Selling Confusion

For the second straight week, the House Special Subcommittee on Investigations poked and prodded at the A. C. Nielsen Co., the colossus of the television rating industry.

Nielsen, which collects 90% of all dollars spent on national radio and television ratings, knew it was the committee's prime target, and its executives came to the hearings armed with a vanload of statistical charts. But the committeemen were not to be diverted by the long-winded, jargonized explanations of the Nielsen modus operandi. "You gentlemen amuse me," California Republican J. Arthur Younger told the Nielsen men. "I have never yet seen anything that sells confusion before like you people do."

Lights & Buzzers. Yet from the confusion, a picture of Nielsen's operation slowly, fuzzily emerged. The company's "sample universe" is peopled with two species of audience: Audimeter families and Audilog-Recordimeter families. In some 1,100 U.S. homes (selected by computer), all radios and television sets are monitored continuously by Audimeters--black boxes about the size and shape of a car battery. Each Audimeter comes equipped with eight weeks' worth of film, which records the family's listening and viewing activity. When a spool of film is replaced (either weekly or every other week, according to Nielsen's need for speed), the Audimeter rewards its keeper by ejecting two quarters (Nielsen also pays half the family's TV repair bill). The film is mailed to headquarters in Chicago, where its coded streaks and dots show what channels the set was tuned to, when, and for how long. By correlating this information with program schedules, Nielsen's tabulators and a clacking IBM 1410 computer crank out the ratings that spell life or death for most shows.

The Audilog-Recordimeter family has a more challenging role to play. Planted in some 12,000 homes, the Recordimeter, a small, clocklike instrument, tots up the number of hours the radio or television set is operating. But it cannot tell what channel or station the set is tuned to. Every half hour, if the set is on, the Recordimeter briskly rouses the absorbed or snoring viewer by flashing a white light behind the picture tube. Radio listeners are alerted by a buzzer. At this signal, the viewer is supposed to pick up his Audilog, a soft-backed book with a page for each day's viewing, and record the time, the channel he is watching, and the channel changes he has made. The A-R family is rewarded with $1 a week for allowing itself to be called to work by lights and buzzers. The unanswered question in everyone's mind: Were families that consented to having these electronic watchdogs truly representative of all TV viewers? An ex-Nielsen field man testified that he once had to try 92 homes in Grand Rapids, Mich., before he could place a single Audimeter. On another occasion, in Washington, D.C., he rang 400 doorbells before finding anyone willing to take on the more laborious A-R chores.

In the Wasteland. The actions of the small Nielsen sample are extrapolated to determine the habits of the entire tele viewing and radio-listening population. How risky is this? Committee Investigators Robert E. L. Richardson and Rex Sparger had some jolting examples: -- In Texarkana, Ark., a woman "didn't like what Jack Paar said on his show . . . concerning the recent Meredith situation in Mississippi, so she turns him off every time he comes on, even though she likes the show." Since each Audimeter represents some 50,000 TV homes in the Nielsen projections, she cost Paar 50,000 "listeners" just by flipping a switch.

>An Audilog yielded this entry: "Turned TV on this morning so baby could watch it. I had too much to do to day because I had to go away for a while." Nielsen counted the baby's viewing times as valid in its rating equation.

>-A San Diego woman recorded in her Audilog that she kept her radio going most of the day, and "my dog enjoys it as much as a dog can." Nielsen's research division manager, Henry Rahmel, explained: "We don't count dogs in our au-'dience sample," but admitted that the entry was counted.

> An Audimeter in Louisville showed that one television set ran almost continuously for two weeks in November 1961, although the two channels to which it was tuned in that period broadcast only eight hours a day. Richardson wondered who had been watching, if anybody.

>Ten Nielsen TV families constitute approximately one rating point, and the loss of a point can be fatal. One producer testified that had he been able to get eleven more homes, one of his TV shows would not have been canceled.

>There is not a single Audimeter in the entire Rocky Mountain time zone (pop. approximately 7,000,000, including Denver, Albuquerque, and Salt Lake City). Nielsen argued that it did not matter, since Rocky Mountain opinions would not be different enough to make a significant change in the ratings.

Richardson moved on to the company's handling of the data itself. In a Nielsen report on a local television market, Recordimeter data must jibe with Audilog entries only one day in five to have that day counted as valid. In radio, though, if one day on a Monday-Friday basis is "bad," the entire period is discarded.

This "messing" with the sample, Richardson maintained, hurts the record of radio sets in use, since up to one-third of the radio diaries may be blanks. This pulls down radio ratings and helps push time buyers to other media. Also, individual stations can be hurt. One Louis ville report showed radio station WKLO in first place (with 21% of the listeners) during one Monday-Friday quarter-hour period. WAKY was second with 20% and WAVE third with 18%. But when Richardson added the valid days from 39 Audilogs that were dumped because of flaws, WAVE climbed to 23%, WAKY to 21%, while the erstwhile leader, WKLO, slipped to 19%. "It is obvious," Richardson said, "that this could be important . . . the way time buyers use this data."

Counting Eyeballs. As the hearings drew to a close, Arthur C. Nielsen Jr., president of the company, watched impassively from the hearing-room audience. "We have received some worthwhile suggestions," he said. "If they prove to be desirable improvements, we will be glad to make them." Board Chairman A. C. Nielsen Sr., who thought up the whole idea, was in Europe and stayed there.

But his equanimity did not calm the advertisers, whose faith in rating services was clearly shaken. Said Paul E. J. Gerhold, director of marketing services for Foote, Cone & Belding: "Restoring the confidence of business in television audience figures, after these hearings, may well require setting up a completely independent facility to produce and publish spot checks on the accuracy of the syndicated data."

At the annual convention of the Na tional Association of Broadcasters in Chicago, President LeRoy Collins declared: "It is incredible that the rating services have not instituted greater changes than they have, with all the indicated faults and weaknesses of their methodologies and services." Addressing the convention next day, FCC Chairman Newton Minow told the broadcasters he hoped the hearings "may encourage you to put more trust in the people and more faith in your own judgments of the public's capacity to respond to the best that is in you. I should hope that sometimes you would cancel the ratings and keep the programs. It's not accuracy I'm particularly worried about. I just don't think it's the function of broadcasters simply to count eyeballs."

This file is automatically generated by a robot program, so reader's discretion is required.