Failures of Memory--Big and Small

By now, we all know that Brian Williams “misremembered” that a helicopter he was in was hit by enemy fire in Iraq twelve years ago. The video below shows how his account of the incident changed over the years. Since the memory was self-serving, many people have characterized Williams' account as a lie rather than an accidental memory lapse. To people who have never been in a war zone, the idea that someone might forget whether his helicopter was hit by gunfire is a pretty tough sell. NBC News implicitly took this position when it punished him by suspending him without pay for six months. Several other well-known people, such as Hillary Clinton, have been caught in similar fabrications, and several of Williams' other recollections have come under suspicion.

Cognitive psychologists point out that the average person dramatically underestimates the malleability of human memory. False memory syndrome refers to cases in which people have an important memory that they truly believe, but which is factually incorrect. Some people have been convicted of crimes they did not commit, such as sexual abuse of children, on the basis of memories “recovered” by alleged victims, usually with the aid of mental health practitioners. Other innocent defendants have been convicted of crimes to which they confessed under police interrogation, and of which they believed they were guilty, at least for a time. A new study by Julia Shaw and Stephen Porter provides some of the best evidence yet for the construction of detailed and important false memories.

The authors used a paradign developed by Elizabeth Loftus for her “lost in a shopping mall” study, in which some college students were induced to develop this false memory. After 126 Canadian students had volunteered to participate in the study, their parents or primary caregivers were interviewed. The purpose of the interview was to identify one highly emotional event of a non-criminal nature that had occurred to the students during their adolescence, and to ascertain that they had never been arrested and charged with a crime. The first 60 students who met these criteria participated in a series of three interviews, each about one week apart.

In the first interview, they were asked to recall and describe the true event and a false event allegedly reported by their parents. In half the cases, the false event was that they were arrested and charged with a crime—ten cases each of assault, assault with a weapon, and theft. The other thirty false events were non-criminal: an accidental injury, being attacked by a dog, and losing a large amount of money. Since none of the participants reported remembering the false event during the first interview, they were given false contextual cues, were asked to try to recover the memory, and were given advice as to how to recover it. They were asked again to describe both events and answer several questions about them during the second and third interviews. At the end of the third interview, they were fully debriefed.

Participants were said to have accepted the false memory if they met two criteria: (1) they reported ten or more details of the incident not suggested by the interviewer, and (2) they reported during the debriefing that they truly believed the event had occurred. Forty-four of sixty participants (73%) met these criteria. The percentages did not differ very much among the six types of false events.

The researchers were impressed not only by the large number of participants who accepted the false memory but also by their “richly detailed” descriptions of it. They compared responses to the true and false memories among the 44 participants who experienced the false memory. These students recalled more details of the true memory (92, on average) than the false memory (72)—but 72 details is still fairly impressive for a false memory. They had greater confidence in the true memory, and experienced it as more vivid, but the amount of anxiety generated by the true and false memories did not differ.

There are some important differences between Brian Williams' false memory and those created in the Shaw and Porter study. As far as we know, Williams' memory was not prompted by anyone else. However, Loftus has suggested that he might have been influenced by “audience tuning.” To paraphrase Jon Stewart, audience responses to his tale might have excited the pleasure centers in his brain and caused his story to drift toward greater bravery over time. In addition, the memories created in the study were not self-serving. Just the opposite. They remembered being charged with a crime. Would their misremembering have been even greater if the memory had been flattering to them?

Leslie Savan expressed her satisfaction that at least someone is being punished for lying about our invasion of Iraq, even if it's not one of the right people. Others have spoken about the relative unimportance of Brian Williams' battlefield exaggeration when compared to the many other lies he routinely told on behalf of the government and his corporate sponsors every weekday night. In fact, we might speculate that NBC's main reason for suspending him was not that he told a false story but his loss of credibility, which might adversely affect public acceptance of their nightly propaganda report.

Pittsburgh Post-Gazette columnist Tony Norman reports that he was in the Monroeville Mall last week when three people were shot, although he didn't see or hear the incident. When he posted his experience on Twitter, one person wryly commented, “in five years, he will say he was shot.”

By the way, did I ever tell you about the time I rescued a dog from a burning building?

You may also be interested in reading:

Hey, Wait a Minute!

A Stroll Down Memory Lane

Hey, Wait a Minute!

On February 1, New York Times columnist Nicholas Kristof published an op-ed entitled “Dylan Farrow's Story,” containing excepts from a letter by Ms. Farrow which repeated allegations that her adoptive father, Woody Allen, sexually molested her in 1992 when she was seven. She claims to still suffer from post-traumatic stress. Allen denies the allegation. A police-appointed medical team concluded that she was not molested and the charges were dropped in 1993.

