Cue the infomercial:
Wanna to know what real people are thinking, but can’t bother to ask actual questions? Have no fear, the Perception Analyzer is here!
This nifty gadget fits in the palm of your hand! Read viewers’ minds without lifting a finger. It slices and dices complex, nuanced ideas into bite-sized factoids. It’s never been easier to toss in a handful and infuse meaningless data into your broadcast!
Order now! Operators are standing by. (Shipping and handling not included.)
The Perception Analyzer is the brand name of the dial doodad (the generic version is only available in Canada) that CNN is using to gauge viewers’ reactions to the debates. It’s been around for the last three debates, and is bound to make an appearance on Wednesday. Oh, how I wish it wouldn’t.
The system works like this: thirty voters, roughly equal numbers of Ds, Rs, and Is, hold in their hands a doodad that looks like an egg timer. Turn the dial to the left to register a negative reaction and turn it to the right to express approval. The results are tabulated to create a real-time line graph on the bottom third of the screen. Responses above the neutral line are positive, below the line, negative. Easy enough, right?
The dial promises insight, but only offers confusion, since there’s no way to actually know which words or phrases elicited the viewers’ reaction.
Here’s one reason: In a line graph, you need to identify both axes, in terms of meaning and intervals. In the case of the dial reports, the y-axis represents approval, while the x-axis represents time. The range of reaction is defined from 0-100; below 50 is bad, above 50 is good. Of course, the graph is so small, crammed into the bottom of the screen, that it’s hard to actually qualify how much approval is being expressed. Did Obama get 45-out-of-50 approval on his health-proposal? You can surmise whether the feedback is generally positive or negative, but that’s about it.
But how is the x-axis defined? In terms of intervals, it’s hard to tell if viewers are reacting to the theme of a candidate’s answer, his overall proposal, or his specific phraseology. Unless you have an excellent memory, it’s hard to simultaneously watch the squiggly lines and keep in your brain what is being said. This requires a nimble act of synthesizing audio- and visual-processing, something that could be corrected with a running transcript of the debates presented along the x-axis.
Another complaint against the dial-meter concept is that it operates with a tiny sample. During the broadcast, Wolf Blitzer asked Soledad O’Brien if the dial group could be “extrapolated to the rest of the population.” No, O’Brien responds, “But to some degree, they are a microcosm of the United States.” If that doesn’t suggest extrapolatability, I don’t know what does.
Also, according to studies cited by the Wall Street Journal , “Recent psychological experiments suggest they can influence viewers’ judgments. That might give tiny focus groups outsize influence, especially over undecideds.”
Here’s more from the WSJ:
Often, the groups turn their dials up when they hear specific plans, like Sen. John McCain’s desire for tax cuts or Sen. Obama’s charges against CEOs’ golden parachutes. By contrast, they turn them down when candidates repeat obvious catchphrases or go on the attack, said Rita Kirk, a professor of communications and public affairs at Southern Methodist University who is running CNN’s focus groups.
On-screen charts are just one of a host of outside factors that might influence judgments. Some have argued polls can have a “bandwagon effect.” Debate spin is often based on the idea that framing the discussion can shape views. And reactions among fellow debate watchers can have as profound an effect as a laugh track.
Two studies published in the last two years suggest continuous-reaction graphs can affect opinions — at least in an experimental setting. In one, led by a researcher at Emory University, 253 college students evaluated “American Idol”-like performances with fake audience feedback superimposed on screen. Those who saw negative reactions themselves viewed the performances more negatively.
As it is, voters have a snowball’s chance in hell of forming their opinions based on the words and deeds of the candidates alone, given the barrage of opinion-posing-as-news to which they’re subjected. The dial-group polls are just one more example of these opinion injections. Perhaps, instead, networks could compile the dial-poll responses and present a summary and analysis of the results in the follow-up to the debates.