**3. Vulnerability to error and influence**

Beyond memory issues, additional factors can influence the accuracy of children's memory reports – and their vulnerability to influence. These include children's intelligence, mental health, source monitoring ability, suggestibility and the use of suggestive interview aids.

**Intelligence and mental health**. For child witnesses, intelligence is most relevant to memory accuracy and resistance to suggestion. That is, individuals with higher IQ scores tend to have better memory skills and to provide more detailed recall when reporting an event (Gudjonsson & Henry, 2003; Zhu et al., 2010). Although a legal professional might not need to take into account the intelligence of typically developing (TD) children, it is important to be aware that children who suffer from learning and other intellectual disabilities (ID) may be more vulnerable in forensic contexts (Henry & Gudjonsson, 2007). Indeed, the rate at which children's memory performance improves across age may be slower for ID children than for TD children, with mental age a better predictor of eyewitness memory performance than IQ measures (Henry & Gudjonsson, 2007). Thus, children with ID are at risk of being less accurate and more suggestible in their event reporting than are TD children.

90 Current Topics in Children's Learning and Cognition

Cicchetti, 2004; Zajac & Hayne; 2003).

children who face recurrent stressful experiences, such as children who are victims of abuse. Maltreated children may come into contact with the legal system frequently, especially if the abuse has been substantiated and if they are required to testify. Maltreated children may differ from children who only experience traumatic or stressful events once, in that the constant exposure to stress may affect how maltreated children encode and retrieve information. Some research has yielded evidence for developmental delays in maltreated children, especially those who have been physically abused or neglected (see Veltman & Browne, 2001 for a review). Maltreated children have been shown to perform at levels one or two years behind nonabused peers in areas such as intelligence, language, short-term memory, and executive function (Alessandri 1991; Carrick, Quas, & Lyon, 2010; Cicchetti, Rogosch, Maughan, Toth, & Bruce, 2003; Trickett & McBride-Chang, 1995). These delays can be problematic when obtaining a memory report from a maltreated child because they can affect a child's ability to adequately understand and respond to questions (e.g., Eigsti &

On the other hand, a child's constant exposure to traumatic events might help his or her memory. The constant stress may lead a child to adopt processing strategies that make him or her hyper-vigilant to stressful stimuli, thereby encoding that information better than nonstressful stimuli (McNally, Metzger, Lasko, Clancy, & Pitman, 1998). A maltreated child may also have extensive knowledge of trauma-related information, which could support better memory (Goodman, Quas, & Ogle, 2010). However, there is evidence to suggest that this hyper-vigilance is generally associated with emotion-regulating strategies (McNally, Kaspi,

Riemann, & Zeitlin, 1990; Williams & Broadbent, 1986), rather than cognitive factors.

does it seem to alter the predictors of memory accuracy (Eisen et al., 2002, 2007).

Beyond memory issues, additional factors can influence the accuracy of children's memory reports – and their vulnerability to influence. These include children's intelligence, mental health, source monitoring ability, suggestibility and the use of suggestive interview aids.

**Intelligence and mental health**. For child witnesses, intelligence is most relevant to memory accuracy and resistance to suggestion. That is, individuals with higher IQ scores tend to

**3. Vulnerability to error and influence** 

Overall, despite the findings of developmental delays among maltreated children, research has shown that maltreated children can be as accurate in their memory reports as their nonabused counterparts (Goodman et al., 2010; Howe, 1997; Howe, Toth, & Cicchetti, 2011). For example, Eisen and others (2002) examined 189 3- to 17-year-olds who were allegedly maltreated. These children were questioned about an anogenital examination they received as part of their child maltreatment investigations. Results mimicked those found by studies examining nonabused children: younger children made more errors and reported fewer details than older children. Importantly, the children with prior histories of abuse performed on par with children with no prior history of abuse. Thus, child maltreatment does not necessarily deprive children of their ability to provide accurate event reports, nor These effects may be exacerbated among children who are in the justice system as defendants. ID is overrepresented among juvenile delinquents (Najdowski, Bottoms, & Vargas, 2009), as is mental illness (Redlich, 2007), both of which increase adolescent vulnerability to interrogative pressure, which is already heightened among youths (Redlich, 2007; Richardson, Gudjonsson, & Kelly, 1995). Indeed, research demonstrates that children and adolescents (particularly ID and mentally ill juveniles) are at greater risk for falsely confessing transgressions that they did not commit (Najdowski et al., 2009; Redlich, 2007). Given these findings, it is vital that intellectual ability and mental health be taken into account when considering child and adolescent witness and defendant statements.

