Today it was reported that Tonya Craft, a former kindergarten teacher accused of molesting three young girls, was acquitted of all 22 charges against her. Although the accusations against Tonya were startling in their own right, what made this case even more scandalous was that one of the accusers was Tonya's own daughter. What could lead children to wrongfully accuse their teacher of molestation, and what could lead a daughter to make the same accusation about her own mother? A wealth of research has been conducted on this very topic and provides insight into how people come to make false accusations.
Clearly, there are situations where eyewitnesses are coerced, intimidated and even threatened by law enforcement officials to knowingly make false accusations against a defendant. However, such extreme examples probably represent the exception rather than the rule. Far more likely are situations where eyewitnesses are subtly "pushed" into truly believing a crime occurred when it did not. Social psychologists have long been interested in understanding the variables that can cause such false beliefs. The heart of the problem lies in the fact that contrary to popular belief, our minds do not work like video cameras that record, store and play back our memories with perfect accuracy. Instead, the three stages of memory processing—acquisition, storage, and retrieval—are highly susceptible to error and bias. As a result, false information can make its way into our memories at each step in the process.
The first phase of memory processing is acquisition. This is the stage where eyewitnesses take in information about a crime. There are a number of factors that limit the amount of information that gets past this first stage, yet we typically fail to take such factors into account when judging how confident we are in our recollection of events. For example, if you were to witness someone else being mugged, research suggests your ability to accurately identify the perpetrator will be severely limited if the crime occurred quickly and unexpectedly (aren't all crimes unexpected?), if the crime happened far from view (you see the crime from a few blocks away), and if it happened under poor visibility conditions (at night, rainy weather, etc.). For example, Loftus and Harley (2005) found participants started to have difficulty identifying photos of famous celebrities (e.g., Julia Roberts) when viewed from a simulated distance of 25 ft. And when viewed from 77 ft., only 25% could accurately identify the photos.
Although these factors can clearly lead to inaccuracies in eyewitness testimony, such biases in acquisition do not do a good job of explaining the Tonya Craft case. To better understand what may have occurred in that case, we need to move on to the next phase of memory processing.
The second phase of memory processing is storage. At this stage, information gathered in the first stage is stored in our memory. Although we tend to assume that such information is "filed away" like a letter or photograph that remains untouched or unaltered until we retrieve it, this is not the case. A wealth of research on the topic of reconstructive memory shows that our memories are not indelible and that they can be altered by events that occur after the initial memory event.
For instance, in a classic study by Elizabeth Loftus and colleagues (1978), participants were shown photos depicting different stages of an automobile accident. Some were shown a photo of a car stopped at a stop sign, while others saw an almost identical photo that had a yield sign instead. After viewing the photo, half of the participants were asked "Did another car pass the red Datsun while it was stopped at the stop sign?" and the other half were asked "Did another car pass the red Datsun while it was stopped at the yield sign?" That is, for half of the participants, the question accurately described the traffic sign in their photo, and for half the wording subtlety differed from what they had actually seen. Later, the participants were shown both photos (one with the stop sign and one with the yield sign) and were asked to recall which once they had originally seen. Of those who were asked the question that correctly matched their photo, 75% chose the correct photo. However, for those who were asked the misleading question, only 41% chose the correct photo. This shows that slight changes in the wording of questions can influence what eyewitnesses later remember.
Similarly, Loftus and Zanni (1975) found that eyewitnesses who were asked "Did you see the broken headlight" were twice as likely to have false memories than those who were asked "Did you see a broken headlight." The problem is that first question assumes there was a broken headlight, whereas the second question does not.
