, closed questions (with a cafeteria) were used. In the repeated interviews which took place two weeks later four open-ended questions on the same topics were asked from the same respondents. To secure their cooperation the interviewers explained to them that the interviewing was necessitated by the need to obtain some supplementary information. No refusals were ascertained. Main results of the comparison were as follows: 1. With regard to the question concerning desirability of subscription as a form of acquisition of newspapers, it was ascertained that 50% answers to the open-ended version were identical with those obtained by means of a closed version; 16% of respondents approving earlier of subscription expressed, in the open-ended version, their negative opinion about it. Furthermore, it turned out that 10 respondents were answering to the closed question automatically, i.e. without exactly understanding what the word “subscription” meant. 2. Answers to the questions with a series of pre-categorized alternatives embracing various elements of the contents read in the local newspaper were compared with the answers to the analogous open-ended question with intensive probing. The main difference was that despite very similar ranking of the content elements named by respondents in both questions the percentages of the reading declarations relating to each element were much lower in the open-ended version. The only exception to this rule was much higher rank of the “other” category (59%) in the open-ended version (with a zero in the pre-categorized one). This means that the “other” category in pre-categorized version did not ensure against serious omissions in answers of respondents. It is proposed to demonstrate in interviews a newspaper copy with simultaneous asking about the contents elements read, which procedure helps respondent’s memory. 3. The closed question about book reading within the last two months has brought less affirmative answers than its open-ended version because of the difference in the two wordings; in addition, in 30% of cases the answers were contradictory in both interviews. Respondents were also asked whether they considered it possible to recall the books read within the last year. Their opinions differed, depending on how many books they had read themselves. The conclusion is that questions about book reading must aim at helping respondent’s memory and not to refer to excessively long periods. 4. Answers to two cafeteria questions concerning the reception of the TV and radio programmes were compared with answers to their corresponding open-ended versions. Main conclusions were that: (1) the more popular was the programme item the more reliable were the answers of the respondent, (2) answers to the open-ended questions were scanty and too superficial, (3) "liking” in the pre-categorized questions meant just listening or watching while in the open-ended ones it meant an actual fancy, (4) the reliability level of answers turned out to be rather low, which fact points to the need of further methodological research in this direction.