Kristof's column has been followed by several articles by people claiming to be feminists and progressives who argue that Ms. Farrow should be believed and Woody Allen should be ostracized. I find this disturbing both as a social psychologist and as a citizen. As a psychologist, I am aware of research on the false memory syndrome—a condition in which people's lives are affected by memories that are objectively false, yet strongly believed. Although many people find the idea of false memories counterintuitive, their existence is well-documented. We should be extremely skeptical of claims of sexual abuse, especially of children, unless there is corroborating physical evidence. These expressions of outrage directed at Woody Allen appear to disregard his legal presumption of innocence and to recall some of the moral panic surrounding child abuse in the 1980s.

False memories have been documented by studies in which a number of children were exposed to the same experience. Some of them were randomly assigned to an experimental group that receives misleading post-event information, while the remainder served as the control group. For example, in one study, 3-year-old children were given a routine physical by a doctor that did not involve a genital exam, but following leading questioning, 70% of the children falsely reported that the doctor touched their genitals. Similar results have been found with older children.

Stephen Ceci and Maggie Bruck summarize a number of variables that have been shown to produce false accusations by children:

• Suggestive questioning. A suggestive question presumes a particular answer, such as “Did he touch your pee-pee?” rather than “What happened next?”
• Repeated questioning. Children may deny abuse at first, but if you keep asking the same question, they may assume their first answer was wrong and switch to “yes.”
• Selective reinforcement. If the interviewer ignores denials but reinforces accusations with attention and praise, the child will make further accusations.
• Stereotyping the suspect. This refers to telling the child that the defendant is a bad person.
• Telling the child that others have reported abuse. This increases conformity pressure.
• Anatomically correct dolls. The child's attention is riveted on the one feature that makes them different from other dolls. They are very effective in producing false accusations and should never be used.

Both Ceci and Elizabeth Loftus have shown that very complex and detailed false memories can be created in adults as well as children. Here's a preview of a very professionally done video about Ceci's research on children's memories. The 37-minute video can be found here.

What about the accusations against Woody Allen? The alleged incident occurred after Woody Allen and Mia Farrow ended their 12-year relationship when Allen announced his intention to marry Farrow's adopted daughter, Soon-Yi Previn. Not surprisingly, Mia Farrow has hated him ever since. On the day in question, Allen visited Farrow's home in Connecticut to try to resolve their custody dispute. In a house filled with people, many of them hostile to him, Allen is alleged to have taken Dylan to the attic and touched her inappropriately. Afterwards, it was claimed that her underpants were missing. 

When Mia Farrow questioned her, she is said to have told Mia what happened. Mia recorded the accusations on a videotape that contains several stops and starts, and took it to the police. In subsequent interviews, Dylan changed her story several times. The court assigned an investigative team from the Yale-New Haven Hospital headed by Dr. John Leventhal to investigative the charges. Here are some quotes from Dr. Leventhal's testimony:

• “Even before the claim of abuse was made last August, the view of Mr. Allen as an evil and awful and terrible man permeated the household.”
• “Those were not minor inconsistencies. She told us initially that she hadn't been touched in the vaginal area, and she told us that she had, then she told us that she hadn't.”
• “We had two hypotheses: one, that these were statements made by an emotionally disturbed child and then became fixed in her mind. And the other hypothesis was that she was coached or influenced by her mother. We did not come to a firm conclusion. We think that it was probably a combination.”

One commentary on the case claims that if you believe Allen is innocent, you are calling Dylan a liar. This is not the case. When children are prompted to make false reports, they believe those accusations to be true and sometimes suffer traumas similar to those they would have experienced had they actually been molested. The child has actually been victimized by whoever was responsible for planting the false memory.

Another claim is that failure to believe Dylan Farrow's story is symptomatic of a patriarchal society and that siding with Woody Allen shows that you have been “rape cultured.” In fact, both men and women have been falsely accused and sometimes convicted of sexually abusing both boys and girls. Their guilt or innocence is an empirical, not an ideological, question. The cause of feminism is not helped by false accusations of male abuse of women and girls.

Woody Allen's reputation for eccentricity probably makes it more likely the public will believe this story. But the award for the most unintentionally funny commentary goes to an author who cherry-picked scenes from Woody Allen's 49 films and a segment from one of his many interviews to suggest that they are evidence of his guilt. And speaking of awards, does anyone believe that the timing of Kristof's column was merely coincidental? Will Cate Blanchett lose the Best Actress award as a result?

The probability that Woody Allen is guilty is not zero. However, I think it's much closer to zero than most people think.

Update (2/9/14):

The New York Times has published a reply by Woody Allen in which he blames Mia Farrow for coaching their daughter to make the accusation against him.

