Postevent Information and Eyewitness Memory

Human memory, however accurate generally, is not a perfect processing system. Over time, our memory becomes less accurate, primarily for two reasons. First, our memory is not permanent, and information fades from memory over time. Most people are familiar from experience with this unfortunate feature of memory but are less familiar with the second factor that influences the accuracy of memory—memory can be distorted by the influence of postevent information. Although memory can be influenced by subsequent experiences, there are constraints on the conditions under which this is likely to occur. Nonetheless, when memory accuracy is a premium, such as in forensic situations involving eyewitness memory, it is important to recognize that eyewitness memory can be suggestively influenced. In these situations, the impact of postevent information should be minimized by avoiding misleading questions, and when it is relevant to do so, jurors should be informed about the potential fallacies in eyewitness memory that can result from a suggestive interview.

How can postevent information influence memory? Take the example of eyewitnesses who observe a convenience store robbery. From their observations, they construct a memory for what transpired during the robbery. Most of the time, it is this memory that police officers want to examine. However, virtually all eyewitnesses to crimes who eventually testify in court are interviewed by police officers at least once and typically multiple times. In police interviews, the eyewitness is questioned about what happened, and if the investigating officer has specific suspicions about what occurred, the interview may include some leading questions (e.g., “Was it a white four-door sedan?” “Was he or she wearing athletic clothing?” “Was he or she carrying anything?” “May he or she have had a gun in his or her hand that was in his or her jacket pocket?”). Questioning such as this presents one source of postevent information. Another source of postevent information is self-generation; that is, the eyewitness may introduce new information by just thinking about or talking about the robbery. Either way, the postevent information affects one’s memory of the original observed event, and over time, individuals become less able to differentiate between the information that is in their memory because it was actually observed and the information that was introduced after the event by postevent information.

Most of the time, the influence of postevent information is minimal and inconsequential. However, in the case of an eyewitness to a crime, when it is important to know exactly what transpired, postevent information may be an important source of memory error. For this reason, the distortion in memory that results from postevent information is often referred to as the misinformation effect.

Research on the Effect of Postevent Information

There is a great deal of research on the effect of postevent information, much of it spawned from the early work of Elizabeth Loftus. In a typical experiment on this topic, participants first view a sequence of slides, a videotape, or a film of an event. After viewing this event, they read a narrative or are asked some questions that intentionally mislead them about the identity of a small set of target items viewed in the original event (the misled condition), or they do not receive the misleading information (the control condition). The principal result is that the participants are more accurate recognizing the original target item in the control condition than in the misled condition; that is, they are misled by the postevent information presented in the narrative or questions.

In a related paradigm for studying the effect of postevent information, Elizabeth Loftus had individuals view a video of a traffic accident. They were subsequently asked questions regarding how fast the cars were traveling prior to the accident. Whereas individuals who were asked, “About how fast were the two cars going when they hit each other?” provided a mean response of 34 mph, individuals asked, “About how fast were the two cars going when they smashed each other?” responded with a mean speed of 40.8 mph. Surprisingly, the wording of this question—the postevent information—also affected their memory of the car accident more generally. One week later these individuals were questioned again. When asked whether they had seen broken glass in the film, whereas 34% of the individuals who had received the question, including the word smashed, indicated having seen broken glass in the film, only 14% of those who received the hit question reported remembering broken glass. The impact of postevent information is thus not limited to immediate questioning but can have long-term consequences as well.

Cognitive Interpretations of the Effect of Postevent Information

There are several explanations for how postevent information influences memory, and there is evidence that the cognitive mechanisms underlying each of these explanations play some role in the misinformation effect. Take for an example Elizabeth Loftus’s now classic study in which participants viewed a car approaching an intersection with a stop sign. In the misled condition, the sign was later suggested to be a yield sign (“Did another car pass the red Datsun while it was stopped at the yield sign?.”). In a subsequent test, individuals were less likely to recognize the stop sign if they had been in the misled condition. One interpretation of this result is that the postevent information impairs memory of the observed event. That is, being presented the yield sign literally degrades and replaces memory of the observed stop sign. A second interpretation of the misinformation effect is that the original information (stop sign) and the suggested information (yield sign) are both present in memory, but individuals make what Marcia Johnson calls a source-monitoring error and confuse what was seen with what was subsequently suggested. A third interpretation of the effect of postevent information is that the misleading postevent information simply substitutes information in memory when there is no accessible memory for the relevant details of the original event. In fact, even when we pay attention, we do not remember everything about an event that we observe. According to this interpretation, only if individuals cannot remember seeing a stop sign are they likely to incorporate the suggested yield sign.

