February 4, 2005

SCIENCE JOURNAL By SHARON BEGLEY

People Believe a 'Fact'
That Fits Their Views
Even if It's Clearly False
February 4, 2005; Page B1

Funny thing, memory. With the second anniversary next month of the U.S.-led invasion of Iraq, it's only natural that supporters as well as opponents of the war will be reliving the many searing moments of those first weeks of battle.

The rescue of Pvt. Jessica Lynch. U.S. troops firing at a van approaching a Baghdad checkpoint and killing seven women and children. A suicide bomber nearing a Najaf checkpoint and blowing up U.S. soldiers. The execution of coalition POWs by Iraqis. The civilian uprising in Basra against Saddam's Baathist party.

If you remember it well, then we have grist for another verse for Lerner and Loewe ("We met at nine," "We met at eight," "I was on time," "No, you were late." "Ah yes, I remember it well!"). The first three events occurred. The second two were products of the fog of war: After being reported by the media, both were quickly retracted by coalition authorities as erroneous.

Yet retracting a report isn't the same as erasing it from people's memories. According to an international study to be published next month, Americans tend to believe that the last two events occurred -- even when they recall the retraction or correction. In contrast, Germans and Australians who recall the retraction discount the misinformation. It isn't that Germans and Australians are smarter. Instead, it's further evidence that what we remember depends on what we believe.

"People build mental models," explains Stephan Lewandowsky, a psychology professor at the University of Western Australia, Crawley, who led the study that will be published in Psychological Science. "By the time they receive a retraction, the original misinformation has already become an integral part of that mental model, or world view, and disregarding it would leave the world view a shambles." Therefore, he and his colleagues conclude in their paper, "People continue to rely on misinformation even if they demonstrably remember and understand a subsequent retraction."

For the study, the scientists showed more than 860 people in Australia, Germany and the U.S. a list of events -- some true (the first three examples above), some reported but retracted (the second two), some completely invented ("Iraqi troops poisoned a water supply before withdrawing from Baghdad"). Each person indicated whether or not he or she had heard of the event and rated its likelihood of being true. People were pretty good at weeding out the invented reports. Then, for each report they said they had heard, they noted whether it had subsequently been retracted.

If the report had been retracted, surely people would no longer regard it as true, would they? Here is where memory parts ways with reason. The Germans and Australians responded as you'd expect. The better they recalled that a claim had been taken back, the less true they judged that claim. They did not believe in events they knew had been erroneously reported.

But for the Americans in the study, the simple act of remembering that they had once heard something was enough to make them regard it as true, retraction be damned. Even many of those who remembered a retraction still rated the original claim as true.

That comes as no surprise to memory researchers. Time and again, lab studies show that people have an astonishing propensity to recall things that never happened. If you read a list of words such as pillow, bed and pajamas, and are later asked whether another word was there, you may well "remember" related words that were never presented. "Sleep" was on the list, wasn't it?

In this case, people's mental model is "words about sleep." In the case of memories about Iraq, people's mental model is why the U.S. invaded. The Germans and Australians in this study were skeptical of the official justification, namely, to find weapons of mass destruction. The Americans were more credulous on that point. How suspicious or credulous people were strongly affected whether they judged a retracted claim to be true or not.

"People who were not suspicious of the motives behind the war continued to rely on misinformation," Prof. Lewandowsky said, "believing in things they know to have been retracted." They held fast to what they had originally heard "because it fits with their mental model," which people seek to retain "whatever it takes."

In contrast, those who were suspicious of the WMD justification believed the retractions. The reason is probably that they weren't sold on the original, erroneous reports -- all of which cast the U.S. in a good light and Iraqi forces in a bad one. These people "are more willing to discard elements of a mental model that turn out to be wrong," says Prof. Lewandowsky.

The news media would do well to keep in mind that once we report something, some people will always believe it even if we try to stuff the genie back in the bottle. For instance, six months after the invasion, one-third of Americans believed WMDs had been found, even though every such tentative claim was discomfirmed. The findings also offer Machiavellian possibilities for politicians. They can make a false claim that helps their cause, contritely retract it -- and rest assured that some people will nevertheless keep thinking of it as true.