**Source monitoring.** Children's memory reports can also become inaccurate due to faulty source monitoring, which is the process of determining the origin (or source) of their knowledge, memories and/or beliefs (Johnson, Hashtroudi, & Lindsay, 1993). Child witnesses, like any other witnesses, might be asked to provide specific details about criminal events, such as when a crime happened, who committed it, where it happened, and distinguish whether it was a perpetrator's actions or someone else's that they are recalling. Unfortunately, pre- and post-event information from outside sources can influence one's memory, thus making it difficult to distinguish between real event details and suggested details (Poole & Lindsay, 2002; Roberts & Blades, 1999; Thierry, Spence, & Memon, 2001). The source of one's misinformation can vary from television, to word-of-mouth, to intentional suggestion, to name a few. But regardless of the source, the concern is the same – that a child exposed to incorrect information following an event will fail to identify the true source of the novel information and provide inaccurate testimony as a result. Indeed, there is clear developmental growth in the ability to recall source information, identified at the neural level (Sprondel, Kipp, & Mecklinger, 2011). Hence, researchers have begun to explore what factors might influence source monitoring.

There are several factors that can influence the likelihood of committing source monitoring errors for children. These include the similarity of events, delays in recall, suggestive interviewing, and agent identity (i.e., whether the event was about oneself, a friend, or an unfamiliar person; Roberts, 2002). When pre- or post-event information is similar to the target information, it is more likely that children will make source monitoring errors (Johnson et al., 1993; Lindsay, Johnson, & Kwon, 1991). Indeed, although children with stronger source monitoring abilities seem better able to resist post-event suggestions,

particularly when encouraged to identify the source of the suggestions, they are less able to do so when those suggestions are consistent with their event schema (Roberts & Powell, 2006). When information that is similar in nature is encoded, there are few cognitive cues to help to distinguish the sources (Day, Howie, & Markham, 1998; Roberts & Blades, 1999). This can be important in instances where children experience multiple similar incidents, such as in the case of prolonged child abuse. Unfortunately, in forensic contexts such as court, children may be asked to distinguish between these multiple incidents or identify multiple perpetrators.

Cognition and the Child Witness 93

questioning by an interviewer. For example, a question such as, "Was the man's jacket red or blue?" is suggestive because it implies that the man was wearing a jacket (he may or may not have been). Children may have difficulty accurately answering suggestive questions because they may not have the cognitive or social skills to identify that they do not know the

Studies to date indicate that individual difference variables may be more likely to influence the degree to which people are suggestible than situational factors (Gudjonsson, 1994). Indeed there are several traits that might make an individual more susceptible to suggestive questioning (Bain, Baxter, & Ballantyne, 2007; Singh & Gudjonsson, 1992), such as age. Both adults and children are susceptible to suggestive questioning, but younger children tend to be especially vulnerable to suggestive interview techniques (Ceci & Bruck, 1993, 1995; Chae, Goodman, Eisen, & Qin, 2011; Eisen et al., 2002). These age differences can usually be explained by differences in cognitive development between older and younger children. In fact, when age is controlled, children with better cognitive functioning are overall less suggestible and produce fewer memory errors due to suggestive questioning (Chae et al., 2011; Karpinski & Scullin, 2009). Moreover, as noted above, children with intellectual disabilities are highly susceptible to suggestive questioning (Henry & Gudjonsson, 1999,

Findings for other potential cognitive predictors of suggestibility have been mixed (see Bruck & Melnyk, 2004 for a review). However, research has shown that children with advanced language skills, better memory test performance, (Clark-Stewart et al., 2004; Danielsdottir, Sigurgeirsdottir, Einarsdottir, & Haraldsson, 1993; Henry & Gudjonsson, 2003), better event accuracy (Marche, 1999; Marche & Howe, 1995; Pezdek & Roe, 1995), and/or more creativity (Brown, 1999; Clarke-Stewart et al., 2004) are often less suggestible

Concerns about children's suggestibility are common in legal cases, especially in instances where the child's memory report is key evidence. In fact, in child sexual abuse cases, the child's accusation may be the only evidence available to the court. It is thus important to consider interviewing techniques and the potential for suggestion when evaluating child

