The Council for Research Excellence, an advertising industry and academic body created last year by Nielsen Media Research, has announced a new initiative.

In a project funded by Nielsen, the CRE will shortly begin the first detailed study into one of the fundamentals of television audience measurement: the quality of the research samples from which the ratings are drawn.

The study will explore 'non-response bias' in Nielsen's national and local samples, including People Meters and Diary methodologies.

Most importantly, it will seek to understand if there are differences in viewing behaviour between those households or persons that agree to participate in Nielsen's samples and those that refuse to do so.

The key question is: do these differences have a material impact on reported ratings?

The study will be questionnaire-based, making significant use of incentives and its results could lead to new methods of panel recruitment, according to Nielsen chief research officer Paul Donato.

Led by Ceril Shagrin, svp for corporate research at Univision, the study is expected to take two years to complete at a cost of more than $1 million (€810.4k; £561.3k). Says Shagrin: "Since the ratings are used as currency for more than $70 billion in television advertising, we need to make certain that these estimates are as accurate as possible."

CRE members include clients such as Procter & Gamble , Ford Motor Company and Anheuser Busch; media owners (Fox Network, Viacom, and NBC Universal); while among the media agencies represented are ZenithOptimedia, OMD and Starcom Mediavest

Data sourced from (UK); additional content by WARC staff