Positive photos of politicians published in two major American newspapers foreshadow favorable polling results, four researchers from the University of California at Los Angeles have concluded.
The two newspapers in question are the New York Times and the Washington Post. The researchers also found that international outlets Reuters and The Guardian tend to publish photos that reflect the public’s attitude and have no influence on future polling numbers.
One would be hard pressed to find many instances in which photos of an assertive and confident POTUS have been published in the pages of either the Times or the Post in recent months. Obama’s polling numbers, consistent with this research, remain low.
“Because we believe our own eyes, but know well that people are manipulative, we tend to be verbally skeptical and visually gullible,” write Jungseock Joo, Weixin Li, Francis F. Steen, and Song-Chun Zhu in the report.
The researchers applied a complex mathematical model that mapped the relationship between what they referred to as “syntactical attributes” and presumed intents, to reach their conclusion. A series of steps were undertaken to create the model, the first of which was identifying nine syntactical attributes, among them facial expressions, gestures and the scene’s context.
The four researchers then collected 1,124 images of eight U.S. politicians from various news sites and recorded the attributes.
Generating 4,000 random pairs of photos, each pair showing the same politician, 10 student volunteers were then asked to evaluate the photos, indicating which of the two ranked higher on a given quality.
From here the researchers were able to determine “perception of competence arises from combination of facial displays (smile), gestures (hand wave, hand shake), and scene context (large-crowd).”
With that information they arrived at their model, with the intent to apply it via computer analysis to determine a politician’s likely visual effect on voters.
The researchers singled out Obama for their first case study, applying their model to about 50,000 photos of him that were published in the aforementioned news outlets between January 2008 and September 2013; comparing the findings to aggregate poll results during the same period to arrive at their conclusion.
“The contribution of our paper is to quantify this emotional signal so that it can become machine-interpretable, and to scale up to deal with dozens of thousands of photographs or even videos,” Joo told USA Today in an e-mail.