In the next sections we summarise research on some specific factors influencing the encoding, storage and retrieval stages of memory, focusing on recall and eyewitness testimony. Identification evidence is covered in detail in Chapter 15.
6.4 ENCODING FACTORS
In this section we address the factors affecting the quality and accuracy of memory encoding. We focus on the influence of stress, weapon presence, change blindness, stereotyping and intoxication. We also draw your attention to the importance of more stable witness characteristics, such as age, in influencing the memory encoding stage.
6.4.1 Stress
During a crime, it is likely that witnesses will experience relatively high levels of stress, which can affect the clarity and accuracy of the memory they encode. A number of studies have investigated this factor, with the overall results indicating that stress can have a negative effect on memory recall accuracy (Deffenbacher, Bornstein, Penrod, & McGorty, 2004). Research on stress and memory often refers to the Yerkes-Dodson Law (1908), which plots the relationship as an inverted U-shaped curve. This law may help to explain some of the contradictory results in the literature. According to this law, memory performance is optimal under moderate stress levels, but declines at the two extremes of physiological arousal.
In a laboratory setting it is difficult to simulate the level of stress involved with witnessing an actual crime because of ethical considerations (Deffenbacher et al. 2004). As a result, researchers have had to be creative when designing studies. For example, Valentine and Mesout (2009) asked visitors to a scary tourist attraction, the Horror Labyrinth in the London Dungeon, about their visit. Participants were asked how stressed they felt and to describe an actor they had encountered. Individual differences in reported stress predicted description accuracy. People with high levels of recorded anxiety (indicated by heart rate) whilst in the Labyrinth provided a greater number of incorrect descriptors and fewer correct descriptors. Morgan, Southwick, Steffian, Hazlett and Loftus (2013) tested military personnel whilst they were confined in a mock prisoner of war camp as part of survival training. Memories formed in this stressful situation, which included food and sleep deprivation, were vulnerable to changes following exposure to subsequent misinformation. We consider how memories come to be modified by misinformation later in the chapter.
Stress does not always affect recall accuracy. As we have already seen, witnesses to a real crime remembered the events accurately and reliably (Yuille & Cutshall, 1986). In fact, the most stressed witnesses recalled the event more accurately than those who reported being less stressed (Yuille & Cutshall, 1986). However, outside the lab it is difficult to control for other variables; people who were more stressed were also closer to the shooting. They may simply have been able to encode the event more clearly.
On the other hand, it is also important to consider the relationship between memory accuracy and the type of information recalled in stressful conditions. Traumatic memories, which by definition are extremely stressful, are often very accurate (Reisberg & Heuer, 2007). Some psychologists have presented evidence that in times of high stress and physiological arousal, attention narrows to central aspects of an event (e.g., Safer, Christianson, Autry, & Österlund, 1998). This means that more peripheral aspects are neglected, and encoded less well (Easterbrook, 1959), whilst central information may be remembered very accurately (Christianson, 1992). A recent study by Hope et al. (2015) contradicts this explanation, but offers evidence in support of the inverted U-shaped curve linking stress and recall accuracy. Pairs of police officers took part in a realistic scenario involving an armed assailant; one officer was armed and confronted the assailant while the second acted as an observer. The armed officers exhibited higher levels of physiological arousal but remembered central features of the incident less well than the observers, who showed lower levels of stress. Interestingly, a fifth of the officers reported that the assailant had drawn a weapon and pointed it at them, but this never occurred.
PHOTO 6.1 Witnesses who experience “weapon focus” attend to the weapon but fail to notice details of the offender’s appearance.
