Arbitron vs. Nielsen on recruiting


Nielsen’s entry into the US radio ratings business has raised anew the question of how to recruit people to participate in a sample group for broadcast ratings. Arbitron has resisted calls to expand its address-based recruitment from Houston to other Portable People Meter (PPM) markets, citing the high cost. Consultant Randy Kabrich, long an Arbitron critic, contends that Nielsen will have a better sample in 50 small radio markets than Arbitron has in any of its PPM markets, save Houston.

This is not exactly and apples-to-apples comparison. The recruitment methodology that Arbitron uses in Houston, which was created in consultation with Nielsen when the two were considering a PPM joint venture, recruits sample from an address-based database. If the household can’t be reached by phone or mail, someone actually goes over to knock on the door. That, an Arbitron spokesman noted, is much more expensive than what Nielsen plans for the 50 small to medium markets where it plans to launch radio ratings next year.

The address-based sample (ABS) that Nielsen is using for television and will implement in its new radio markets as well works to match a listed, unlisted or cell phone to the selected address – and then uses mail for contact if the household can’t be reached by phone. No one goes knocking on any doors.

Even so, Kabrich says it is better than Arbitron’s approach because it begins with a database that includes 100% of all cell phone-only (CPO) households. “That eliminates the cell phone-only distortion that Arbitron has,” he said.

Not surprisingly, Arbitron begs to differ.

“Arbitron still believes Random Digit Dial (RDD) is a better system for reaching landline households at this time for a few strong reasons.  From our own testing over the years and test results that we’ve seen from other companies, response rates are generally lower with an address frame.  One notable study showed that Hispanic representation dropped when the address frame was used.  The address frame does not cover group quarters and it can also miss what are known as ‘drop point’ addresses, such as trailer parks and some gated communities.  And while the reports we’ve seen have shown improvement in young adult performance, we’re not certain how much of that is due to having cell phone only households in the sample compared to a sample frame that does not cover CPO households versus actual improvement in the measurement of young people,” said Ed Cohen, Vice President, Research Policy and Communication, Arbitron.

“Arbitron has conducted multiple tests of address frames in recent years and it was only after reviewing test results as well as seeing what others had reported that we came to the conclusion that the hybrid (RDD for landline and address frame for CPO) was the best method at this time,” Cohen said in comments sent to RBR/TVBR.

Arbitron recently touted its success in recruiting cell phone-only households in its PPM markets. That’s moving the sample toward 10% from cell phone-only households, up from 7.5%. But with cell phone-only households approaching 20%, Arbitron is still underweighted. “They’re not getting even half of the cell phone-only,” charged Kabrich. 

50 of Arbitron’s diary markets are scheduled to have cell phone-only households added to the mix beginning with the Spring 2009 survey.