Onderzoeksartikel

Gevolgen van de introductie van STER reclame: een voorbeeld van Box Jenkins interventie-analyse

---

Abstract

In this article an analysis is presented that serves as an illustration of the capacity of the Interrupted Time Series Design and Box Jenkins models for complex questions about the consequences of the rise of new media, especially as income from advertisements of the old or traditional media is concerned. A historical ’’intervention” of a new medium on the amount of advertisements in the Dutch national daily press took place in 1967 when TV advertisements (STER) were introduced. It has often been tried to establish the exact loss of the daily press in consequence of the introduction of the STER but most results were invalid or contradictory. Complications originated often from a lack of data; time series based on 10 or 15 yearly measurements were used for regressions analysis with obvious consequences for statistical conclusion validity and statistical power. A new time series of monthly data concerning the amount of advertisements in the Dutch national daily press (1960-1982) was used to see if Box Jenkins analysis could produces more valid results. Two intervention models were tested. The model postulating that every yearly enlargement of STER air time caused an immediate and permanent decrease in advertisements appeared to fit the data, as well statistically as substantially. A total loss of 151 units of 1000 ’’column-millimetres” was calculated (monthly loss for the average of the Dutch national daily press). It is concluded that the Interrupted Time Series Design and Box Jenkins models provide a promising research strategy to tackle more general intervention and effect questions originating from the rise of new media.

How to Cite:

Van Zoonen, L., (1987) “Gevolgen van de introductie van STER reclame: een voorbeeld van Box Jenkins interventie-analyse”, Massacommunicatie : Wetenschappelijk Kwartaaltijdschrift voor Communicatie en Informatie 15(1), 81–92.