Memory is not reliable; memory can be altered and adjusted. Memory is stored in the brain just like files stored in a cabinet, you store it, save it and then later on retrieve and sometimes even alter and return it. In doing so that changes the original data that was first stored. Over time memory fades and becomes distorted, trauma and other events in life can cause the way we store memory to become faulty. So when focusing on eyewitnesses, sometimes our memory will not relay correct information due to different cues, questioning, and trauma and so forth, which makes eyewitness even harder to rely on. Although memory is highly unreliable and hence affects the validity of eyewitness it is still applied in the criminal justice system.

Eyewitness testimonies are highly important in criminal justice systems, they are fundamental in classifying, charging, and convicting perpetrators of crimes.

Jurors are significantly inclined to believe and follow eyewitness evidence; this is quite unnerving because the criminal justice system, laboratory studies and field studies supports the conclusion that eyewitnesses regularly make errors. A vast amount of studies have found that eyewitness misidentifications are the most common cause of wrongful convictions and by using forensic DNA testing, they have found that this have accounted for more convictions of innocent persons than all other factors combined (Innocence Project, 2009; Wells, Memon, & Penrod, 2006).

Social scientists and members of the legal profession have turned their attention to whether they can rely on the ability of young children to provide accurate eyewitness testimony. They have focused on many cases relying on evidence provided by child witnesses, some of these cases are those of physical or sexual abuse. These have helped bring to the front issues relating to the accuracy and reliability of such eyewitness reports (Ceci & Bruck, 1993). “As a result there has been a related increase of scientific studies of children’s eyewitness competences, with results indicating that very young children perform significantly worse than younger adults. In line-up identification studies, young children perform at a similar level to young adults when the line-up presented contains the actual culprit but commit more false identifications when it does not” (Pozzulo & Lindsay, 1998).

There are number of reasons behind why a child will provide more incorrect eyewitness information, some of these reason are: misleading suggestions by the interviewer, false memories, sexual or some other form of abuse, the presence of someone in authority such as uniform police, attention or lies. Many things may render the allegations made by children as unreliable.

First, Researchers believe that children make the assumption that an adult would not provide the task if the target was not present, so when the children are presented with the line-up array it suggests to the children that the adult expects them to choose someone from the line-up. Therefore the children then will choose someone to avoid either disappointing the adult and at the same time avoiding to admit to uncertainty, or they may even choose someone that looks similar to the target they have seen before. They have also noted that young children may feel pressured to make identification regardless of whether the perpetrator is in fact recognised at all. According to (Davies 1996) the reason children turn towards choosing in identification line-ups is due to feeling pressured or being required to respond to questions regardless to the fact if the target is present or absent. This is also supported by the study by Pozzulo and Lindsay (1997), and they noted that due to the fact that adults are seen as an authority figure or a person to be respected and of status, children fail to realize that ‘I do not know’ is an available option as a response and so they are less inclined to respond using “I do not know” and so may be less likely to use it in comparison with adults, whereas adults may not feel that pressure of having another adult present and will not feel that they have done something wrong by admitting that they are not certain of the target.

As mentioned before it is acknowledged that there is a growing number of case studies evidently reveal that mistaken identifications made by child witnesses contribute to a failure to achieve justice. This can be seen in many examples such as, “Gene Bibbins served 15 years of a life sentence after being convicted based primarily on a mistaken identification made by a 13-year-old victim; Jimmy Ray Bromgard served 14 years of a 40-year sentence based on a mistaken identification made by an 8-year-old victim; Danny Brown served 18 years of a life sentence after being convicted based on a mistaken identification made by a 6-year-old eyewitness; and Larry Youngblood served 9 years of a 10-year sentence based on a mistaken identification made by a 10-year-old victim.” DNA evidence has afterwards proven the innocence of all these persons. So after seeing all these false identifications and wrongly accused persons it is unclear as to why the criminal justice system continues to rely on this method.

Another reason why young children’s memory recall is unreliable is that they lack some sort of understanding of what information needs to be provided in response to questions that are open-ended (Saywitz & Snyder, 1996).

