This paper outlines a new way of reducing the number of titles any one respondent is asked about, while maintaining, or even increasing, the overall media list. In essence, PMLÂ® is a technique to filter which titles a respondent is asked about, based on a small number of demographic and topic interest questions which are asked at the beginning of the interview. The complexity of the filtering is such that PML is only feasible using on-screen prompting. As well as being shown all the titles they have a high propensity to read, respondents are also shown a sample of titles they have a low propensity to read. Thus the majority of readership claims are collected directly from the respondent, and there is a basis to replace the relatively small proportion of â€˜missingâ€™ claims. Many different variations of the techniques used for PML are possible. This paper outlines a prototype, which will no doubt be refined as development progresses, along with the data already available.
Symposium: 2001: Venice, Session 5 - Data Collection Whats New?
Authors: Page, Katherine
Organisations: Ipsos MediaCT
Topics: Questionnaire Design