Salvatore A. Soraci
When an eyewitness sits on the stand in a courtroom recalling details of an incident, how much of what he or she remembers actually happened?
False memories are common in the courtroom and in everyday life. Psychologists have long considered false memories a side effect of efforts to boost memory. New research from Tufts University has answered the question of how to increase memory without also increasing corresponding false memories.
"The better we understand false memory, the more we will be able to explain the factors that lead to the problem in the laboratory and real-world situations," said Salvatore A. Soraci, associate professor of psychology, whose research was published in the July issue of The Journal of Experimental Psychology: Learning, Memory and Cognition.Active learning
Previous research has focused on how to improve memory using "generative learning"—the concept that we remember things more accurately when we are actively involved in forming an idea. For example, if an individual is given a clue and asked to provide a one-word answer, he or she will remember that word better than if simply given the word and told to memorize it.
"Generative learning holds the promise of immunizing people against the pitfalls of false memory," Soraci said.
Funded by a $654,000 grant from the National Institutes of Health (NIH), Soraci and Michael Carlin, assistant professor of psychology at Tufts, set out to determine what impact generative learning has on the formation of false memories.
Study participants were given a list of words to memorize. Some of the words were complete, while others were missing one letter. The complete words fit within subject categories (for instance, "queen," "moat," "knight," etc.), while the incomplete words were in different subject categories (for instance, furniture such as "t*ble" and "cha*r").
After a three-minute period, during which participants were given a math quiz as a distraction, they were then presented with a list of words. The list included some words that had not been included in the original list but were related to the subject categories. The participants were asked if the words were among those that had been shown earlier.
"The incomplete words led to generative learning, because the participants had to determine the words on their own," Soraci said. "People were far more likely to falsely remember words from the list of complete words—such as mistakenly believing that 'king' had been on the list—than they were to falsely identify a word from the generative learning list."
In a similar experiment, study participants were not provided with the second list and instead asked to write down all the words they could remember seeing. The experiment showed the same advantage for generative learning.Negative reinforcement
In another experiment, Soraci determined what kind of cues would help people remember words without increasing false memories. Participants were given a list of words that were missing one letter and could be either of two words, depending on what letter was used to fill in the blank. For example, one of the words on the list was "s*eaker," which could be "speaker" or "sneaker." Some of the participants were given a positive clue, such as "a tennis shoe," and asked to fill in the blank. Others were given a negative clue, such as "not part of a stereo." Soraci found that people were more likely to remember words when given a negative clue than a positive one, and also were less likely to falsely remember a word.
"This method of learning using negative cues is similar to how we find our way when we're driving our cars and looking for a new location," Soraci said. "If we make a wrong turn, we're much more likely to remember the correct route next time by remembering that we shouldn't go the wrong way again."
Editor’s note: Salvatore Soraci, a faculty member at Tufts University since 1993, died unexpectedly on August 26. A memorial service for Prof. Soraci will take place on Tuesday, September 23, at 5 p.m. in Goddard Chapel on the Medford/Somerville campus.