In the UK a huge number of line-up identifications involving children are carried out by police officers wearing uniform. Researchers carried out a study examining the possibility that wearing a uniform contributes an authority figure cue that affects a child’s ability in some way to make accurate eyewitness identifications. They carried out a study where sixty participants aged 9-10 years old would witness a staged crime and were later on be asked to identify the criminal from a line-up. They used four conditions in order to do so, this was a two (2) (uniform: present vs. absent) Ã- two (2) (target: present vs. absent) design. They found that children in the uniform present conditions made significantly more choices than children in the uniform absent conditions. More significantly they found that in the presence of a uniform, children made more significant false identifications in target-absent line-ups. This therefore suggests that the children experienced uncertainty if the target was absent from the line-up and this may be because they were looking to some authority figure to somehow ensure them that the possibility of the burglar being present was high, but this uncertainty was not expressed when the line-up administrator wore a uniform because the that authority figure was present, leading to an increase in false identifications.

It was also found that children feel that they are helping the police, and in the eyes of children this will be deemed as something highly important and so they will not want to disappoint them in any way. They also assume that the police may have already arrested the guilty persons and need some final confirmation to be able to convict them (informational influence; cf. Steblay, 1997). Therefore in some way they have relied on the police and believe that no mistake was made. The heightened levels of uncertainty and stress in the target-absent condition could be interpreted as that they may be failing to make a proper the identification as there was nobody who they may have been assisting such as the police or an adult and at the same time found no one who matched their memory of the target. Trying to appear knowledgeable they would then have lowered their identification threshold and therefore wrongly identify someone.

Another point looked at on why children give false recollections can be looked at in terms of sexual abuse and the relation between stress and the children’s memory when asked to recall the traumatic event. First, it may be that children made false accusations from the beginning and was aware of it all along. If that was the case then this implies that they did not form false memories, unlike what many researchers would have claimed (Ceci &Bruck, 1993). Instead, the children would have been lying to please the adults or may have even been trying to seek attention. It was found that where they may have promoted lies and not false memories the children who later on, as adults, withdrew their claims. So now inferences are important because these withdrawals of their claims would mean that children’s memory flexibility was not as great as were the adult social pressures applied to the children. There were however some of the children who still held to their original charges of child sexual abuse. As well, for those who may have indeed experienced a sexual or physical traumatizing event their memory may become terribly inaccurate from the stress they had experienced, stress causes a person to see things and recall them in many incorrect ways. Even if left to calm down for sometimes or some weeks the memory can be even more lost to the correct information, in that the child can suppress the memories and recall them in ways to make them feel better or to remove the guilt and pain caused by the event.

Recent research by (Alexander et al., 2005; Widom & Morris, 1997) points out that men are more likely than women to define certain acts of child sexual abuse as not abusive and have less accurate memories for child sexual abuse experiences, they usually make the ordeal less important than it really or distorting what happened

Researchers assume that children may typically reveal sexual abuse to their mothers (Berliner & Conte, 1995), and that the person who may be sexually abusing or abused the child may be known to the child (Finkelhor, 1984), and that because the crimes and abuses may not be reported promptly (Goodman et al., 1992; Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003), it can be argued that mothers may have led the children to make a false report, or the children and/or mothers may have had concealed intentions for making the accusations, and that the child’s memory may have been either distorted, trained, or become faded with time. It was that many prosecutors have reported that such defences are often used in child sexual abuse cases (Goodman, Quas, Bulkley, & Shapiro, 1999). They argue that children do not always disclose abuse readily, may at times require some leading questions to tell accurately and completely what happened, which in doing so may bring about false reports as to what may have truly happened and also leading the children into thinking that it happened in the way that the questions were asked and they may even withdraw their claims even if they were true (Malloy, Quas & Lyon, in press; Saywitz, Goodman, Nicholas, & Moan, 1991; Summit, 1983). They have also found that it is more difficult to mislead children to report negative or events related to abuse than positive or events that is not related to abuse, but some children at times may even consent to false negative, personal experiences.

It has been well-known that the younger the children, the more likely it is for them to report false information (Bruck & Ceci, 1999; Howe, 2000). For instance, if after witnessing an event young children are given misinformation about it (e.g., “Do you remember when the doctor gave you a candy?”), their reports of the event would be more likely to include the (mis)information that the “doctor gave them candy” when, in fact, the doctor did not. Clinicians and researchers have observed that some very young children are capable of providing accurate reports of events with the use suggestive questioning, whereas some older children are not able to do so (Baxter, 1990; Bruck & Ceci, 1999; Geddie, Fradin, & Beer, 2000).

