Who’s in First? Rival Ratings Services Tell Different Story : RATINGS: Surveys Sometimes Differ

During the recently concluded November ratings sweeps, syndicated reruns of “The A-Team” on KHJ-TV Channel 9 were the most-watched television show in Los Angeles between 6:30 and 7 p.m., beating everything from old “Family Ties” episodes to Dan Rather and the evening news.

It was, that is, in the Arbitron ratings.

In the ratings survey from the A.C. Nielsen Co., “The A-Team” was dead last among the seven most prominent local stations. KTTV Channel 11’s “Family Ties” was No. 1.

This conflict over the current popularity of Mr. T and his gang was only the most glaring discrepancy between the two ratings services for the November sweeps that ended last week. Two different winners were produced again in the local news race: KABC-TV Channel 7 in Arbitron and KNBC-TV Channel 4 in Nielsen.

Depending on who you talk to, these variations between the two reports, which purport to measure the viewing habits of the same 4.8 million households over the same four weeks through a survey of 500 homes, are either nothing to be alarmed at, terribly confusing or symptomatic of how inadequate and inaccurate both ratings services are.

“With the diversity of the L.A. market, you just can’t do justice to a population of this size,” said Felix Gutierrez, a journalism professor at USC who teaches graduate courses in survey research. “If you’re trying to divide such a small sample between different age, racial, lifestyle and language groups, each sample household within each segment of the population carries more weight than it should. If one Hispanic changes the channel, it’s like 300,000 just did.”

While advertisers, station executives and even representatives of the two ratings services agree that the current measuring system is anything but an exact science, however, no one seems willing to dismiss it entirely. It is, they all say, the only game in town for gauging how many people are watching a particular program and determining how much advertisers should pay for a commercial on the program.

“You could never put a man on the moon with this kind of research,” said Elaine Kaplan, senior vice president of Western International Media, the No. 1 buyer of advertising time in Los Angeles. “But you create your own monster, and now you have to live with it.”

Indeed, six of the seven major commercial stations in Los Angeles subscribe to both services. KNBC-TV Channel 4, which for years has been rated lower by Arbitron than by Nielsen, discontinued using Arbitron numbers two years ago and refused to renew its subscription with the company last October, saving the station approximately $1 million a year, according to John Rohrbeck, Channel 4’s general manager.

The real problem with the current ratings system, said Terry Pittman, research director at KCBS-TV Channel 2, is that many of those who use the numbers try to push them beyond what they are designed to do. In their quest to be able to promote themselves as No. 1, stations crunch the numbers down to a tenth of a point--when the fact is, he said, that the margin for error in these surveys is generally one rating point.

For example, in the 4 p.m. time period last month, the Nielsen ratings showed Channel 4 on top with a 7.2 rating, while Channel 2’s “Geraldo” and the Channel 7 news tied for second with a 6.6 mark. A clear victory for Channel 4? Not really.

Ratings experts said that because of the margin for error, 7.2 and 6.6 are essentially the same number. In their final books, both Arbitron and Nielsen round off their ratings and would report all three as 7 (each local rating point represents 48,002 homes).

Advertisers, Kaplan said, usually round the numbers down, in this case to a 6, when considering where to place their ads and how much they are willing to pay.

Still, none of that explains why Channel 4 does significantly better in Nielsen or why “The A-Team” is rated nearly twice as high by Arbitron.

Executives from both Nielsen and Arbitron insist that the techniques they use to gather their ratings data are scientifically sound, and both are confident that their results are “absolutely correct.” Both use samples of approximately 500 households, both claim that every household in the area has an equal chance of being selected, and both insist that their samples properly represent the diversity of the market.

The easiest explanation for the discrepancies between the two is simply that each monitors a different set of 500 homes. Roy Anderson, executive vice president at Nielsen, said that if Nielsen monitored two sets of homes using the exact same methodology, it would come up with two different sets of results.

