• by Erik Sass  @eriksass1, August 24, 2007

Arbitron reissued its radio ratings from Houston for the last week of July in an attempt to calm panicked radio broadcasters. The Portable People Meter problems were traced to a software issue, rather than actual listening patterns.

While the news is probably welcome, the confusion surrounding the ratings foul-up has galvanized Arbitron's customers, who are increasingly vocal about the perceived shortcomings of its PPM system.

Houston-area broadcasters, and the industry in general, were alarmed by what appeared to be a steep drop-off in listening for the week July 26-August 1. After presenting several hypotheses, including the possibility that members of the sample group were going AWOL on summer vacation, Arbitron determined that the records of about 200 individuals had not been included in its weekly unified analysis.

Arbitron's chief research officer, Bob Patchen, said this oversight "should have been caught in the quality control process and was not." The process is currently under review.

Radio broadcasters, however, are still complaining about unrelated sample problems in Houston and Philadelphia. Patchen admitted that the Houston sample is still below the target size Arbitron set for itself when it rolled out commercial ratings in mid-July. Barring other explanations, the problem seems to lie with sample members who are neglecting to carry the PPM device, leaving it docked in its charger instead.

There is also a fall-off in Philadelphia, where the sample size for the 18-34 age cohort is just 197--roughly 55% of what it should be.

A spokesman for Arbitron said the company is implementing a number of measures to bring the samples up to their target size in both Houston and Philadelphia. These include increasing the weekly compliance bonuses for households that include a young person ages 18-24, and in-person coaching by Arbitron reps. In Philadelphia, the company is also planning to begin deliberate over-sampling of young households so the number of participants matches the in-tab target.

To restore sample integrity in both major markets, Arbitron must weed out non-participants and recruit more reliable replacements. In a conference call on August 15, Patchen said the Philadelphia panel would be fixed by the end of September, and Houston by the first half of October.

In a related issue, the National Association of Black Owned Broadcasters (NABOB) said last week that Arbitron's Philadelphia sample also under-represents African-American men ages 18-54--a key demo for urban broadcasters like Radio One.

NABOB chairman Jim Winston said that his organization early on expressed their reliability concerns to Arbitron, especially "by the low sample size used, particularly among African-Americans in the 18-24 demo. The samples in both Philadelphia and Houston for this demo have been consistently and substantially below the proportion of the population represented by this demo." The Arbitron spokesman also said the company is increasing weekly compliance bonuses for households that include a young African-American ages 18-24.