**Interview aids.** A special case of suggestibility might be said to arise through the use of some interview aids, such as anatomically detailed dolls. First used clinically as a means of communicating nonverbally about children's emotional issues, forensic interviewers began to use anatomically detailed dolls to assist children in reporting difficult or embarrassing events (i.e., sexual abuse; Poole, Bruck, & Pipe, 2011). Unfortunately, the dolls seemed to be inherently suggestive, leading some children to demonstrate genital or anal cavity insertions on the dolls that did not occur (Bruck, Ceci, & Francoeur, 2000). Thus, researchers and policy groups have concluded that children's interactions with such dolls are not diagnostic of abuse, they should not be used to elicit abuse reports, their use is particularly problematic for children under age 5, and they do not appear to facilitate or enhance children's reporting

even when used after verbal reports have been provided (Poole et al., 2011).

answer and do not have to choose an answer.

2003, 2004, 2007).

than other children.

witnesses' memory reports.

On the other hand, children tend to make fewer source monitoring errors when the target information regards the self rather than others (Baker-Ward, Hess, & Flannagan, 1990; Roberts & Blades, 1998). In other words, they are better able to accurately report on events that happened to themselves than on events they witnessed other people experiencing. For example, Roberts and Blades (1998) found that 6- and 9-year-olds were more accurate and made fewer source monitoring errors answering questions about activities they had performed versus those they observed, while 4-year-olds showed the reverse relation. Further, children remember better actions performed by a familiar person compared to an unfamiliar person (Baker-Ward et al., 1990). Some have posited that this is due to the cognitive processes that are elicited when encoding familiar (self and peer) information (Baker-Ward et al., 1990; Foley et al., 1989). Specifically, children in these situations are using cognitive processing that requires them to use more cognitive effort—they are anticipating what must be done, anticipating the action's consequence, and reflecting on the cognitive processes that were required to perform that action (Ratner, Foley, & Gimpert, 2000). Children can then use records of these cognitive processes as cues to discriminate between sources (Roberts, 2002).

All of these factors are relevant when considering child witnesses and techniques for interviewing them. For example, misleading or suggestive interview questions that are script-consistent and plausible are more readily accepted by child witnesses and are more likely to elicit source monitoring errors (Pezdek, Fingers, & Hodge, 1997). Additionally, the nature of the inquiry can impact the accuracy of the report. For instance, Roberts and Blades (1995) found that younger children were more accurate in their source identifications when asked to report information nonverbally compared to verbally. Moreover, children committed fewer source monitoring errors during free recall than when answering specific questions (Roberts & Blades, 1998, 1999), reinforcing the importance of free recall when interviewing child witnesses. To date, however, explicit training in source monitoring appears to be beneficial for older (i.e., 7- to 8-year-old), but not for younger (3- to 6-year-old) child witnesses (Poole & Lindsay, 2002).

**Suggestibility.** Another potential source of children's errors in forensic contexts is via suggestibility. Suggestibility has been defined as the ease with which a person is persuaded and influenced, and the degree to which one accepts and changes one's own judgments, opinions, or patterns of behavior without critical response (Drukteinis, 2001). In a forensic context, this is usually exhibited by changes in memory reports due to suggestive questioning by an interviewer. For example, a question such as, "Was the man's jacket red or blue?" is suggestive because it implies that the man was wearing a jacket (he may or may not have been). Children may have difficulty accurately answering suggestive questions because they may not have the cognitive or social skills to identify that they do not know the answer and do not have to choose an answer.

92 Current Topics in Children's Learning and Cognition

multiple perpetrators.

sources (Roberts, 2002).

child witnesses (Poole & Lindsay, 2002).

particularly when encouraged to identify the source of the suggestions, they are less able to do so when those suggestions are consistent with their event schema (Roberts & Powell, 2006). When information that is similar in nature is encoded, there are few cognitive cues to help to distinguish the sources (Day, Howie, & Markham, 1998; Roberts & Blades, 1999). This can be important in instances where children experience multiple similar incidents, such as in the case of prolonged child abuse. Unfortunately, in forensic contexts such as court, children may be asked to distinguish between these multiple incidents or identify