In 2004 Bruck and Melnyk published a review of the literature on individual differences in suggestibility. Out of 69 studies they looked for evidence of relationships between three categories of possible predictors: one which is demographic (socioeconomic status and sex), secondly the other is Cognitive (intelligence, language, memory, theory of mind, executive functioning, distractibility, and creativity), and thirdly, Psycho-social (social engagement, self-concept/self-efficacy, stress/emotional arousal/state anxiety, maternal attachment styles, parent-child relationship, parenting styles, temperament, and mental health). No relationship with suggestibility for some variables could be found but for others, the results were inconsistent. The predictors which showed the potential appeared to be that of the parent-child relationship, language ability, creativity, self-concept/self-efficacy, and maternal romantic attachment. Children who were vulnerable to being impressionable were more creative and had less superior language skills (Clarke-Stewart et al., 2004), inferior self-concept or self efficacy (Davis & Bottoms, 2002), less supportive relationships with parents, either fathers or mothers (Clarke-Stewart et al., 2004), and mothers who were attached in their romantic relationships in an insecure manner (Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1997; Quas et al., 1999).

Different studies were done to test and improve the accuracy of eyewitness testimony in children as well as correct memory recall. One of these tests is the Event Report Training (ERT), this is a training procedure intended to help the improvement of the memory recall of children and at the same time to reduce suggestibility. To test this efficiency of the training procedure they carried out a study. In the study fifty-eight (58) children took part in two forensically significant play events. After two weeks, the children received (ERT) or participated in control procedures, after that they were given a memory interview. The results pointed out that the Event Report Training procedure decreased suggestibility to questions related to abuse in preschoolers; their responses were greatly accurate and the difference in age was removed. (ERT) procedure did not raise the amount of information that preschoolers provided to open-ended questions. However, using the Event Report Training procedure 32% more information was reported by 7 to 8-year-olds which included a 32% enhancement in actions, without an associated raise in wrong information. (John Wiley & Sons, Ltd. 2009)

Another focus on improving accuracy is a narrative style approach. A wide-ranging study to date of 3 ½ – 9 year old children narrative development, it observed over a 1000 narratives, and used diverse techniques of analysis of narrative structure, (Peterson and McCabe 1983) acknowledged three common narrative styles in 4-year-old children. The most common being a ‘leap frog pattern’ in which children ‘jump’ from one particular feature to another, thereby excluding important aspects. An important aim of Event Report Training is to bring out intricate chronological narratives in preference to leap-frog narratives.

Researchers have developed a number of structured interviews in order to achieve accurate memory reports from children in forensic situations. One of these interviews is The cognitive interview (CI) which relies on techniques to assist in retrieval and this consists of reporting everything, temporal recall and reverse order recall, context reinstatement and recall of the event from different viewpoints of the people involved in the event (Geiselman, Fisher, MacKinnon, & Holland, 1985).

Another structured interview procedure for children is Narrative Elaboration (NE), which relies on the grouping methods to improve the narratives of children. In this, researchers first teach the children how to recall an event by organizing information into specific categories, they then instruct children about the appropriate information that is involved in complete recall of that particular category (Camparo, Wagner, & Saywitz, 2001; Saywitz & Synder, 1996; Saywitz, Snyder & Lamphear, 1996).

Afterwards, the children are given cue cards to remind them to explain each group. A third procedure developed by researchers at the National Institute of Child Health and Human Development (NICHD; Lamb, Sternberg, Esplin, Hershkowitz, & Orbach, 2000) was to improve the children’s event reports while at the same time fortifying the forensic investigator’s interviewing skills. The NICHD protocol lets the child of interview rules which they are allowed to say ‘I do not know’. This procedure builds an understanding and supplies the children with practice in describing recent events and separating precise instances of an event recurring. In addition, the procedure uses related cueing whereby after a child-generated material is given the interviewer then asks specific questions.

It was found in two investigations that 8 to 10 years old children showed a significant decline in the false response to misleading questions after being interviewed using the Cognitive interview (Memon, Holley, Wark, Bull, & Kohnken, 1996; Milne & Bull, 2003).

In general, the evidence from previous research that children are more likely to choose from a line-up is relatively strong, although the underlying mechanisms are not yet fully understood. Children may be more vulnerable to any perceived social and environmental demands to choose, or they may have a less sophisticated understanding of the purposes of an identification test and the potential consequences of their decision (Brewer, Weber, & Semmler, 2005). In any case, studies have shown that these difficulties are exacerbated in the presence of a uniform line-up administrator. As well, leading questions are problematic because it can cause the children to rely on the adults for information or may think that by the way the question was phrased they expect a particular answer. It is hence, important that such methods are limited in order to improve the accuracy of child eyewitnesses and increase confidence that the identifications made are correct.