Interactivity produces statistical morass for TV networks
The mixed blessings of our technologically connected world include advancing a trend Monty Python might have lampooned as the Bureau of Silly Interactivity and Meaningless Polls.
The old guard’s infatuation with reaching out through social media, and allowing viewers to reach back in and contribute to content, surely isn’t going to improve math scores. Witness the proliferation of insta-polls and real-time reactions designed to look cool and hip, but which have no statistical significance or evidence of representing anyone but a few lonely souls on Twitter.
Granted, the temptation is understandable. It’s appealing to quantify what “the public” wants — even if it’s only a small subset — and now that we have these new toys, well, why not play with them?
Yet people are hardly going to become more informed when the quality of the data being fed to them amounts to garbage in, garbage out.
ABC trotted out the latest exercise in interactive mania with “The Glass House,” a series in which 14 sequestered contestants are supposed to be directed by viewers, who can vote online about what sort of things they want to see happen in the “Big Brother”-like setting. ABC’s website noted fans can influence contestants’ actions by voting to “decide everything from what they wear and eat to the games they play, even where they sleep.”
As it turns out, “the show they didn’t want you to see” — as ABC coyly referred to a premiere CBS took legal steps to block on copyright grounds — was a show most of the public didn’t want to see, either. And it’s a good bet an extremely tiny fraction of the 4 million people who watched the premiere took advantage of the opportunity to participate.
“Glass House” nevertheless represents an exaggerated example of the desire to weave social media into TV and communications arsenals in ways intended to be showy and provocative but which don’t tell anything to anyone with the slightest knowledge about polling methodology.
Simply put, as the public-opinion research organization Public Agenda explains it, “pseudo-polls” with self-selected respondents have no scientific validity and thus “should be ignored and not reported.”
News operations like Fox News nevertheless love pushing their audience toward websites with such polls, just as networks like TNT or CNN pass along real-time Tweets from viewers — a form of gimmickry that appears especially ill-suited to older-skewing channels, whose viewers are less likely to be multitasking in that fashion.
The hunger for information inspires the media to throw around all kinds of bogus measures, from dial-in contests to consumer grades for movies from CinemaScore which, predicated as they are on those who rush to see films during their opening weekend, tend to be inordinately charitable.
Interactivity has long been considered a sort of Holy Grail in television, with experiments — several involving children’s shows — dating back to the medium’s infancy. The dual-screen environment, with people watching while simultaneously on a computer or using a hand-held device, has hastened the pressure, if only in anticipation of what might come next.
The allure of layering on extracurricular elements is obvious, seeking to foster higher levels of engagement and commitment among viewers. The practice also reflects the media’s habit of pandering to youth, who are presumably more apt to tweet or phone in during marquee events than are their parents, even if those under 25 are less likely to watch in the first place. (Be wary of grownups who squander undue time logging calls to select the next “American Idol.”)
The drawback is bad information that can mislead more than it illuminates, especially when networks play fast and loose with results, as they often do, making sweeping “America has spoken” assertions.
These misgivings are hardly new. As Stanford U. researchers concluded in 2009, compared with the high accuracy of random-sample polls, Web-based studies using self-selected or opt-in panels “were always less accurate, on average, than probability sample surveys.”
The unreliability, and in many instances vitriol, associated with online discourse also explains why so many reputable websites have curtailed comment sections, realizing the nebulous effect of such input on serious discussions.
Finally, the uncertainty goes beyond what’s said, to verifying who’s saying it. As the New York Times reported, the social-networking app Skout recently discovered adults masqueraded as teenagers in an online youth forum, and even now it remains “extremely difficult to tell whether someone is an 11-year-old girl or a 45-year-old man.”
The appeal of employing the Web to assemble man-on-the-street reaction or take the public pulse is easy to see; it’s quick and, best of all, cheap. Small wonder print reporters increasingly use corner-cutting tools like Twitter to locate premise-supporting quotes or anecdotes.
It’s just that most of this digital dialogue is, in any statistical way, meaningless — something that should be transparent to anybody, even if they don’t live in a glass house.