Stripping down TV's poll dance
Namely, don't believe 94% of the polls you see (with a plus-or-minus 3% margin of error) regarding the public's attitude toward TV.
Jack Rosenthal's piece in the Times' "public editor" space explored various misrepresentations that are disseminated by journos, many of whom would likely be doing something more lucrative, frankly, if they were any good at math.
Paraphrasing, three areas of caution that Rosenthal cited are particularly germane to TV:
- How intensely do people feel about an issue?
This is the most vexing aspect of polls pertaining to media in general and TV in particular -- where misgivings about sex and violence run a mile wide and, I'd argue, an inch deep.
Asked if they are "concerned" about televised sex, violence and profanity, vast majorities will usually say, "Sure." Yet ask the same to walk across the street to sign a petition that would reduce those levels, and a similar majority would say, "All the way across the street?"
The number truly agitated about "indecency" is relatively small, not only because few adults have minors at home but because a lot of us like seeing naked bodies, hearing coarse language and watching good guys beat the living daylights out of bad guys. As my sister said, "It's amazing how cathartic watching violence can be when it's done right."
'Net, however, has become a valuable rallying tool, magnifying the sound of limited voices like a megaphone and enabling half-hearted allies to "take action," expressing "outrage" by simply hitting "submit." When required to write a letter or even make a phone call, participation plummets.
- Respondents want to look like "good citizens." With most surveys, everyone knows what the answer should be. That's why more people say they intend to vote in elections or know the issues than actually do. This is especially true when discussing TV viewing habits. Nobody wants to admit that they can name all the panelists on "The View" but not a single Supreme Court justice.
Very few people will say, "I love watching sex and violence, especially if it's potentially harmful to young kids." Somehow, though, all those slasher movies, episodes of "Taxicab Confessions" and latenight Cinemax skinfests aren't watching themselves.
- How questions are phrased matters.
Because scientific polls are expensive, few major news orgs waste time and money on media matters, which is why many "studies" addressing TV are skewed by groups with axes to grind.
Find any parent who answers "no" to a question like "Are you concerned about the sex and violence your children are exposed to in the media?" and I'll show you a candidate for a visit from Child Protective Services. But the "yes" tells us nothing, and most surveys lack any sense of nuance about those parents who wouldn't want their children watching "The Sopranos" or "Nip/Tuck" but who savor every last drop after dutifully tucking the kids into bed.
Then again, in these statistically minded times, all kinds of information swirls in the ether of dubious worth. TiVo, for example, is making "behavioral data" available to clients, but that only demonstrates characteristics of its subscribers -- an early-adopting, opt-in sample. Ditto for voluntary surveys such as "The O'Reilly Factor's" online polls, where participants overwhelmingly agree with -- surprise -- Bill O'Reilly!
The plain truth about popular entertainment is that people have always said one thing and watched another. Hell, 35 million of them somehow were paying enough attention to tune in to at least part of the Emmy Awards in August, according to Nielsen data.
Not that I believe everything Nielsen says either, but since they use those peoplemeter thingamajigs, they have to be more reliable than the polls.
