Eyewitness testimony can be defined as the evidence given by an individual in court about people and events of a crime. Eyewitnesses testify about many things, for example, the identification of a perpetrator, which hand a gunman used, the colour of a car, or even recollections of a conversation (Wells & Olsen, 2003). Comprehensive psychological research into eyewitness testimony has been compiled all over the world and findings have mainly been united in finding eyewitness testimony vulnerable unreliability. Despite these findings, the criminal justice system still relies heavily on eyewitness testimonies for the investigating and prosecuting of crimes. Wells & Loftus (2002) state that mistaken eyewitnesses account for more convictions of innocent people than all other causes combined.
In a recent publication, Wells & Olson (2003) provided a comprehensive review of psychological research into eyewitness testimony since the 1970’s. In their publication, they provided a user-friendly framework that assists in determining the reliability/unreliability of an eyewitness testimony. The framework includes firstly investigating, the characteristics of the witness, secondly, the characteristics of the event, thirdly, the characteristics of the testimony and fourthly, the ability of testimony evaluators to discriminate between an accurate and inaccurate eyewitness testimony (Wells & Olson, 2003). In this essay, the four elements will be discussed, some in more detail than others, in conjunction with relevant research to provide a critical review of eyewitness reliability in both children and adults.
Before discussing the four elements determining eyewitness reliability, it is important to discuss some basic concepts of human memory. In reference to Siegler (1998), it seems many people understand human memory to be simular to that of a series of photographs, or a movie of our experiences. If this statement were correct, eyewitness testimony would be very reliable, as the witness would simply recount what happened. This understanding however is not the case; in fact, human memory at any age is not close to being this complete or accurate. Research has found that both adults and children often fail to remember what they saw, often have memories of events that never happened and often combine separate experiences into a single memory (Siegler, 1998).
The study of human memory understandably incorporates many components. One of these components is the three phases of memory. The first of these phases is titled the encoding phase, which refers to an individual’s acquirement of information; in the case of an eyewitness this is the phase when they witness the crime. The second phase is titled the storage phase whereby encoded information is stored in the memory for later use; research has found some eyewitnesses memory storage is very susceptible to ‘suggestibility’, meaning they are greatly influenced by experiences that occur after the original event but before the time of retrieval. The third phase is titled the retrieval stage where an individual recalls or recognises information from a memory, this would be the phase of an eyewitness giving their testimony (Galotti, 1994, Haberlandt, 1994 & Siegler, 1998).
After summarising the three phases of memory we can now examine Wells & Olsen’s (2003) first component of determining the reliability/unreliability of an eyewitness testimony, the characteristics of the witness. Psychological research into eyewitness testimony focusing on the characteristics of the witness has explored areas such as the gender, age, intelligence, race and personality of the eyewitness. We will investigate in more detail recent research of the eyewitness characteristics of age and race.
The hypothesis that an eyewitnesses age may be linked to how they will perform and the accuracy at which they will perform in testifying and whether some age groups are more vulnerable to suggestibility has been investigated by many psychologists. This research is of great importance, as children are increasingly testifying in courts, particularly in relation to family violence and sexual abuse (Leippe, Manion & Romanczyk, 1991; Roebers, Bjorklund, Schneider & Cassel, 2002). Wells & Olsen (2003) state that very young children and the elderly have been found to perform significantly worse than younger adults in eyewitness identification tasks.
In a study examining child eyewitnesses, Davies, Tarrant & Flin (1989) (cited in Pozzulo & Warren, 2003) found that younger children of 6 – 7 years of age, recalled fewer descriptions about a stranger they had observed than older children of 10 – 11 years of age. Leippe, Manion & Romanczyk (1991) found simular results in their study, with 5 – 6 year old children giving less complete free recall responses, free recall meaning the task of telling someone everything you can remember about a person or event, and more inaccurate responses to objective questions about the stranger they encountered and what he did than 9 – 10 year old children or college students.
Cognitive psychologists have posed many reasons why research has found such a marked difference in recall performance between young children and young adults, some believe the differences may lie in the development of human memory, in particular the three phases of memory. Brainerd, Reyna & Kingma (1990) (as cited in Siegler, 1998) believe the encoding phase of memory provides some answers. Brainerd, et al., (1990) state that when people encode information, they form two types of representations. The first type of representation is titled ‘verbatim’, meaning the literal details of the situation, for example, the exact words spoken, the expressions on people’s faces and the colour of the walls. The second type of representation is titled ‘gist’, meaning the essence of the situation, for example, who did what to whom. It has been found that individuals encode both types of representations into their memory, however representations of gist last much longer than representations of verbatim and this is where differences between adult and child memory becomes apparent. Brainerd, et al., (1990) believe part of the reason that young children remember somewhat less well than older individuals is that their encoding phase places a greater emphasis on verbatim information than gist information. Since there is a tendency for individuals to forget verbatim information more quickly than gist information, the young children’s emphasis on verbatim information therefore leads to more forgetting.
