Nielsen Ratings fail to accurately represent Americans’ preferences

Editorials featured in the Forum section are solely the opinions of their individual authors.

Until a little less than a month ago, you were completely and utterly insignificant in the eyes of one of the most powerful market research firms in the world. You — the college student, the future of America — had absolutely no say in a cause that is, in all likelihood, near and dear to your heart: television.

The Nielsen Ratings have been an established part of the TV industry for over four decades. The ratings, which are one of the only metrics determining which shows succeed and which fail, are computed in two ways. The first method is “ratings diaries,” which require people who have opted into the Nielsen program to record all of the TV they watch for a week. The other method the system uses is the set-top-box method: Participants have boxes installed on their televisions that take note of who in the family is watching which shows. There are only about 25,000 Nielsen boxes in the country, and the viewing diaries are distributed to about two million people over the course of a year.

Hold on a second: Nielsen Media Research estimates there are about 110 million TV-equipped homes in the U.S. But Nielsen ratings only take into account 1.8 percent of Americans’ TV preferences. Not even 2 percent of TV viewers have the power to dictate what shows get renewed and which get axed mid-season. Doesn’t that seem unbalanced?

Anyone who has taken a statistics class knows that you can never sample the whole population; that’s why it is called a sample. Any good research will have a random sampling of a population, which stands in for the larger population. Ideally, the research will demonstrate external validity: The results can be generalized to the larger population.

The problem, in Nielsen’s case, is that the American population is anything but homogenous, especially with regard to TV viewing habits. When the Nielsen system was conceived in the 1960s, Americans watched television differently. There were only a few networks and television was more of a shared experience, with families gathering around the only TV in the house and probably watching the same shows as the family across the street. Random sampling of the population probably worked great back then; but the times, they are a-changin’.

With the advent of cable and the wider adoption of television sets, individual networks often have personalities and “target markets.” A random sampling can’t necessarily hit all the niches, and it’s not a binary choice. The people who watch Battlestar Galactica every Sunday night would not be watching Desperate Housewives otherwise.

Until January 29 of this year, Nielsen families with students at college did not record the viewing habits of their children away from home. College students, of course, have far more free time than most other demographic groups, and Nielsen estimates they watch an average of 24 hours of television every week. Only now have they realized that those 24 hours actually matter.

The best example of how the Nielsen system doesn’t work was the cancellation and renewal of Family Guy. Fox cancelled the show after three seasons of poor ratings, only to realize its mistake when DVDs of the show sold extraordinarily well. Oops.

Nielsen also recently decided to have viewer diaries include recorded content when applicable. If a Nielsen family didn’t have time to watch Lost on Wednesday night, and instead watched it on TiVo over the weekend, Lost would no longer have a little gap in their ratings.

Still to be considered is online streaming. Many networks allow their prime-time shows to be watched on the network’s website via streaming videos and downloads. Apple’s iTunes Store sells thousands of television shows every week, and even more television shows are downloaded over peer-to-peer networks like BitTorrent. These viewers are watching a show, but if they don’t have a Nielsen box, their vote doesn’t count.

Admittedly, Nielsen Media Research is working on solutions to these problems, but every day they don’t figure it out is another day that your voice as a college TV viewer goes unrepresented. And you are the one who feels the hammer fall when your favorite show gets cancelled. Here’s hoping the future holds some answers.