A fresh look at estimating readership frequency distributions

Date of publication: June 15, 2004


This paper focuses on the frequency approach to measuring both average issue readership and publications turnover/casualness rates, with an emphasis on the measurement of turnover (or casualness) figures, and finds a respondent's proportion of reading derived from the frequency question cannot be used as personal probabilities to predict future behaviour. While the frequency question may produce in some cases reasonable average issue readership estimates, some estimates are very wrong! The findings have major implications in evaluating the value of a print-media schedule or a multi-media (print, TV, radio, Internet, etc) schedule. Campaign schedules which use the frequency of reading question will in many instances use too many publications and an insufficient number of issues for the publications used.

  • PDF
  • This could also be of interest