How big a deal is the loss of 1.1 million viewers from 2002? Or the relatively small decline of viewers—200,000—between 2001 and 2003?
"It's not a significant shift" in viewers, says Anne Elliot, vice president of marketing and communications at Nielsen. Nielsen doesn't count viewers; it estimates them based on the 5,100 People Meters attached to TV sets that measure the TV consumption of the 14,000 to 15,000 people in those households. As with any scientific sample, there is a margin of error built into Nielsen's estimates. Different Nielsen estimates have different margins of error, Elliot says, and while she couldn't cite the margin of error in this particular example, she cautioned against extrapolating too much from the raw numbers.
Rutenberg undermines his own news burnout thesis by noting that the summer of 2002 might have been an outlier news year because of several very telegenic stories: the Elizabeth Smart disappearance; the rescue of the Pennsylvania coal miners; and the Los Angles International Airport murders.
In other words, this summer's trend is that there is no trend.
Spotted a bogus trend in the news? Send e-mail to firstname.lastname@example.org.