Nielsen hopes to perfect button-pushing


It probably comes as no surprise to you that Nielsen has discovered that not everyone in a People Meter household is diligent about pushing buttons properly. But the ratings company says its research is helping it come up with ways to encourage better compliance.

To figure out how accurately data generated from People Meters matches actual viewing behavior, Nielsen conducted studies in which it called metered households to inquire about who was actually watching television and then compared that to the meter data. Depending on your point of view, the meter data was either 90% accurate or 10% inaccurate. Here’s a summary of findings from research conducted in June 2008:

— “Overall agreement between coincidental reporting of viewing and People Meter button-pushing records is approximately 90% (89.9% for all persons across all dayparts measured). This is the net of two forms of agreement (and disagreement) between the coincidental results and the People Meter entries.

— Under-reporting – Instances where the coincidental identifies viewing that is correctly reflected by button-pushing records (called the In-Audience Agreement Rate) is 82.3%. In other words, 17.7% of the time, the coincidental identifies viewing when a panelist’s People Meter button is not pushed.

— Over-reporting – Instances where there is no viewing and the coincidental report agrees with the button-pushing records (called the Not-In-Audience Agreement Rate) is 94.3%. In other words, 5.7% of the time, the People Meter indicates viewing when the coincidental reports none.

— The above two measures of under and over-reporting offset each other to some degree. The net effect is reflected in a measure called the Viewing Index. The overall Viewing Index for this study is .92, which implies that when both the under- and overstatements of viewing are combined, the coincidental shows 8% more viewing than that which is reported by the People Meter. Or, in other words, the viewing shown in the People Meter data is 92% of that extracted from the coincidental calls.”

Now that Nielsen has quantified non-compliance, it is working to improve button-pushing compliance. “We will develop general modifications to our coaching process to improve button-pushing compliance and apply them where necessary – for example, in situations with multiple viewers and specific demographic targets,” the ratings company said. “We will consider specialized treatments for specific households that have been identified in this study as being non-compliant. While this is something that we can implement quickly and in a targeted fashion, it will take great care to implement it in such a way that we don’t risk household cooperation by casting the coincidental call as a punitive measure,” it continued.

In addition, Nielsen said it will continue to analyze compliance and report on future waves of the study.