You may also be interested in reading:

"We Need You to Tell Us That This is What Happened"

"We Need You to Tell Us That This is What Happened"

Not surprisingly, news coverage of Tuesday's defense presentation at the Jerry Sandusky trial focused on the testimony of Dorothy Sandusky. People are understandably interested in the reactions of a woman who, if her husband is guilty, has been publicly humiliated. However, there was other testimony that day that is potentially more important, testimony about the possibility of suggestive questioning of alleged victims by the Pennsylvania state police.

First of all, it is important to remember that Victim 4 was the prosecution's star witness. He was the first alleged victim to testify, and he gave the most detailed account of his alleged molestation. He is the one who claimed that oral sex happened “almost every time I'm over there. And I'm over there two or three times a week.”

The following four paragraphs are from the inside page of Wednesday's Pittsburgh Post-Gazette, about halfway through the article.

Earlier in the day, Mr. Amendola [Mr. Sandusky's attorney] called state police Cpl. Scott Rossman and retired Cpl. Joseph Leiter in a bid to show jurors that investigators led the accusers to make up stories about the alleged abuse. He played a tape of an April 21, 2011, interview in which Cpl. Leiter tells the man who later would become known as Victim 4 that he was not the first person interviewed, that there were nine others, and that their stories were similar.

“There's a pretty well-defined progression in the way he operated and still operates I guess to some degree,” the officer said. He went on to tell the would-be accuser that in some cases, Mr. Sandusky allegedly went beyond touching or fondling children. “There have been acts of oral sex that have taken place,” and an incident “classified as rape.”

“I don't want you to feel ashamed because you're a victim in this whole thing. What happened happened. He took advantage of you. We need you to tell us that this is what happened. We need you to tell us as graphically as you can what took place,” the corporal said.

Under questioning by the lead prosecutor, Senior Deputy Attorney General Joseph McGettigan, Cpl. Leiter defended his method and said he never intended to elicit anything but the truth from the interview.

Although I only know what I read in the paper, this seems almost like a textbook illustration of how not to interview a possible victim of child abuse.

In the ' 80s and '90s, there were a number of high profile cases—such as the Little Rascals Daycare in North Carolina and the McMartin Preschool in California—in which adults were falsely convicted of molesting children on the basis of testimony elicited either by police officers or by psychologists and social workers employed by the prosecution. The existence of false memories is well documented by research. The false memory syndrome refers to cases in which people who sincerely believe they have been sexually abused, but where their recall of abuse is apparently a memory disturbance, or confabulation. Elizabeth Loftus, among others, has shown that adults can be induced by repeated suggestive interviews to have highly emotional false memories about events their childhood that actually never happened.

At the root of suggestive questioning is the fact that interviewers often exhibit confirmatory bias—the tendency to search for information that confirms their beliefs and ignore information that might disprove them. Stephen Ceci and Maggie Bruck have summarized the research literature on interviewer behaviors which encourage false accusations of child abuse. Among them are:

• Stereotyping the suspect as a bad person.
• Telling the witness that peers have reported abuse.
• Suggestive questioning. A suggestive question refers to a possible answer in the question itself, i.e., “Did he touch your privates?” rather than “What did he do next?” In the worst cases, interviewers describe an abuse scenario reported by others and ask the witnesses whether something like that happened to them.
• Repeated questioning. Interviewers ask the same questions several times until they get the desired response.
• Selective reinforcement of responses. We usually think of reinforcement as attention and praise that occurs after a response. However, reinforcement can be anticipated before the fact when interviewers state or imply that they will be pleased if the witness can make an accusation.

All of these conditions can be read into the description of yesterday's testimony except repeated questioning. The New York Times fills in that particular blank.

The interviewers acknowledged that some of the witnesses at first insisted that nothing untoward occurred and told of abuse only on a second or third police interview, but that such reluctance was not unusual.

It is important to remember that this is not just hearsay. We don't have to speculate about what was said to Victim 4. The interview is on tape. How many other interviews were tape recorded? Was the same interrogation strategy followed with other alleged victims? How many suspected victims were interviewed who did not report abuse despite repeated questioning?

The fact that the investigators knew that the tape was running makes it clear that they were not consciously attempting to falsely implicate Mr. Sandusky. It suggests instead that they were poorly trained in the technique of interviewing alleged victims of sexual abuse and were unaware of the implications of their behavior.

The existence of suggestive questioning does not directly address the question of Mr. Sandusky's guilt or innocence. In either case, this is an unfortunate situation. If he's actually guilty, the incompetence of these officers could have jeopardized the case against him. If he's innocent, it may have contributed to a false conviction.