But there are some differences in methodology that might explain the difference in the results, according to Bill Shafer, Arbitron’s vice president of television sales. Nielsen monitors its sample of homes and simply reports the data as it has been collected. Arbitron gives differing weights to certain households depending upon the demographic makeup of its reporting sample.

Based on statistics supplied by a research company in New York, Arbitron believes, for example, that Hispanic households comprise about 20% of all households in the local viewing area. If in its daily measurements Arbitron finds that only 19% of its sample is Hispanic, then the results from each of those Hispanic homes would count a little more than the results derived from every other kind of home. These adjustments, up or down, are made for several age and ethnic groups.

Over the years, Arbitron has also been known for skewing toward younger viewers. Though KCBS’ Pittman said this difference is much less prominent today than it has been in the past, Shafer, in trying to explain why “The A-Team” was rated so much higher in his service, pointed out that Arbitron prides itself on including in its sample a representative number of teen-age and young adult households--the most difficult demographic category to measure, he said.

“I don’t think either service does anything that makes it better than the other,” Pittman said. “It’s a matter of art. It’s like trying to decide if one tie looks better than another with your gray suit when they both match.”

Both Pittman and Shafer said that the discrepancy between the two was significantly less last month than it has been in the past several years. Shafer admitted that up until two years ago, Arbitron had too many meters clustered together in too few local communities.

KNBC’s Rohrbeck said that while Nielsen had spread its meters all across the area, Arbitron had set up 10 meters in Mission Hills and had zero meters in Santa Monica, Beverly Hills and several other wealthy communities on the West Side.

Though Rohrbeck said he still has “no confidence” in Arbitron’s research, Pittman said the company has solved this problem and now spreads out its meters across all zip codes in the L.A. market. He insisted that though the specific numbers may vary, both services last month showed similar trends in regard to which stations or programs are going up and which are falling off.

Advertisers, who collectively purchase more than $1 billion of local television spots each year, have their own opinions about who’s right. It is estimated that about 60% of the agencies rely on Nielsen, while 40% use Arbitron.

Kaplan said that while the total number of households watching particular programs are important to advertisers, the demographic makeup of the audience is often more critical. Even if the raw household numbers for two stations are close, Kaplan said, one station often gets the nod over the other because of the demographics--age of viewer, income of viewer--that it can deliver.

Though the accuracy of demographic information supplied by the two ratings services can also be questioned, Kaplan said people who buy ad time have the advantage of their own personal judgement based on experience, intuition, the atmosphere of the program, counterprogramming and time period when analyzing this data and making decisions.

Still, advertisers would prefer a more definitive system, Kaplan said. “It would be foolhardy to say that one can be totally secure with the tools that we have.”

The only way to diminish the discrepancies between the two services and produce a more accurate picture of what local viewers are watching is to increase the size of the sample. Advertisers, television stations and the ratings services themselves all say they would like to see this happen.

But research scientists point out that in order to cut the margin for error in half--from 1 rating point to .5--the ratings services would have to quadruple their sample to 2,000 households. And neither the television stations nor the ad agencies, who already pay a great deal for the data they currently receive, are willing to foot the bill.

“If I were placing ads for Vons or McDonalds, I’d have a lot of questions about these numbers,” Gutierrez said. “But until the people who buy ads demand it, until enough of them tell the ratings services that they can’t rely on these numbers anymore, everyone will have to make do with what they get.”

WHICH RATINGS WOULD YOU BUY?

Where the L.A. dial was tuned at 6:30 p.m. in November.

ARBITRON

Rank Show % of audience 1. The A-Team 13 2. Family Ties 13 3. Gimme a Break 12 4. NBC Nightly News 11 5. Magnum, P.I. 10 6. Eyewitness News 10 7. CBS Evening News 10 8. All others 21

A.C. NIELSEN

Rank Show % of audience 1. Family Ties 16 2. NBC Nightly News 14 3. Gimme a Break 12 4. Eyewitness News 12 5. Magnum, P.I. 10 6. CBS Evening News 9 7. The A-Team 7 8. All others 20