On the other hand, children tend to make fewer source monitoring errors when the target information regards the self rather than others (Baker-Ward, Hess, & Flannagan, 1990; Roberts & Blades, 1998). In other words, they are better able to accurately report on events that happened to themselves than on events they witnessed other people experiencing. For example, Roberts and Blades (1998) found that 6- and 9-year-olds were more accurate and made fewer source monitoring errors answering questions about activities they had performed versus those they observed, while 4-year-olds showed the reverse relation. Further, children remember better actions performed by a familiar person compared to an unfamiliar person (Baker-Ward et al., 1990). Some have posited that this is due to the cognitive processes that are elicited when encoding familiar (self and peer) information (Baker-Ward et al., 1990; Foley et al., 1989). Specifically, children in these situations are using cognitive processing that requires them to use more cognitive effort—they are anticipating what must be done, anticipating the action's consequence, and reflecting on the cognitive processes that were required to perform that action (Ratner, Foley, & Gimpert, 2000). Children can then use records of these cognitive processes as cues to discriminate between

All of these factors are relevant when considering child witnesses and techniques for interviewing them. For example, misleading or suggestive interview questions that are script-consistent and plausible are more readily accepted by child witnesses and are more likely to elicit source monitoring errors (Pezdek, Fingers, & Hodge, 1997). Additionally, the nature of the inquiry can impact the accuracy of the report. For instance, Roberts and Blades (1995) found that younger children were more accurate in their source identifications when asked to report information nonverbally compared to verbally. Moreover, children committed fewer source monitoring errors during free recall than when answering specific questions (Roberts & Blades, 1998, 1999), reinforcing the importance of free recall when interviewing child witnesses. To date, however, explicit training in source monitoring appears to be beneficial for older (i.e., 7- to 8-year-old), but not for younger (3- to 6-year-old)

**Suggestibility.** Another potential source of children's errors in forensic contexts is via suggestibility. Suggestibility has been defined as the ease with which a person is persuaded and influenced, and the degree to which one accepts and changes one's own judgments, opinions, or patterns of behavior without critical response (Drukteinis, 2001). In a forensic context, this is usually exhibited by changes in memory reports due to suggestive Studies to date indicate that individual difference variables may be more likely to influence the degree to which people are suggestible than situational factors (Gudjonsson, 1994). Indeed there are several traits that might make an individual more susceptible to suggestive questioning (Bain, Baxter, & Ballantyne, 2007; Singh & Gudjonsson, 1992), such as age. Both adults and children are susceptible to suggestive questioning, but younger children tend to be especially vulnerable to suggestive interview techniques (Ceci & Bruck, 1993, 1995; Chae, Goodman, Eisen, & Qin, 2011; Eisen et al., 2002). These age differences can usually be explained by differences in cognitive development between older and younger children. In fact, when age is controlled, children with better cognitive functioning are overall less suggestible and produce fewer memory errors due to suggestive questioning (Chae et al., 2011; Karpinski & Scullin, 2009). Moreover, as noted above, children with intellectual disabilities are highly susceptible to suggestive questioning (Henry & Gudjonsson, 1999, 2003, 2004, 2007).

Findings for other potential cognitive predictors of suggestibility have been mixed (see Bruck & Melnyk, 2004 for a review). However, research has shown that children with advanced language skills, better memory test performance, (Clark-Stewart et al., 2004; Danielsdottir, Sigurgeirsdottir, Einarsdottir, & Haraldsson, 1993; Henry & Gudjonsson, 2003), better event accuracy (Marche, 1999; Marche & Howe, 1995; Pezdek & Roe, 1995), and/or more creativity (Brown, 1999; Clarke-Stewart et al., 2004) are often less suggestible than other children.

Concerns about children's suggestibility are common in legal cases, especially in instances where the child's memory report is key evidence. In fact, in child sexual abuse cases, the child's accusation may be the only evidence available to the court. It is thus important to consider interviewing techniques and the potential for suggestion when evaluating child witnesses' memory reports.

**Interview aids.** A special case of suggestibility might be said to arise through the use of some interview aids, such as anatomically detailed dolls. First used clinically as a means of communicating nonverbally about children's emotional issues, forensic interviewers began to use anatomically detailed dolls to assist children in reporting difficult or embarrassing events (i.e., sexual abuse; Poole, Bruck, & Pipe, 2011). Unfortunately, the dolls seemed to be inherently suggestive, leading some children to demonstrate genital or anal cavity insertions on the dolls that did not occur (Bruck, Ceci, & Francoeur, 2000). Thus, researchers and policy groups have concluded that children's interactions with such dolls are not diagnostic of abuse, they should not be used to elicit abuse reports, their use is particularly problematic for children under age 5, and they do not appear to facilitate or enhance children's reporting even when used after verbal reports have been provided (Poole et al., 2011).

