The ends can sometimes indict the means. But indictment is not, as we all should know, conviction. Neilson is changing their methodology in major markets from using viewer logs and…
The ends can sometimes indict the means. But indictment is not, as we all should know, conviction.
Neilson is changing their methodology in major markets from using viewer logs and older monitoring devices to using just PeopleMeters. These measure what you’re really watching, vs. what you record in the manual viewer log.
But recent results in some test markets have some folks alarmed over this. Why? Because the results show that, as far as I can tell, when you actually monitor what people really watch, vs. what they say they watch, “minority-oriented programming” doesn’t seem to be watched as much..
THE N.A.A.C.P. and leading members of Congress from both parties, including Senator Hillary Rodham Clinton, have added their voices to the growing number complaining that the proposed changes in how Nielsen gathers local television ratings will drastically undercount the number of black and Hispanic viewers.
Mrs. Clinton, in her letter, cited concerns that the new system was “undercounting minority viewers.” Why this is so is unclear. But the senator cited results of a test of the local people meters during the winter, during which “virtually all top-rated shows among African-American adults witnessed significant declines in viewership, in some cases by more than 60 percent.” Among the shows affected are “The Parkers” and “One on One,” both on UPN.
“Similarly, large declines were seen in the ratings for top Spanish-language networks,” Mrs. Clinton’s letter continued. “Without a thorough investigation into these statistical aberrations, I think it is fair to say that Nielsen would be remiss in pushing forward with its rollout plan.”
In a separate letter to Ms. Whiting, Kweisi Mfume, president of the National Association for the Advancement of Colored People, expressed similar concerns, saying Nielsen’s plans could imperil the “future of programming aimed specifically at African-American and Latino audiences.”
Now, obviously, a differential result going from testing method A to testing method B raises some concerns. The problem is, nobody (who’s complaining) seems to be interested in whether the old method was actually more or less accurate than the new method. They’re just concerned about the results being different, and different in a way they don’t care for.
Or, perhaps, it’s just obvious to them that, well, naturally, Hispanic and African-American audiences are most interested in watching programming that is “oriented” in their direction (regardless, say, of its quality), and so if there’s a drop in the measured audience, then Neilson must be disenfranchising minorities. Right?
But Nielsen Media Research, whose ratings have been used for decades to help set TV advertising rates, takes issue with the critics, asserting that the number of households sampled with African-American and Hispanic viewers will actually increase under the proposed changes. Senator Clinton, Democrat of New York, in a letter yesterday to Susan D. Whiting, president and chief executive of Nielsen, urged a delay in plans to adopt, effective April 8, the new methodology in New York. The change involves adopting locally the so-called people meters Nielsen has used since 1986 to gather national ratings data.
The people meters would replace the paper diaries Nielsen has provided to viewers in local markets since 1950, as well as set-top boxes that are not as technically sophisticated. Nielsen intends to switch New York, Chicago and Los Angeles to local people meters as part of plans to have all the 10 largest local markets using them by next year. One big market, Boston, shifted to people meters in 2002.
[…] The African-American households will compose 21.2 percent of the sample, compared with an estimate that such households make up 17.3 percent of New York households. And Hispanic households will make up 16.8 percent of the sample, compared with an estimate that such households currently account for 16.1 percent of the total.
Also, Mr. Loftus said, during the test in January and February, total viewership among black men ages 18 to 49 in New York increased 4 percent from the viewership total obtained by using diaries and set-top boxes.
Among the networks that gained viewers in that coveted demographic Mr. Loftus listed CNN, Cartoon Network, VH1, and WNJU, a Spanish-language station.
So, WNJU aside, one reason for the changes in results is that cable stations are being better monitored. Which means that UPN and Fox, which schedule a lot of “minority-oriented programming” lose share.
That UPN and Fox’s “minority-oriented programming” consists of craptacular sitcoms that would be craptacular even if the faces were all-white or all-green or all-blue seems to make little difference. Minority viewers will obviously prefer programming that has minority faces on it, end of story, so …
… well it means something’s wrong, dammit! Something’s rotten in the state of Neilson!
Again, it’s certainly possible that the new PeopleMeter system is frelled up in some fashion. If critics wanted to harp on the discrepency in results, and drive for some answer as to why that was happening, I’d have no objections. If there’s analysis of that sort going on, it seems absent from the article, at least. Which is a shame, because that seems like basic … well, science.
Instead, the answer seems already predetermined by the critics. Neilson is undercounting minorities, because minority-oriented program viewing is down. QED.
Which strikes me less as science than as politics. And, frankly, racist politics at that.
(via BoingBoing)