Source: © Vetta/iStockphoto
6.4.2 Weapon Focus
In crimes with violence or a threat of violence, it is likely that a deadly weapon, such as a knife or gun will feature. It is easy to imagine that if you were confronted with a gun or a knife, you might focus a lot of attention on the weapon due to fear or surprise, but fail to notice other details that could be relevant to the police. This is exactly what psychologists have found when investigating whether the presence of a weapon influences memory in the laboratory (e.g., Pickel, 1998; Hope & Wright, 2007). The impact of weapons in actual crimes is less clear-cut, which might suggest that the weapon effect may not be as large as psychology studies suggest (Cutshall & Yuille, 1989; Behrman & Davey, 2001). However, unless the crime was recorded on CCTV, there are problems with using real crimes to make judgements about how memory operates, as it is impossible to compare the witnesses’ memory to what actually happened (the ground truth) (Davis & Valentine, 2009). A recent review of the existing literature set out to reconcile laboratory and real-life findings. The results support the argument that overall, weapon focus is an important factor in eyewitness memory accuracy (Fawcett, Russell, Peace, & Christie, 2013). The review also indicated that the weapon focus diminishes the longer the witness is exposed to the perpetrator. As exposure length increases, witnesses shift their attention away from the weapon and towards the perpetrator.
Two possible explanations have been put forward for the weapon-focus phenomenon. The first is in keeping with the cue utilisation hypothesis (Easterbrook, 1959), and argues that people narrow their attention to the weapon because of the threat associated with it. Consistent with the threat hypothesis, research with children showed that the presence of a “frightening” object (a syringe full of red liquid) during an experiment resulted in them remembering less about the researcher’s appearance than those who encountered neutral objects (Davies, Smith, & Blincoe, 2008). On the other hand, if the threat explanation totally explained weapon focus, we would expect the effect to increase in line with the level of threat. This does not appear to be the case (Pickel, 1998, 1999). Another possible explanation is that the weapons capture attention because it is so unusual to encounter one in everyday life. Studies have found that unusual items specific to a certain context (e.g., a man entering a store holding a feather duster: Hope & Wright, 2007) have a similarly detrimental effect on memory. In comparing the effect of unusual non-weapon items to weapon items across studies, Fawcett et al. (2013) found no difference in the extent to which they impaired memory.
Interestingly, being aware of the existence of a weapon focus may enable people to mitigate its effect, and to orient their attention to other important aspects of the crime scene. Pickel, Ross, & Truelove (2006) informed their participants about the phenomenon by way of a lecture. People listening to the weapon-focus lecture recalled more correct details and less incorrect details about a subsequent event featuring a weapon, than people listening to a lecture about a different topic.
6.4.3 Change Blindness
Whilst observing a crime, you might shift your attention around, looking in one direction then another, to try and make sense of what is happening. Research suggests that if a change occurs in the brief time you are not looking, you might not notice it when you revert your attention. You might assume that this only applies to small changes, but you would be wrong. In one study, 61% of people reported that they did not notice that the identity of a burglar changed after the camera angle changed halfway through a crime video (Davies & Hine, 2007). Participants who detected the identity change tended to have greater recall accuracy compared to those who did not. Police personnel, assumed to be skilled observers of crime, are just as susceptible to change blindness as laypeople (Smart, Berry, & Rodriguez, 2014).
This surprising phenomenon is caused by lapses in perception. Levin, Simons, Angelone, & Chab
ris (2002) explain change blindness as being due to the lack of a well-established object/person representation, which might allow the new representation of the object/person to overwrite the first. Change blindness is particularly likely to occur when witnesses make assumptions about continuity. For example, participants were less likely to notice changes in identity when a crime event featured a continuity illusion: an “innocent” person walked behind a stack of boxes and the perpetrator emerged from the other side (Davis, Loftus, Vanous, & Cucciare, 2008). As yet, very few studies have investigated change blindness in a forensic setting. More research is needed to help establish when witnesses are most at risk of committing these kinds of perceptual errors (Laney & Loftus, 2010).
6.4.4 Stereotyping
Crime events can be highly complex and difficult to make sense of, resulting in a high cognitive load at encoding. Available information from the crime alone can be confusing, so as a way of creating a more coherent memory, people might unknowingly make use of schemas and scripts. Thanks to books, films and TV, we probably all have a representation of what might happen during a shoplifting incident, what the offender would do, and how he/she would act. This representation might be used to try and fill memory gaps (Tuckey & Brewer, 2003b).
