Arbitron Cites PPM Enhancements

by Mike Boyle of Radio and Records Magazine

Customers get 'increase in transparency'

In accordance with its recent agreement commitments with the attorneys general of New Jersey, New York and Maryland, Arbitron says it has committed the same key methodological enhancements to all PPM markets.

While there is variation in metrics and milestones based on particulars of the local market environments, all Arbitron PPM customers are in the process of receiving enhancements to the PPM methodology in four key areas: cell-phone-only sampling, address-based sampling, in-tab compliance rates and response metrics (Sample Performance Indicator or SPI).

In addition, customers in all PPM markets will be seeing an increase in transparency for more of the sample metrics in the Arbitron PPM survey research, including the distribution of sample by ZIP code and by cell-phone status. Arbitron will also continue to share with all customers any current and future findings of the impact of non-response on the PPM service.

"I am committed to continue our focus on the concerns of our PPM market customers in terms of the sample size, sample proportionality and sample quality of our PPM panels," Arbitron president/CEO Michael Skarzynski says. "I want to assure the industry that Arbitron is actively engaged in the continuous improvement of our PPM and diary market services. Thanks to specific customer input, Arbitron is already making good progress on improvements in many operational areas. As an integral part of the company's continuous improvement programs, Arbitron is on track to meet or exceed 100% of the settlement agreement criteria with the attorneys general of New York, New Jersey and Maryland."

Arbitron's key methodological enhancements include:

* Arbitron will increase the sample target for cell-phone-only households in all PPM markets to an average of 15% by year-end 2010. In an interim step, the current target of 7.5% will be raised to 12.5% in PPM markets by year-end 2009. (Monthly milestones within each year will vary by market.)

* In keeping with the commitment to increase cell-phone-only sample in all markets, Arbitron is committing to the use of address-based sampling technique for at least 10% of its sampling efforts by late 2009 and for at least 15% of its recruitment efforts by the end of December 2010 in all PPM markets. (Monthly milestones within each year will vary by market.)

Arbitron has already begun using addressed-based sampling for cell-phone-only households in PPM markets. The initial returns from the first round of address-based sampling are indicating that markets such as Nassau-Suffolk and Middlesex have a cell-phone-only penetration that is significantly less than 15%.

* Arbitron is applying its average-daily in-tab benchmark of 75% of installed sample to all PPM markets. Arbitron is currently implementing in all PPM markets many of the techniques and methods that Arbitron is using to minimize differential compliance rates in order to fulfill the terms of the New Jersey, New York and Maryland agreements. (Specific programs and techniques are being applied based on an individual market's in-tab performance.)

* The PPM markets that were first recruited in 2008 have received the benefit of the latest response rate and compliance programs starting with the very first recruitment contact. In these markets, SPI metrics are already above the commitments to New Jersey, New York and Maryland.

Arbitron will continue its focus on improving the SPI in all PPM markets. Based on the progress to date in the company's response rate programs, Arbitron has informed the MRC that it hopes to achieve a 21% SPI average across all PPM markets by the end of 2010.

* The company's commitments to New Jersey, New York and Maryland include greater transparency in terms of information regarding PPM sample composition and other metrics of PPM services. Arbitron intends to provide this additional information in all PPM markets that many of its clients have indicated they would find useful and would help their understanding the strengths and limitations of PPM samples.

* Arbitron is providing in all PPM markets installation and in-tabulation data by individual ZIP code, along with market population data for blacks, Hispanics and others (non-black or non-Hispanic) for each individual ZIP code.

* Arbitron will continue its practice of publishing the percentages of installed and in-tab cell-phone-only sample in the front of all monthly Arbitron electronic PPM market reports.

* The company intends for all PPM markets to benefit from the findings of a non-response bias study that it will conduct in the New York market by July 15. Should the findings indicate any measurable statistical bias, Arbitron will inform all clients of the identified sources of that statistical bias in the PPM methodology and will take reasonable measures to address those sources.

* The option to see the current composition of the Hispanic sample by country of origin has been a topic of discussion with Hispanic broadcasters. Arbitron has conducted a test of three markets selected by the Spanish Radio Association (Houston, Los Angeles and New York) to determine if asking the country of origin question during one of the regular status calls to PPM households would have an adverse impact on the PPM sample. Early results from the test are indicating that Arbitron would be able to collect and track country of origin without affecting the quality of the sample. The decision about moving forward with country-of-origin tracking is pending a full analysis of the test.