Radio Survey Bulletin,
Winter 2006
Small Market Reporting Enhancements
Effective with the Fall 2005 survey, Arbitron implemented a special initiative in non-embedded Condensed markets—i.e., Condensed markets other than: Frederick, MD; Morristown, NJ; New Bedford-Fall River, MA; and Stamford-Norwalk, CT (since the Metros of these four markets are embedded).*
The smaller sample sizes in Arbitron's Condensed markets (typically markets ranked 150+) have historically resulted in a "bounce" in estimates from one survey to the next. Working collaboratively with the Arbitron Radio Advisory Council, radio stations, and other customer groups, Arbitron developed a program that addresses this issue by utilizing a full year's sample in these markets, rather than the smaller sample from the traditional three-month Spring and Fall survey periods. The sample size for a single report is effectively doubled as a result of using the entire yearly sample from the market.
As an example, if a market had 500 diaries in Spring and 500 in Fall, rather than having a report based on a sample of 500, the new report in that Condensed market would be based on all 1,000 diaries. Eventually, those 1,000 diaries will be spread over 12 months of the year, rather than only during the traditional Spring and Fall survey periods. This approach results in doubling the reporting sample and eventually providing our customers in Condensed markets four reports per year.
Two-Step Implementation
To help ensure an expedient and smooth transition, this program is being implemented
in two steps:
Step 1—Fall 2005
With the release of Fall 2005 radio survey results, Arbitron will include diaries
from the Spring 2005 and Fall 2005 surveys when tabulating audience estimates,
and will report a two-book average based on the Spring 2005 and Fall 2005
surveys, in (non-embedded) Condensed markets. Each report thereafter
will be a two-book average of the current (Spring or Fall) survey and the
most recent previous (Spring or Fall) survey, delivering two times the current
sample size for audience demographics and dayparts.
* Notes:
Single-survey Fall 2005 audience estimates, rather than combined Spring 2005-Fall 2005 audience estimates, will be released for the following non-embedded Condensed markets:
- The Florida Keys, FL – This market became a new market effective Fall 2005.
- Rocky Mount-Wilson, NC – This market became a new market effective Fall 2005.
- Sebring, FL – This market will become a Standard market effective Spring 2006.
Single-survey Fall 2005 audience estimates, rather than combined Spring 2005-Fall 2005 audience estimates, will be released for the two markets that were Condensed markets for Spring 2005 and became Standard markets effective Fall 2005:
- Santa Barbara
- Trenton
Step 2—Targeted for 2007
Beginning 2007, ratings reports will be issued quarterly (four times per year)
in Condensed markets. The same combined number of diaries formerly placed
during the traditional Spring and Fall survey periods will be spread over
all 12 months of the year. Each quarterly report will add listening from
diaries from the most recent quarter and drop listening from diaries from
the oldest quarter.
Level of Information and Granularity
The full six-month report period in Step 1 and the full 12-month report
period in Step 2 will be the standard report periods for (non-embedded)
Condensed markets. While estimates based on these report periods
will be considered the standard for audience estimates, estimates for
three-month intervals in Step 1, and for three-month and six-month
intervals in Step 2, will also be available on a more restricted
basis.
Estimates for intervals of less than a year, as outlined above, will be available to Maximi$er and Media Professional Respondent-Level Data subscribers. To provide greater insight into programming changes in the marketplace, Respondent-Level Data subscribers will be able to look at a variety of listening estimates for three-month and six-month periods, to capture a snapshot of the change in listening patterns. Although there will be restrictions on the available dayparts and demographics for these shorter periods of time, this level of granularity will give Maximi$er and Media Professional subscribers the opportunity to see the high-level impact of programming changes.
Updated Maximi$er and TAPSCAN Software
New Maximi$er 10.1 and TAPSCAN™ 9.4 software mails
to subscribers of the respective services during January 2006. Although the
revised software contains key
service updates for subscribers in all markets, subscribers in markets subject
to the Small Market Reporting Enhancements must install the software
in order to run data for Fall 2005.