The higher the cognitive load, the more likely people are to employ stereotypes to increase coherence. Van Knippenberg, Dijksterhuis, & Vermeulen (1999) activated social stereotypes by describing the suspect in a crime scenario as either a drug addict (negative condition) or a bank worker (positive condition). In the negative condition, participants who had encoded the information under a high cognitive load remembered more incriminating details than exonerating details. These results show memory operating in keeping with the negative stereotype.
Whether or not information is consistent or inconsistent with stereotypes and schemas influences how well it is remembered (Tuckey & Brewer, 2003a). Older adults appear to be particularly susceptible to stereotype and schema-based processing of crime information (Overman, Wiseman, Allison, & Stephens, 2013), with less typical features of a crime being particularly likely to be forgotten by older witnesses (García-Bajos, Migueles, & Aizpurua, 2012).
It is also necessary to consider the facilitative effects of stereotypes on memory performance. Encoded information that is consistent with a stereotype or schema might actually be preserved very well (Tuckey & Brewer, 2003a). In fact, there is evidence that when people try to repress stereotypical thinking they make memory errors. Half of Peters, Jelicic and Merckelbach’s (2006) participants were told to avoid responding stereotypically, but to be as accurate as possible in response to questions about a violent crime. When asked whether the main suspect was foreign, participants who were told to suppress their stereotypes and had been primed by observing a foreign group photograph were more likely to make recognition errors compared to those exposed to a neutral prime.
6.4.5 Intoxication
When a crime occurs, it is relatively common for the witness or victim to be intoxicated (Evans, Schreiber Compo, & Russano, 2009; Palmer, Flowe, Takarangi, and Humphries, 2013). This might be particularly true of sexual or violent assaults because these types of crime are more likely to occur at night (Office of National Statistics, 2011/2012). In an early eyewitness study involving alcohol, Yuille and Tollestrup (1990) observed a detrimental effect on the detail of immediate recall for a staged theft. Furthermore, after a week-long retention interval, participants who were intoxicated during encoding recalled less, and were less accurate than their sober counterparts.
According to the alcohol myopia framework (Steele & Josephs, 1990; Steele & Southwick, 1985), intoxication affects attention and encoding by narrowing focus to central details, at the expense of peripheral details. In one of the few studies testing how alcohol influences the recall of an interactive event, Schreiber Compo et al. (2011) reported that intoxicated participants were more likely than sober participants to remember central details, such as the bartender’s blonde hair, but were less likely to recall peripheral details, such as whether there was a dartboard in the bar.
Some studies, however, do not support the myopia hypothesis. For instance, the memory of Harvey, Kneller and Campbell’s (2013) intoxicated participants was equal to that of their sober group in terms of both central and peripheral features of a photo. Some results even suggest that people’s preconceptions about the effect of alcohol has on memory may be more detrimental than the alcohol itself (Assefi & Garry, 2003). However, this finding has not been supported by subsequent research by Clifasefi, Takarangi, and Bergman (2006). In this study, whilst all participants were told that they were drinking alcohol, only participants who actually drank alcohol failed to notice an unexpected event, while those who were in fact sober, did notice it.
Other research highlights differences between intoxicated and sober participants in their completeness and accuracy of memory. Flowe, Takarangi, Humphries, & Wright (2015) found that although participants who were intoxicated during the encoding of a sexual assault scenario provided significantly less information overall than their sober counterparts, the accuracy of the information they provided did not differ one day or four months later. These results suggest that people who witness crimes whilst under the influence of alcohol are just as likely as other witnesses to provide answers to questions when they are uncertain about the accuracy of information, or when they do not know the answers. This study has important implications for understanding the accuracy of victims’ accounts of actual sexual assaults.
PHOTO 6.2 Intoxicated witnesses report less than their sober counterparts, but their accuracy did not differ one day or one month later.
Source: © Lorenz Timm/Shutterstock
6.4.6 Age
So far we have concentrated on how transient factors of the witness or witnessed event operate on memory accuracy. However, more stable witness characteristics also affect how memories are encoded. Most notably, research has focused on the influence of age, with findings tending to emphasise that, of all age groups, young adults exhibit the most reliable memories. Older adults may be less accurate because of age-related declines in encoding quality, which, for example, make it more difficult to bind new information (Li, Naveh-Benjamin, & Lindenberger, 2005). This means that older witnesses are likely to remember fewer details about events, and/or recall less accurate details than young adults (Aizpurua, García-Bajos, & Migueles, 2009; Searcy, Bartlett, Memon, & Swanson, 2001).
Children are also frequently found to make less reliable witnesses, although the completeness and accuracy of children’s memories do improve in line with their age (see Chapter 7 and Lamb, Orbach, Warren, Esplin, & Hershkowitz, 2007). Children’s encoded memories might be less accurate for a number of reasons. For example, younger children lack the life experience that would enable them to employ useful scripts and schemas to help make sense of an event, and therefore to recall it accurately (Baker-Ward, Gordon, Ornstein, Larus, & Clubb, 1993).
6.5 STORAGE FACTORS
As we have just seen, a host of situational and individual differences factors influence how we attend to a criminal event as witnesses. In this section, we will consider factors that influence how events are stored in memory.
6.5.1 Post-event Information
Separating witnesses while others testify has long been relied upon by police investigators and in court as a means to preserve the independent character of witness testimony. These practices are enacted in the legal system to help deter social conformity and misremembering by eyewitnesses. Yet, people often witness crimes in the presence of others, and discuss what they saw with them (Paterson & Kemp, 2006; Skagerberg & Wright, 2008). Witnesses can pay attention to, remember, and misremember, different aspects of the crime. Through subsequent discussion, witnesses might learn information about the event that they did not actually observe themselves. If the witness incorporates this information into his or her own testimony, this can be problematic for justice.
CASE STUDY 6.1 THE OKLAHOMA BOMBIN
G AND THE MURDER OF JILL DANDO
Confabulation caused by witnesses talking to each other can have a detrimental impact on police investigations. In the Oklahoma City bombing case, in 1995, three employees were present when Timothy McVeigh, later convicted of the bombing, rented the truck used in the attack. Initially two of them thought McVeigh had been alone, but after discussing this with a third witness, they too came to believe that McVeigh had not been alone. (The FBI subsequently concluded that the “accomplice” was an innocent person who hired a different truck the next day.)
Another example comes from the Jill Dando murder case in 1999. Initially, only one of 16 witnesses identified the police suspect, Barry George, from the original line-up. The witness who made the positive identification seems to have influenced the other witnesses. After the witnesses shared a taxi home together, one of the witnesses who had previously failed to make a positive identification changed her mind, reporting that she was now 95% sure that the man she had seen was in the line-up (see Wright, Memon, Skagerberg, & Gabbert, 2009 for a discussion of these cases and possible psychological mechanisms underlying them).
The foregoing examples are an illustration of the misinformation effect, which occurs when the accuracy of a memory report is decreased by exposure to misleading information (Loftus & Hoffman, 1989). In their seminal paper on the effect, Loftus, Miller, and Burns (1978) presented participants with a slide sequence that depicted a simulated accident involving a car and a pedestrian. Afterwards, the participants answered a series of 20 questions concerning the event. For experimental participants, one of the questions suggested misleading information about the type of traffic sign at the scene of the accident (it described it as a Stop sign when in fact it was a Yield sign or vice versa). Control participants received consistent information about the type of traffic sign on the questionnaire rather than misleading information. A two-alternative forced-choice recognition test was given in the final phase of the experiment, featuring two slides, one showing the correct and the other the misleading sign. Compared to controls, experimental participants were far more likely to choose the incorrect slide, suggesting that their memory of the accident had been altered by exposure to the misleading post-event information.
Forensic Psychology Page 30
