A Critical Assessment of the Frequency Question and its Ability to Provide Valid Measurements of Readership Figures: Lessons from a Study Involving a Voice Response Based Panel of Respondents

Abstract

The paper presents a new way of gathering data: A panel of respondents is instructed to perform a daily call to a voice response computer (VR) which has been programmed to pose pre-recorded questions regarding media contacts yesterday (day-after recall). Respondents are then providing answers using the buttons on their push-button phones i.e. they push the numbers of the titles read - these numbers figuring on a list containing approx. 50 titles which has been mailed to the respondents. The paper describes how the voice response technology might be used for media research purposes. The paper also discusses the construct validity of a frequently used readership measurement. The study involved indicates a considerable disagreement (low convergent validity) regarding readership reported when comparing data which were gathered (1) using continuous measurement (VR-method) with data gathered using the so-called (2) frequency question (mail-interview). Some crucial questions regarding a popular way of measuring readership are raised. Implications for advertising strategy are discussed.