This latter interpretation, referred to as the trace strength theory of suggestibility, has important implications for some of the constraints on the effect of postevent information. Although memory is a constructive process, and we can incorporate postevent information into our memory for what we observe, in fact our memory is generally quite accurate. What is the vehicle by which the veracity of memory is preserved? It is now clear from a number of research studies that stronger memories are more likely to resist suggestion than weaker memories; we are less likely to be misled about details of events if we saw them very clearly to begin with and remember them well than if we did not remember them well. Kathy Pezdek conducted a research study in which children were presented a sequence of slides depicting an event. The target slides that were presented once or twice each were included in the sequence. A postevent narrative was then read that described the same event; in the nar-rative were several misleading sentences that suggested a change in what had been observed in the slides. The children were significantly less likely to be misled by the postevent narrative if they had observed the target slides twice rather than once; the stronger memory was more resistant to suggestibility. This conclusion is also supported by the results of other studies in which it has been reported that children are less suggestible in domains in which they have greater knowledge (i.e., greater memory strength).

Can Postevent Information Plant False Events in Memory?

Research on the influence of postevent information is often used to imply that it is relatively easy to suggestively mislead a person to believe that an entirely new event had occurred when it had not. This assumption is at the heart of the false-memory debate. This debate concerns the veracity of delayed memory for incidents of childhood sexual abuse. Analyses by Jennifer Freyd and her colleagues have established the pervasiveness of childhood sexual abuse. Nonetheless, some claims of childhood sexual abuse might be based on “false memories”; that is, the abuse never occurred, but memory of the abuse was planted by postevent information, such as suggestive questioning by overzealous police officers, social workers, or therapists who interviewed the child.

According to the trace strength theory of suggestibility, if a child is recalling an event that was experienced several times, he or she would be expected to have a more accurate memory of the event and be less vulnerable to suggestive influences such as biased interviewing procedures than if the event had occurred only once. This is especially important in child abuse cases because it is common for a perpetrator to frequently abuse the same child. A child’s memory for an incident that occurred frequently would be expected to be relatively reliable, even in the face of possibly suggestive interviewing.

Another important constraint on the effect of postevent suggestion is that implausible events are less likely to be suggestively planted in memory than plausible events. In one study on this point, Kathy Pezdek and her colleagues had 20 confederates read descriptions of one true event and two false events to a younger sibling or close relative. The more plausible false event described the relative being lost in a mall while shopping; the less plausible false event described the relative receiving an enema. Three events were falsely remembered: all were the more plausible events. Thus, all memories are not equally likely to be suggestively planted in memory, and individuals with differing prior experience and prior knowledge are not equally vulnerable to suggestion. The relative ease of suggestively planting false childhood memories of sexual abuse versus being lost in a mall while shopping should be related to the relative plausibility of these two events to a particular individual. This research demonstrates that although memory can be influenced by postevent information, there are constraints on the conditions under which this is likely to occur.

References:

  1. Freyd, J. J., Putnam, F. W., Lyon, T. D., Becher-Blease, K. A., Cheit, R. E., Siegel, N. B., et al. (2005). The science of child sexual abuse. Science, 308,
  2. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4, 19-31.
  3. Pezdek, K., & Banks, W. P. (Eds.). (1996). The recovered memory/false memory debate. San Diego, CA: Academic Press.
  4. Pezdek, K., Finger, K., & Hodge, D. (1997). Planting false childhood memories: The role of event plausibility. Psychological Science, 8, 437-441.

See also: