Abstract:

Many newspapers have a distinct and regular readership pattern across the days of the week, a function of general newspaper readership patterns and of differing editorial content. The current press readership model, based on 'average issue' measurement, does not adequately measure daily audience fluctuations. This paper addresses two issues: first the collection of the readership of daily newspapers and their supplements on a day-by-day basis; and secondly the use of that data to estimate schedule reach and frequency. The authors report how a 'bi-source' method of data collection was used in Sweden to estimate variations in daily readership levels and how the traditional binomial model used within the SESAME media analysis software was significantly modified to allow for conditional dependencies of reading a newspaper on successive days of the week and for the reading of the parent newspaper and its supplement.