Friday, June 10, 2005

Lose the Election? Looks May Be to Blame

Split second judgments about a politician's competence can predict an election's outcome better than chance alone, a new study reveals. The results indicate that superficial inferences can contribute to voting choices, a process hoped to be rational and deliberative.
Alexander Todorov and his colleagues at Princeton University showed more than 800 people pictures of two candidates who competed against each other in races for either the U.S. Senate or House of Representatives. The researchers asked subjects to rate the politicians on characteristics such as age, trustworthiness, charisma and competence, based on a glance that lasted less than a second. Analysis of the data showed that the rankings of competence correlated with election outcomes: nearly 70 percent of the time the candidate thought to appear more competent was the race's winner. "Although the study doesn't tell us exactly what competence is---there are many kinds, including physical strength, social dominance and intellectual shrewdness--babyfaced people are perceived to be lacking in all these qualities," explains Leslie A. Zebrowitz of Brandeis University, who penned a commentary that accompanied the study in today's issue of the journal Science.
ADVERTISEMENT (article continues below)



Judging a nuanced character trait such as competence solely on facial features uses a decision-making method known as System 1 processes. As a rule, these choices are fast, unreflective and effortless. A second kind of evaluation that has been implicated in voting choices uses System 2 processes, which are slow, deliberate and require more effort. Conclude Zebrowitz and co-author Joann M. Montepare of Emerson College: "Understanding the nature and origins of appearance biases has real world value, not the least of which may be identifying electoral reforms that could increase the likelihood of electing the most qualified leaders rather than those who simply look the part." --Sarah Graham

No comments: