The Council for Research Excellence (CRE) this morning unveiled plans to study various TV audience measurement methods in order to better understand so-called "response bias" in Nielsen's
address-based audience samples. The CRE said it would test these methods in three-markets with the objective of coming up with methods to improve Nielsen's paper diary-based TV audience measurement
service.
The three markets include Dallas-Fort Worth (which is also a Nielsen local people meter market),
Albuquerque-Santa Fe. (which is also a Nielsen TV set meter market), and
Paducah, Kentucky-Cape Girardeau, Missouri-Harrisburg, Illinois (which is a diary market).
The CRE said the markets were selected due to the differences in their measurement methods, market
sizes, geographies and populations.
During Nielsen's May 2012 diary measurement period, a separate diary sample was selected for Dallas, while identified “non-TV homes” in all
three markets will receive a modified diary.
Sampled homes in all markets will be mailed a short questionnaire seeking answers on media equipment ownership and general viewing patterns. Homes
identified as not having a television set will be contacted and asked if they have any source of viewing television programming; those saying “yes” will be sent a modified diary in which
they will be asked to record what they view and the device used for viewing.
This new study is the second major effort for the CRE’s Sample Quality Committee (formerly known as the
Non-Response Bias Committee), following its 2009 study, “Measuring the Unmeasured Viewer,” was the most comprehensive of its kind to determine the impact of non-response on ratings --
revealing more about unmeasured viewers than any prior effort.
Findings from the new study are expected to be made available by first quarter 2013.