In addition to the comprehensive research on age and eyewitness testimony, some psychological researchers have also been interested in racial differences and the accuracy of eyewitness memory. Pezdek, Blandon-Gitlin & Moore (2003) found that children from 5 years of age through to adults had more accuracy in recognising faces from their own race than those faces from other races. From this research there was also found to be age consistency in the cross-race effect, meaning both children and adults had the same ability level in recognising faces from their own race rather than other races. Other previous studies have not found this result, in a study by Pozzulo & Lindsay (1998) (as cited in Pezdek, et al., 2003); it was found that face recognition memory improves with age. In providing support for their controversial findings, Pezdek, et al., (2003) suggested the cognitive processes underlying face recognition memory were simular for participants at each age level and the reason previous studies had found face recognition memory showed signs of improving with age was because older participants could simply perform these cognitive processes better. Valentine (1991) and Valentine & Endo (1992) (as cited in Pezdek, et al., 2003) provide a model titled ‘the multidimensional space framework’. The model proposes how faces might be represented in memory. According to this model, when we view a face, it is represented in a hypothetical space that retains each face on the basis of various dimensions representing features or sets of features. These dimensions are developed from individual’s prior experience with faces, therefore it makes sense that individual’s would have more dimensions in their memory storage representing features of individuals faces from their own race. This model therefore explains why people can more accurately recognise faces from their own race than those from other races; our memory is more familiar with different individuals facial features from our own race. There have been many instances in eyewitness testimonies where the witness has had to identify a suspect who is not of their own race, it is clear extra caution should be taken in these instances.
A brief overview of the second component of Wells & Olsen’s (2003) framework for determining the reliability/unreliability of an eyewitness testimony, the characteristics of the event will now follow. There are a variety of factors of an event that could affect the ability of an eyewitness to identify the perpetrator of a crime at a later time. Some of the psychological research that has been compiled has investigated factors such as the amount of time the perpetrator is in view, the lighting conditions, whether the perpetrator wears a disguise, the distinctiveness of the perpetrators appearance, the presence or absence of a weapon, and the timing of knowledge that one is witnessing a crime. (Wells & Olsen, 2003) It is clear that all these factors would certainly play an important role in the ability of an eyewitness to accurately identify a perpetrator of a crime, and research into these factors has provided fascinating results.
The third component of Wells & Olsen’s (2003) framework is the characteristics of the testimony. Considerable research has been compiled regarding what characteristics indicate an eyewitness has made an accurate or false testimony of events or identification of a perpetrator. Some of these characteristics include the confidence/accuracy relationship, frequency of suggestibility and an eyewitnesses timing in identifying a perpetrator. We will investigate recent research about suggestibility.
Suggestibility refers to the rate at which an individual’s stored memories can be altered by presenting misleading questions and procedures (Galotti, 1994). In general, research into suggestibility has found that the younger the child, the more susceptible he or she is to suggestion, with preschool children being especially likely to comply with the suggestion of an adult (Roebers, Bjorklund, Schneider & Cassel, 2002). Comprehensive research by Leichtman & Ceci (1995) examined, amongst many other things, the role that ‘source misattributions’ in the suggestibility of children. Source misattributions are found when individuals are asked either to perform some act or else to merely imagine performing it. Many research findings have unilaterally found younger children are more likely to misattribute performing an act that they had only imagined performing. In Leichtman & Ceci (1995) research it was concluded that younger children certainly showed an increased difficulty than older children in determining whether their mental images were the result of past experiences in the world or products of their past imaginings. Furthermore, if a child was given the opportunity to give an eyewitness account free of suggestible questioning by an adult, factual reporting by the child was said to be highly likely. Wells & Loftus (2002) discuss a case in which an eyewitness was subjected to an interview of suggestive questioning by detectives, that later convicted an innocent man. There were many reasons why the questioning was regarded as suggestive, most concerning was the way the detectives tried to interpret the eyewitnesses behaviour for her, for example ‘you went to number three (photograph #3) just like thataˆ¦ you became flushed’ (Wells & Loftus, 2002, Pg 152). It is clear that suggestibility can also be powerful in adult eyewitness testimony, as witnesses and victims of a crime long to convict the perpetrator, shaping of stored memory by suggestible procedures can often distort these individual’s original memories of the true perpetrator.
The fourth and final component of Wells & Olsen’s (2003) framework for determining the reliability/unreliability of an eyewitness testimony is the ability of testimony evaluators to discriminate between an accurate and inaccurate eyewitness testimony. By this, Wells & Olsen (2003) refer to the ability observers, e.g. jury members have in discriminating between accurate and inaccurate eyewitness testimony. There has been a small amount of research into this area of eyewitness testimony, researchers providing several methods to assess the adequacy of people’s judgements about eyewitness testimony accuracy. Some examples of these methods include surveys, prediction studies and cross examination of eyewitnesses. Research using these methods has shown that subject-jurors in an experimental setting have little or no ability to make discriminations about eyewitness accuracy levels.
To conclude, it is clear from this discussion of recent cognitive psychological research why eyewitness testimony is in fact often so unreliable, particularly in child eyewitness accounts. In understanding why eyewitness testimony is so unreliable, it has been helpful to discuss the four characteristics of Wells & Olson’s (2003) framework, the characteristics of the witness, the characteristics of the event, the characteristics of the testimony and observers judgements of accuracy. The most predominant findings from the research discussed have firstly been evidence of children’s eyewitness memory being less accurate than adults due to developmental stages in memory. Secondly, evidence has found children and adults have a better ability of recognising the faces of individuals from their own race and thirdly, evidence has found that suggestibility in the questioning procedures of eyewitnesses can have a dramatic effect on the original stored memory. Considering all these factors and so many others that could not be discussed, it clearly more apparent why innocent individuals are so often falsely convicted by the use of eyewitness testimony in the criminal justice system.