THE PARALLAX VOTE:

Undecided? It's more partisan than you think: Researchers find that voters who haven't made up their minds actually have unconsciously made a decision based on deep-seated attitudes. (Denise Gellene, 8/22/08, Los Angeles Times)

The research, to be published Friday in the journal Science, used a computerized test in which participants were asked to react as quickly as possible to images arbitrarily deemed "good" or "bad." The test measured how long it took to respond.

Scientists selected 33 residents of Vicenza, Italy, who stated they were undecided about a controversial proposal to expand a nearby U.S. military base.

They were instructed to press the letter "D" when they saw a picture of a military base or one of five positive words, such as joy, pleasure or happiness, and the letter "K" when they saw one of the negative words, which included pain, ugly or danger.

The researchers then reversed the test so that the image of the military based was linked to the negative words.

The theory behind the test is that people will hesitate when required to perform actions incompatible with their unconscious attitudes. So subjects who unconsciously favored the base expansion took more time to react when it was associated with negative words, and subjects against the expansion delayed when it was associated with positive words.

The lag in reaction time averaged between 100 and 200 milliseconds, said Gawronski, who collaborated on the project with scientists from University of Padova in Italy.

One week after administering the test, nine previously undecided subjects said they now favored the base, 10 said they had decided against it, and 14 remained undecided. Participants' responses on the week-earlier computerized test and an accompanying opinion survey were about 70% accurate in predicting their actual decisions, researchers said.

The test hasn't yet been adopted by political consultants, although one, Virginia-based TargetPoint Consulting Inc., experimented with it during the recent Republican presidential primary race.

A research team from the University of Virginia, the University of Washington and Harvard University currently is offering the https://implicit.harvard.edu/implicit/demo/selectatest.html of the test, tracking reactions to the presidential candidates.

I got:

Your data suggest a moderate automatic preference for John McCain compared to Barack Obama.

Depending on the magnitude of your result, your automatic associations may be described as 'slight', 'moderate', 'strong', or 'little to no preference'. It is interesting to ask how much IAT-measured candidate preferences agree with self-reported preferences. The agreement tends to be relatively strong for political attitudes. One topic of particular interest for the 2008 Presidential campaign is whether self-reported candidate preferences agree or disagree with implicit associations. Were yours consistent? If not, why might that be, and what consequence does it have? These are some of the questions we are addressing with on-going research.

Posted by Orrin Judd at August 21, 2008 6:52 PM