Undecided voters not so undecided

These days, you can’t even trust undecided people. Individuals who honestly believe that they can’t choose between two available options may in fact already know what to do, thanks to attitudes that lurk outside their awareness, a new study indicates.

Rapid mental associations made by individuals who were undecided on a controversial political issue frequently predicted opinions these people later formed on that issue, psychologist Bertram Gawronski of Canada’s University of Western Ontario in London, Ontario, and his colleagues report in the Aug. 22 Science.

“One could say that people sometimes have already made up their minds, even though they do not know it yet,” Gawronski says.

“Political pollsters might learn that there are some questions better left unasked,” remarks psychologist Timothy Wilson of the University of Virginia in Charlottesville. The new findings suggest that pollsters should be skeptical of voters who label themselves as undecided because those voters’ unconscious minds may have other ideas. The conscious answers these people give could be misleading.

Gawronski recommends that pollsters consider adding measures of automatic associations to their survey repertoires.

His team interviewed 129 residents of Vicenza, Italy, during the last two months of 2007 about the impending enlargement of a U.S. military base in their community, a polarizing issue at the time. Earlier this year, the Italian government approved the base expansion without holding a public vote.

In initial interviews, 32 residents favored the expansion and 64 opposed it. Another 33 participants said they were undecided. Everyone completed a 10-item questionnaire assessing their views on probable consequences of the proposed expansion.

Participants also took a computer-based test measuring automatic mental associations. First, they practiced pressing a left-hand key as quickly as possible when positive words, such as joy and lucky, appeared on the screen and a right-hand key as quickly as possible when negative words, such as awful and pain, appeared on the screen. Then they were instructed to press a left-hand key when they saw either images of the U.S. military base or positive words and to press a right-hand key only when they saw negative words.

In a third trial, participants pressed a left-hand key only for positive words and pressed a right-hand key for either images of the U.S. military base or negative words.

One week later, none of the decided citizens had changed their minds. Among the others, 14 were still undecided while 19 had switched to a pro or con position. Everyone repeated the automatic mental associations test.

The undecided people who later made up their minds showed distinct responses on the first association test. Those who ended up endorsing the expansion were faster and more accurate at pressing the key for both positive words and military base images. Those who eventually opposed the expansion were faster and more accurate at pressing the key for both negative words and military base images. The same pattern held on the second association test.

Participants who stayed undecided showed no strong positive or negative associations to the military base images at any time.

Participants who began with strong pro or con opinions showed strong positive or negative associations to military base images when first tested, and even stronger associations one week later. Their automatic associations consistently coincided with their conscious views.