Onderzoeksartikel

Patronen van printmedia-gebruik 1975-1995: een latente-klasse analyse van veranderingen in lezerstypen

---

Abstract

Using data from the 1975&1995 Dutch Time Budget Surveys latent class analysis (LCA ) was applied to identify through their reading patterns the set of mutually exclusive latent classes of readers. Newspapers (quality and popular; nationwide and regional) and consumer magazines were each clustered into three categories according to their focus on information or entertainment & human interest. Model selection with LEM (Vermunt 1997) resulted in a model with one latent variable and five classes for both 1975 & 1995. This model informs about respondents’ conditional probabilities of scoring Yes on an item, given the class to which they belong. Simultaneous latent structure analysis in the 1975 and the 1995 groups revealed that the structure (that is, the conditional probabilities) of the classes across groups was similar. The five types of readers corresponding to the classes are entertainment readers, information searchers, regional readers, non-readers, and omnivores. The first three types (size in 1975 about 25, 10 and 25%, respectively) are characterized by an almost exclusive focus on one of the three levels of the reading items. Nonreaders represent 32%, while the omnivores account for 11% of the respondents. In 1995 the population was distributed differently across the latent classes (22, 11, 12, 35, and 17%), indicating that the class of regional newspaper readers was halved, while that of non-readers and of omnivores both increased. The decrease in reading behavior becomes manifest only from the changes in conditional probabilities. The class of omnivores may have increased in size, but compared to 1975 the conditional probabilities of 1995 respondents’ scoring yes decreased for all reading items, except for quality newspapers. The same holds for the class of entertainment readers and that of information searchers. A multinomial logit model was used to test hypothesized causal relationships between kinds of reading pattern and five covariates (age, education, gender, occupation, book reading).

How to Cite:

Van Eijck, K. & Van Rees, K., (1999) “Patronen van printmedia-gebruik 1975-1995: een latente-klasse analyse van veranderingen in lezerstypen”, Tijdschrift voor Communicatiewetenschap 27(2), 125–151.