Part of the reason for the dolls' ineffectiveness with very young children could be children's immature representational abilities (DeLoache, 2000). That is, young children may have difficulty perceiving the doll as a representation or symbol of themselves and acting on it as such. They might also have difficulty ignoring the dolls' salient, play-like qualities and unique features to use it as a symbol instead (DeLoache, 2000; Poole et al., 2011).

Cognition and the Child Witness 95

participating accurately and effectively as witnesses. They often partake in family court decision-making, provide evidence in eyewitness identification cases, and testify against their assailants in child abuse cases. There is a great deal still to be learned about important cognitive factors, such as whether there are means of improving younger children's source monitoring skills, how to improve children's eyewitness identification accuracy when the perpetrator is not present in a lineup, and how to help children resist misleading questions and suggestions. More needs to be learned about the impact of emotion on children's eyewitness memory and recall as well. Perhaps the growing research on the neural correlates of children's memory will further our understanding and allow for greater forensic sensitivity to child witness strengths and vulnerabilities. These and other results, it is hoped, will contribute to more effective interviewing, testimony and treatment of child

Aldridge, J., Lamb, M. E., Sternberg, K. J., Orbach, Y., Esplin, P. W., & Bowler, L. (2004). Using a human figure drawing to elicit information from alleged victims of child sexual

Alessandri, S. M. (1991). Play and social behavior in maltreated preschoolers. *Development* 

Alexander, K., Goodman, G. S., Schaaf, J. M., Edelstein, R. S., Quas, J. A., & Shaver, P. R. (2002). The role of attachment and cognitive inhibition in children's memory and suggestibility for a stressful event. *Journal of Experimental Child Psychology*, *83*(4), 262-

Bahrick, L. E., Parker, J., Fivush, R., & Levitt, M. (1998). The effects of stress on young children's memory for a natural disaster. *Journal of Experimental Psychology: Applied, 4*(4),

Bain, S. A., Baxter, J. S., & Ballantyne, K. (2007). Self-monitoring style and levels of

Baker-Ward, L., Hess, T. M., & Flannagan, D. A. (1990). The effects of involvement on children's memory for events. *Cognitive Development*, *5*(1), 55-69. doi:10.1016/0885-

Bigler, R. S., & Liben, L. S. (1993). A cognitive-developmental approach to racial stereotyping and reconstructive memory in Euro-American children. *Child Development,* 

Bjorklund, D. F., Cassel, W. S., Bjorklund, B. R., Brown, R. D., Park, C. L., Ernst, K., et al. (2000). Social demand characteristics in children's and adults' eyewitness memory and suggestibility: The effect of different interviewers on free recall and recognition. *Applied* 

interrogative suggestibility. *Personality and Individual Differences, 42*, 623-630.

witnesses.

**Author details** 

**5. References** 

308-331.

2014(90)90012-I

Daisy A. Segovia and Angela M. Crossman *John Jay College and the Graduate Center, Cuny, USA* 

*and Psychopathology, 3*, 191–205.

290. doi:10.1016/S0022-0965(02)00149-2

*64,* 1507-1518. doi: 10.2307/1131549

abuse. *Consulting and Clinical Psychology, 72,* 304-316.

More recent efforts have focused on the use of body diagrams – pictures – to assist children in reporting abuse allegations. Although pictures seem less challenging for children to use as symbols, concern remains about their suggestiveness (Poole et al., 2011). Specifically, as with dolls, use of the diagrams seems to increase false reports of bodily touch, even among school-aged children (Poole & Dickinson, 2011). Further, their use seems to lead interviewers away from recommended, open-ended questioning styles toward more leading, specific questions (Aldridge et al., 2004; Poole et al., 2011). Finally, their use does not seem to enhance children's disclosures or lead to greater elaboration of their reports (Bruck, 2009). Thus, these symbolic interview aids seem to be suggestive, particularly for young children who might not be capable of using them symbolically, and their forensic use is discouraged (Poole et al., 2011).
