• by The Riff , August 30, 2006

COMO ESTA USTED, NIELSEN -- Nielsen's decision to conduct a test of a new financial incentive plan on its live TV ratings samples in New York and Los Angeles marks another fissure in the line separating media research pragmatism and Puritanism. Testing new research methods on live samples is normally deemed verboten, because the tests can impact the behavior of panel members, which can skew the results of ratings that are used as the currency for advertising buys and programming decisions. In this case, Nielsen believes the risk is worth the reward, which would be to find a way to boost the number of Spanish-speaking households participating in its ratings panels.

The industry's ratings watchdog apparently agrees. George Ivie, executive director and CEO of the Media Rating Council, a guy not known to play fast and loose with the rules of the ratings business, says the MRC has extremely stringent guidelines when it comes to such tests, especially ones involving financial incentives to prospective sample members--and doubly so when the plan is to reduce the incentives for some. And that is exactly what Nielsen plans to do, playing the Hispanic American equivalent of Peter and Paul, er Pedro and Pablo.

Beginning in October, it will redistribute the cash it pays Hispanic households to be in its ratings panel, taking some money away from those families where English is the dominant language spoken at home and giving it to those households where Spanish is the dominant language. It's a simple economic solution to a vexing research problem: getting difficult households to participate in the ratings process. In fact, it was that very issue that was spun so effectively by News Corp. to amass public scrutiny and government pressure on Nielsen during the whole Don't Count Us Out gambit. Nielsen and News Corp. have settled their differences, DCUO is now dormant, and lawmakers and regulators appear to have shifted their eyes off Nielsen once again, but the ratings researcher remains committed to its mandate of being all-inclusive, even if it means bending the rules from time to time to get the necessary results.

This is not the first time, of course. Nielsen has periodically put artificial "weights" in place, adjusting it ratings to make up for poor sampling. Those weights have been vetted, and, for the most part, accepted by the industry, though they have yielded all sorts of aberrations of their own, including an odd one that emerged when Nielsen began releasing data on time-shifted viewing in DVR households. Occasionally, ratings in those homes are lower when playback is added back in than for "live"-only ratings. That's not possible, of course. At worst, those ratings would have to be the same, and that would assume nobody played back anything they recorded or delayed. The anomaly, Nielsen said, was due to its weighting scheme, raising questions about what other kinds of anomalies occur from time to time without anyone ever becoming aware of them.

It seems that the MRC's Ivie sanctioned Nielsen's live sample test with a mixed heart--and only because the potential upside of a better, more holistic ratings sample appears to outweigh the threat that a few extra dollars in the pockets of Spanish-speaking Americans would significantly affect the way they watch TV.

But as one ratings researcher whom we have a great deal of respect for notes, "Experimenting on a live panel is potentially introducing a non-systematic bias into the survey and its results. Expressed differently, it alters the methodology in a manner no longer consistent with that which was accredited." Another problem, the researcher said, is that Nielsen initially will only be making these changes in two markets--New York and Los Angeles--creating even more inconsistencies within its local and national people meter samples. If the method is deemed successful, Nielsen says it will introduce it in other markets.

Nielsen Chief Research Officer Paul Donato has spent a considerable amount of time researching the issue of cash incentives over the past year, particularly what happens when the ratings company pays Nielsen households significant sums that could impact their lifestyles. Among other things, the extra cash could be used to engage in more social activities outside the home--dining out, going to the movies, etc.--which could alter their TV usage patterns. Another possibility is that they could use the money to buy new electronic equipment such as DVRs or DVDs that could enhance or compete with conventional TV viewing.

So far, the results of Donato's research have shown that higher cash incentives do not alter viewing behavior, but for some of the sums we've heard of, we'd consider speaking more Spanish at home. Under certain scenarios, Nielsen's cash payments could be high enough to require households to report the sums to the Internal Revenue Service as part-time income. Nielsen says it hasn't quite reached those thresholds yet, but if its recruiters come knocking on our door anytime soon, we know exactly what we'll tell them: "Bienvenido a mi en casa."

Until then, hasta la vista, baby.