Campaign ad may have swayed voters subliminally. (Dirty RATS).

Psychological research sparked by a controversial campaign advertisement aired during the 2000 presidential election suggests that a 30-second spot--which briefly flashed "RATS"--may have negatively affected viewers' opinions of Democratic candidate Al Gore.

In one segment of the ad, which was funded by the Republican National Committee, short fragments of the phrase "BUREAUCRATS DECIDE" dance about the screen while a narrator criticizes Gore's prescription-drug plan for seniors. A frame-by-frame analysis of the campaign spot reveals that in one particular image, lasting only one-thirtieth of a second, RATS nearly fills the screen.

Some Democrats cried foul, accusing Republicans of planting subliminal messages--those shown too quickly or faintly to be consciously noticed--to turn voters against Gore. A bevy of Republicans, including then-candidate George W. Bush, dismissed that idea as absurd. Intrigued by the controversy, Joel Weinberger, a psychologist at Adelphi University in Garden City, N.Y., constructed an experiment that mimics the notorious commercial.

For the project, Weinberger and his colleague Drew Westen of Emory University in Atlanta developed a questionnaire in which people visiting an Internet site were asked to rate a purported candidate. After participants viewed the candidate's photo, they rated the contender in relation to 10 statements, such as, "This candidate looks competent" or "I dislike this candidate." Before the photo appeared on the screen, however, the researchers flashed one of four short messages--RATS, STAR, ARAB, or XXXX--for a mere six-thousandths of a second.

Weinberger described the pair's research this week in Denver at the annual meeting of the American Association for the Advancement of Science.

He and Westen used STAR as an option for the subliminal image because it's RATS spelled backward. They used the word ARAB to investigate whether those who answered the questionnaire held negative stereotypes related to the word. Subjects who viewed XXXX, a presumably neutral nonword, formed the control group for the research. About 250 people took part in the study.

For survey statements that were framed in an affirmative way, such as, "I like this candidate," the subliminal message that participants viewed didn't seem to affect their opinion. However, for statements phrased in a negative manner, participants exposed to RATS, on average, judged the candidate much more harshly than did people who viewed the other three subliminal messages.

Exposure to RATS had the same effect among men and women in the study. Also, participants who identified themselves as Republicans responded to RATS just as negatively as Democrats did. The good news from this research, says Weinberger, is that the experiment didn't detect an unconscious bias against Arabs among study participants.

The results suggest that negative impressions of candidates may be more easily affected by subliminal messages than positive ones, says Weinberger. However, it's also possible that questionnaire responses for affirmative statements didn't vary significantly because study subjects didn't strongly link STAR with a positive characteristic. More research would be needed to bolster the contention that negative campaigning really works, he notes.

Philip S. Holzman, a Harvard University psychologist, points out that the effects of subliminal messages, once ignored by many scientists, must be studied further. "Otherwise, we're at the mercy of the politicians," he quips.