Any student of communications worth his or her salt will have studied the infamous Nixon-Kennedy Presidential debates of 1960. Why? Because they were the first ever televised presidential debates, and they marked an inflection point in American politics, where hearts and minds were not won merely by talented rhetoricians and charming radio personalities, but increasingly by physical appearances and a demonstrated ease in front of a camera.
As the story goes, Nixon was ugly and evil looking normally, but on the date of the first of the four debates he would have with Kennedy, his physical appearance was worse than usual: “Nixon had seriously injured his knee and spent two weeks in the hospital. By the time of the first debate he was still twenty pounds underweight, his pallor still poor. He arrived at the debate in an ill-fitting shirt, and refused make-up to improve his color and lighten his perpetual ‘5:00 o’clock shadow.’” I think we can all imagine.
However, Kennedy’s appearance was another story, “Kennedy, by contrast, had spent early September campaigning in California. He was tan and confident and well-rested. ‘I had never seen him looking so fit,’ Nixon later wrote.”
Whether Kennedy’s handlers were much more prophetic about the impact of TV, or whether Kennedy just lucked out, we may never know. What we do know is that Kennedy’s appearance on TV during that debate changed the path of American politics forever. A majority of Americans who listened to the debate solely via radio pronounced Nixon the winner. A majority of the over 70 million who watched the televised debate pronounced Kennedy the easy winner.
Are you beginning to see why this appeals to comms geeks? The suggestion that a newly introduced medium could so profoundly impact the perspectives of so many people in the context of a very high stakes popularity contest was tantalizing. It remains tantalizing today.
Fast forward 51 years to Obama conducting a Townhall meeting streaming on Facebook, and to GOP Presidential candidates using Twitter and Facebook metrics to potentially supplant traditionally collected polling information.
What would happen if you could use Twitter, Facebook or good old Google Analytics to accurately predict the outcome of the 2010 Presidential Election? Some growing social media analytics companies such as Likester are doing just that by measuring the uptick in popularity of pages and social networking presences. In fact, Likester accurately predicted this year’s American Idol winner way back in April.
But how scientific is this data, and what exactly is being measured? As Mashable reports, Likester mostly measures popularity and predicts winners based on the aggregation of “likes” on Facebook in concert with high-profile events. For the GOP debate, “The stand-out frontrunner was Mitt Romney, who ended the night with the greatest number of new Facebook Likes and the greatest overall Likes on his Page.” As we can see, Likester basically began the ticker right when the debate began and distinguished between unique “likes,” or “likes” that occurred after the debate had started, from overall likes. In the end Romney had 19,658 unique or new “likes” during the debate, resulting in 955,748 total “likes,” representing a 2.06% increase in overall “likes” during and directly following the televised debate.
Likester reported, “Michelle Bachmann ranked second in the number of new Likes on her Facebook Page.” In numbers that came out to 9,232 unique or new “likes,” 326,225 total, representing a 2.75% increase.
Naturally, AdWeek threw their two cents into the discussion, arguing:
“Polling has always been an important element to any electoral bid, but now a new type of impromptu assessment is coming to the fore. Third parties, such as analytics startup Likester, are carving out a space for themselves by processing data that is instantaneously available.”
I’ll give you instantaneously available, but, again, how scientific is this? After all, no one seems to be taking into account what I would call the “hipster correlate.” The hipster correlate is the number of Facebook users who would have “liked” a Romney or Bachmann or Ron Paul page in a stab at some hipster-ish irony, thus proving to those who check their Facebook page or read their status updates their outstanding skills of irony in becoming a fan of a Tea Partier web page, etc. If we’re really doing statistical regressions here, what’s the margin of error here, Likester?
Additionally, how closely can we attach the fidelity of someone who “likes” a Facebook page to a living, breathing politician? On my Facebook page I think I have “liked” mayonnaise, but if there were to be a vote between mayo and ketchup to determine which condiment would become the new official condiment of the U.S., would I necessarily vote mayo? That’s actually kind of a crap analogy, but you get what I mean.
Before we are forced to read more blogs and news articles (like this one!) pronouncing exit polls dead and Facebook and Twitter as the new polling media, I’d like to see a very solid research study conducted as to how closely becoming a fan of a political Facebook page correlates to Americans’ actual voting behavior. In other, more web-based marketing terms, what’s the voting conversion rate for political Facebook pages?
Has anyone seen anything like that?
In other words, please, social scientists and pollsters, show us whether yet another new medium is disrupting the way that Americans individually see and interact with their political candidates, and how that medium has begun to shape the way those political candidates are regarded by an American audience as a whole.