Weryfikacja technik badania częstotliwości i intensywności oglądania telewizji w oparciu o dane z obserwacji uczestniczącej ukrytej

Abstract

Answers given by respondents to questions concerning their cultural consumption (TV watching, reading of books and magazines, cinema and theater attendance) are treated by the author as empirical indicators of actual facts assessed by means of hidden participant observation. The present article deals only with the questions concerning TV watching. Twenty six persons were observed during one month by 18 observers being their close relatives. Within four days after the observation period questionnaire interviews were conducted with the same 26 persons using closed and open-ended questions, relating to their self-evaluation of frequency and intensity of TV watching, by intensity being understood average amount of time devoted daily to television; one week later the same questions were asked in second interviews. Results obtained in the observation were compared with those obtained in the two subsequent interviews. The comparison has shown that the more detailed were frequency categories the less accordant with the observation results were the interview answers. In addition it turned out that a high degree (20 cases) of accordance (i.e. reliability) between the two interviews was not accompanied by a similar degree of accordance of the observation data with interviews, the conclusion being that the reliability of answers does not indicate their validity (truthfulness). As the six frequency categories used in this analysis were very detailed (everyday, six, five, four, three, and once/or twice a week), they were combined in three ways, to three categories, and further analysis was performed in order to see which way of reduction would yield the highest degree of accordance between the observation data and the interview answers. The most valid answers (i.e. most accordant with the observation) were those obtained with the following cafeteria: 1 — everyday or nearly everyday; 2 — several times a week; 3 — once or twice a week. Matrix analysis has revealed that with this cafeteria understatements and overstatements in respondents’ answers balanced each other. As regards the intensity of TV watching, the actual amount of time devoted daily to television was calculated by combining two kinds of information: observed acts of watching particular TV programms, and duration of these programms as recorded in press announcements. The comparison of the actual average amount of time devoted daily to TV, and the amount declared in interviews has shown that: 1° In both interviews the respondents strongly overstated their TV watching time which means that high interview reliability does not indicate the validity of the answers; 2° The correlation between actual intensity of TV watching and that declared in the first interviews was as low as 0,31. Various measures aiming at the improvement of validity of the information obtained in interviews are discussed and proposed such as: mailed fill-in questionnaire, monitoring letters, concretization of the time periods of TV watching referred to in the questions asked.