Nielsen: Finding Lost Boys

Nov 17, 2003  •  Post A Comment

Nielsen Media Research this week will deliver to its clients a white paper whose bottom line is expected to be that there is no need to correct anything in Nielsen’s methodology.
It’s a message guaranteed to anger the TV rating agency’s clients, who have long been frustrated that Nielsen is their only source of such data. They have been particularly troubled this season by steep declines in young male viewership.
Nielsen spokesman Jack Loftus said the white paper will include recaps of three months of data produced by Nielsen for itself and at the special request of clients, as well as analysis that supports Nielsen’s contention that the disappearing act pulled by 18 to 34 male viewers during the past three months is correcting itself and is not a reflection of a flaw in Nielsen’s National People Meter sample.
“The sample breakdown is at a record quality high,” said Mr. Loftus.
The season-to-date decline in 18 to 34 men’s prime-time TV usage was at 8 percent last week. “If you trend the data, it is right on trend for the past 10 years,” Mr. Loftus said.
That brought a quick reaction from some TV executives who have been troubled by the numbers, which ultimately translate into revenue from advertising based on the size of the audience.
“I can’t believe they’re actually saying that,” said one network executive.
“Their own numbers suggest it is not correcting itself,” said another.
Indeed, Nielsen is certain to find such intuitive assertions hotly debated by network research executives, who say none of the data given to them by Nielsen is consistent with such an optimistic view.
The debate over the missing young men came to a boil last week after an NBC analysis pointed to young Hispanic men who are new to the National People Meter sample as a possible explanation for skewed results.“I am not suggesting that what we found is the answer,” said Alan Wurtzel, NBC’s president for research and media development, who has been among the more vocal questioners of Nielsen’s methodology.
Then CBS suggested that Nielsen had laid the foundation for the steep defections by letting a disproportionate number of historically “problematic” survey participants-young men living in a household not headed by themselves-into the 5,100-home National People Meter sample this year as it attempted to meet client requests for a survey that better reflects the country’s demographic makeup.
David Poltrack, executive VP for research for CBS and UPN, asked Nielsen to compare the behavior of the men 18 to 34 who were in the national sample last year and are still in it this year. That analysis showed that in 40 percent to 50 percent of the unified sample, young men are actually watching more television and that declines in viewership are “concentrated in a relatively small percent of the overall sample.”
“Of all the men 18 to 34 in the sample, 20 percent are accounting for 70 percent of the decline,” Mr. Poltrack said. “Now, we look at that 20 percent and the first thing we notice about them is that the young men in that 20 percent are significantly more likely to live in homes headed by someone else, as opposed to young men living in their own homes.
“Throughout the history of the Nielsen system and of measuring audiences, this has always been the most problematic group, because these young men living at home with their parents, with other relatives, or whatever, tend to be involuntary participants in the survey. Someone else agreed to do the survey and they were dragged into it reluctantly,” Mr. Poltrack said. “As a result, they’ve been, as a group, the least diligent in their button pushing. They’ve caused the most problem for Nielsen over the course of this People Meter measurement era.
“When we see the decline concentrated among this particular group, the red flags automatically go up and we say, `Oh, them again.”’
When the issue first arose, Nielsen suggested one factor might be soldiers stationed in Iraq. However, that is not expected to be addressed in the white paper. What is likely to be mentioned are other explanations ranging from attendance at college (while the young man remains listed at home as a zero viewer) to increased use of other technologies. “If we took DVDs, video games and the Internet,” Mr. Loftus said, “it might account for less than half of it.”
The WB last week compared October diaries from the top 25 markets (which represent some 47 percent of the country) to data from the October national sample and found an unusually high level of discrepancy, even when the data was drawn from the same markets. The comparison showed that the level of men 18 to 34 using television was actually up 1.5 percent among those filling out diaries, while it was down 6.2 percent for the demo in the national metered survey.
Moreover, The WB’s analysis showed year-to-year declines among men 18 to 34 to be smaller in the local diary sample than they were in the national metered sample. For The WB, that translated into a 20 percent decrease nationally while the audience segment registered flat in the 25-market diaries. ABC was down 19 percent nationally and down 12 percent in the diaries. CBS was down 30 percent nationally and down 15 percent in the diaries. NBC was down 26 percent nationally and 14 percent locally. UPN was off 28 percent nationally and down 21 percent in the diaries.
The anomaly was Fox, whose strong postseason baseball performance was used to explain much of the displacement of young men in October. Fox showed a 44 percent increase in the diaries, which included five big-city baseball markets, vs. a smaller 39 percent increase in the national metered sample.
The WB data said the level of men 18 to 34 using television (otherwise known as the PUT level) in the diary samples was up 1.5 percent, while the national sample showed the same PUT level down 6.2 percent.
A 25-market “cutback” from Nielsen (a report on only those markets) showed the PUT levels at plus 1.2 percent and minus 7 percent, respectively.
Mr. Loftus explained the difference between the two samples as an indication that “the meter is more sensitive.”
“This leaves us feeling a little schizophrenic,” said The WB Television Network co-Chairman Garth Ancier. “On the one hand, we’re very excited for the success many of our stations are having selling, on average, basically the same ratings as last year, the best in our history; and at the same time, we’re scratching our heads as to why the national sample shows us down 19 percent year to year. So depending on which Nielsen sample you go by, The WB is either having a great year, or not so great, like many of the other networks to date.”
Andy Fisher, the CEO of Cox Broadcasting, intends to run his own comparisons in which the national sample during the November sweeps is measured against the diaries and local meters. He has been told by Nielsen that such data will not be available to him until after the November book is published.
The Atlanta-based group executive’s skepticism of local People Meter methodology is well-known. “The whole process is skewed to smaller and smaller numbers,” he said. He has frequently done local versions of Mr. Poltrack’s unified sample analysis in search of the reasons for sudden dislocations in Cox markets.
Mr. Fisher is pessimistic about the solidity of meter-produced data to accurately reflect viewing in younger demographic groups. “To me the people meter has always been a time bomb waiting to explode for younger demographics,” he said. “I consider it inevitable that younger demographics will be increasingly resistant to constantly pushing buttons to prove what they are viewing, just as they would be unwilling to make their beds or call their parents every half-hour or otherwise adhere to rigid requirements.”