So why does this error occur? The likely answer is that people who received the misleading question now have two pieces of information in their memory (e.g., the stop sign from the photo and the yield sign from the question). This is fine, as long as people remember where these two pieces of information came from, but people often mix up where they heard or saw something. I bet you can easily think of a time where you remembered a fact or quote, but could not for the life of you remember where you heard it from. That same process leads eyewitnesses to confuse the sources of their memories. When you combine this perceptual bias with the fact that the eyewitnesses in the Tonya Craft case were kindergarten children being interviewed by authoritative police officers and lawyers, it makes sense that these young eyewitnesses could have easily become confused in their recollection of events. For this reason, it is important that officials who interview eyewitnesses do so carefully, making sure not to use questions that include misleading information or assumptions. Interviewers should, whenever possible, use questions that allow eyewitnesses to offer their own unprompted recollections. So rather than asking a child "Where did person X touch you?" interviewers should instead ask "What happened when you were with person X?"
The third phase of memory processing is retrieval. During this final stage, people attempt to recall information that has been stored in their memory. Just as with the previous two stages, there are often problems with how people retrieve eyewitness information from long-term memory.
Imagine you are sitting in a lecture hall waiting for class to start and you see someone duck into the room, grab the professor's laptop from the podium and run out of the room with it. Given the unexpected nature of this event, you probably wouldn't get a good look at the person who committed the crime. At best, you could probably tell what their gender was and an approximate height, but not much else. Now imagine you later were asked to look through a series of photos and pick out the person who committed the crime. This series of photos to pick from is what police officials call a lineup. Research shows that features of the lineup options can actually influence how accurate your identification will be. In fact, the majority of research on retrieval errors has focused on lineups and this is not surprising considering that lineup errors are the most common cause of wrongful convictions in the US.
The main error that occurs with lineups is something called the "best guess problem." This refers to people's tendency to pick the person in the lineup that most closely resembles what they think the criminal looks like. For instance, if you recalled that the person who stole the professor's laptop was a man with dark hair, you will probably look through the photos and pick the person in the lineup that most closely fits this description, even if the rest of his features do not identically match your memory.
To avoid this problem, researchers suggest that police officials avoid simultaneous lineups (witness views all lineup members at same time) and should instead use sequential lineups (witness views each lineup member individually, making a yes/no decision about each one before proceeding to the next). This avoids the comparison strategy that inevitable leads to the best guess problem.
In addition to the problem of lineups, the retrieval phase can also be influenced by the process of simply retelling the event. When we are asked to retell a past event, that retelling also gets entered into memory. Just as participants in the Loftus and Zanni study got confused as to whether they saw a stop sign or were just asked about it later, people can also get confused as to whether they actually witnessed a particular crime or were just asked about it at a later date. Therefore, the more we tell a story, even if the story is false, the more we become committed to that particular recollection.
To investigate this possibility, Wells and colleagues (1981) actually staged a theft in their class that closely matched the example described above. After students witnessed the "crime," they were asked to pick out the criminal from a stack of photos. Most students identified one of the photos as the perpetrator, even though in truth the real perpetrator was not featured in any of the photos. Next, the students were told they would have to take the witness stand to testify about what they saw and were instructed to rehearse their answers beforehand. The researchers found that when participants rehearsed their answers, they became more confident about what they saw, even though all the eyewitnesses were wrong. Even more concerning is that when these eyewitnesses did take the stand in a mock trial, the jurors were more convinced by the eyewitnesses who had rehearsed their testimony, and therefore were more likely to convict the innocent defendant.
In some cases, eyewitnesses are persuaded by police interviewers or lawyers to enhance their testimony or embellish it with details that they were not originally aware of. For example, a man working at the car rental shop where Timothy McVeigh rented the truck used in the Oklahoma City bombing incident was pressured by police to recall who accompanied McVeigh on that day. Under this intense pressure, the mechanic was persuaded to come up with specific details for the second bomber (muscular build, black T-shirt) that later turned out to be false. The work by Wells suggests that the more this eyewitness was asked to recall this description, the more confident he became in his recollection.
In sum, we all suffer from the illusion that a witness's mind records and retrieves memories in a manner that is free from error. But this is not the case. Our memories are highly susceptible to bias and misinformation can creep in at any of the three stages in the memory process. To avoid such errors, research suggests that those involved in every level of the judicial system should be cautious when interviewing eyewitnesses and designing lineups.
Be sure to read the following responses to this post by our bloggers: