Keywords

In this chapter, we discuss basic principles of interviewing alleged child victims of sexual abuse . We have divided our review into three distinct yet complementary (and at times, somewhat overlapping) sections: the interviewee, the interviewer, and the interview . Issues for forensic interview ers to consider when questioning child victims are outlined. In this paper, we assume the interview goal is to elicit accurate eyewitness accounts. We acknowledge that child forensic interviewing is a large and growing field of study; therefore, the topics included in each section are not exhaustive of all factors examined in the extant literature. Moreover, despite the vast research base, there are still many factors that can affect actual forensic investigations that remain largely unexplored. For present purposes, we consider who is being interviewed, who is conducting the interview, and how the interview is conducted. Before discussing our selected factors, we briefly summarize evidence regarding children’s memory for stressful events as these are inherently the types of events interviewers ask children to recount.

Children’s Memory for Stressful Events

The existing literature indicates that children, even as young as age 2 or 3 years of age, form memories of highly arousing, personally significant, stressful events (e.g., Fivush, 2002; Peterson, 2011; Tustin & Hayne, 2010), and can often accurately recall such memories later, sometimes even after long delays (Jack, Simcock, & Hayne, 2012; Quas et al., 1999). Compared to young children, older ones typically provide more detailed accounts and answer questions about such events with greater accuracy (e.g., Eisen, Goodman, Qin, Davis, & Crayton, 2007; Jack et al., 2012; Morris & Baker-Ward, 2007; Peterson, 2011; Usher & Neisser, 1993). The high arousal that even relatively young children experience during stressful events may help them attend to details that are central to the event, thus resulting in particularly accurate reports of central information, potentially at the cost of poorer memory for peripheral information (Chae, Goodman, Eisen, & Qin, 2011; Christianson, 1992; Eisen et al., 2007; Quas et al., 1999). However, children also may attend to and recall idiosyncratic or peripheral information, likely because it was personally significant, represents “the worst moment,” or was otherwise distinctive for the children’s young minds (Goodman, Rudy, Bottoms, & Aman, 1990; Howe, 2006; Pynoos & Eth, 1984). Errors in memory for stressful events can also occur in children and adults (e.g., Eisen et al., 2007).

The legal system often requires “particularization” of memory , that is, the child being able to describe each assault. This can be a difficult task for anyone, but experiencing multiple or repeated traumatic events may cause more blending of memories in young children’s (e.g., 4-year-olds) reports than it does for older children (e.g., 7-year-olds) and adults, making it more difficult for young children to recall a specific event (Howe, Courage, & Peterson, 1995; Terr, 1988; but see Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994). However, interview strategies (e.g., “What happened the last time?”) can benefit children in “sourcing” details to specific events, helping children report separate events accurately (Brubacher, Powell, & Roberts, 2014; Drohan-Jennings, Roberts, & Powell, 2010; Lyon, 2005). Some researchers are finding that asking children what “usually happens” is helpful when later probing specific instances (Brubacher et al., 2014).

Although even very young children (2-year-olds) are capable of remembering core information from personally significant stressful events, it is likely that over time, “infantile amnesia” will increasingly block access to recall of experiences, including stressful ones, that occurred in infancy or toddlerhood (Usher & Neisser, 1993). Moreover, there is little to no evidence that preschoolers, older children, or adults can verbally recall events from their first year of life regardless of the traumatic nature of the event (but see Myers, Clifton, & Clarkson, 1987), although remnants of the early experiences may show in their behavior (Terr, 1988). Furthermore, a number of socioemotional factors can affect children’s motivation, ability, and willingness to share traumatic information. These socioemotional factors in child sexual abuse cases include shame and embarrassment; fear of retribution, getting in trouble, and implicating a loved one or even a stranger; the distress produced by recalling and thus “reliving” a traumatic event; and intimidation by or lack of rapport with an interviewer (e.g., Bidrose & Goodman, 2000; Cederborg, Lamb, & Laurell, 2007; Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003; Pynoos & Eth, 1984; Saywitz, Goodman, Nicholas, & Moan, 1991).

Knowledge about stress and memory in children, combined with age-appropriate, child -friendly tactics, can be used to help children provide their eyewitness accounts accurately. In the next sections, additional factors are discussed that may also affect the quality of children’s memory reports.

The Interviewee

Researchers have examined children’s eyewitness abilities in laboratory and forensic settings for several decades (Goodman, 2006; Goodman & Melinder, 2007; Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007; Lamb, Orbach, Warren, Esplin, & Hershkowitz, 2007; Paz-Alonso, Ogle, & Goodman, 2013). This has increased our understanding of children’s developing capabilities to be reliable and accurate eyewitnesses and has uncovered important individual differences for forensic interview ers to consider when questioning children about child sexual abuse . Here we discuss findings regarding children’s age, gender, psychopathology, and maltreatment histories.

Age and Gender

With age, children’s memory abilities improve. Likely due to their cognitive growth in knowledge base, memory organization, retrieval strategies, and language acquisition (Bjorklund, 2011; Howe, 2011), older children tend to remember more information and make fewer errors compared to younger children about stressful events and nonstressful events (e.g., Eisen et al., 2007; Eisen, Qin, Goodman, & Davis, 2002; Jack, Leov, & Zajac, 2014; Tustin & Hayne, 2010). Older compared to younger children have better source memory (e.g., remembering whether they witnessed something or heard someone talk about it) with sometimes vast improvements occurring between ages 3 and 8 years (Brubacher et al., 2014; Foley, 2014). The general trend is that as children age, they become better eyewitnesses, at least in terms of their memory abilities, by providing greater amounts of correct information and succumbing less often to false suggestions (Malloy, Johnson, & Goodman, 2013; McWilliams, Narr, Goodman, Ruiz, & Mendoza, 2013; but see Brainerd, Reyna, & Ceci, 2008, regarding reverse developmental trends in memory).

However, remembering and disclosing details about child sexual abuse are not identical processes. Younger children may fail to disclose child sexual abuse for several reasons, such as because they do not understand the importance or point of a forensic interview , because they think the interviewer already knows what happened, or due to short attention span or intimidation. Moreover, they are less likely than older children to consistently disclose sexual abuse (Keary & Fitzpatrick, 1994). Older children may be particularly at risk of reluctance to disclose due to fearing consequences of disclosure ; they may also recant sexual abuse details, particularly in intrafamilial abuse cases (e.g., Malloy, Brubacher, & Lamb, 2011; Malloy, Lyon, & Quas, 2007). Regardless of age, children who have been sexually victimized at times even deny that the abuse occurred (Pipe, Lamb, Orbach, & Cederborg, 2007). Older children have greater cognitive ability to remember and recount events accurately, but they may struggle with the social components of disclosing what occurred, whereas younger children may experience both cognitive and social difficulties during forensic interviews . Interviewers may avoid age-related problems by reducing cognitive demands (e.g., avoiding ambiguous or complex questioning, offering reminders to children to correct interviewers if they make mistakes) and providing social support and sufficient rapport building.

Regarding child gender, it remains unclear if memory accuracy for traumatic events differs between boys and girls, and if gender differences exist in willingness to disclosure and discuss past emotional events (e.g., Areh, 2011; Grysman & Hudson, 2013). A small subset of boys (younger than 5) have been observed in research studies to confabulate reports of touch (e.g., Poole, Dickinson, Brubacher, Liberty, & Kaake, 2014). Girls tend to provide more complete accounts than boys in their descriptions of past emotional events (e.g., Buckner & Fivush, 1998; Fivush & Zaman, 2014). However, gender may serve as only a proxy indicator; that is, other variables, such as verbal ability and cognitive control, likely serve as better predictors of memory performance and mitigate gender differences (e.g., Grysman & Hudson, 2013; Poole et al., 2014). Although gender difference in such abilities may exist at young ages (e.g., Bornstein, Han, & Haynes, 2004; Wallentin, 2009), these differences tend to disappear with age such that gender differences in the ability to recall traumatic details and answer questions accurately do not typically exist later on (Eisen et al., 2007; Quas et al., 1999). However, boys report fearing more consequences (e.g., Malloy et al., 2011), may be more reluctant to disclose sexual abuse details (e.g., Malloy et al., 2007), and provide less complete reports (e.g., Eisen et al., 2002; Grysman & Hudson, 2013) than do girls.

Taken together, findings indicate that boys and girls can report their (traumatic and nontraumatic) pasts at comparable rates of accuracy when additional factors are considered (e.g., verbal ability, cognitive control). Yet in the absence of considering additional factors, when gender differences are found, they typically show young girls to report past events with greater accuracy than young boys (Poole et al., 2014). Moreover, boys, compared to girls, may feel more social stigma or embarrassment, experience reluctance to disclose abuse -related details, or provide minimal details or less complete reports about sexual abuse.

Child Psychopathology, Maltreatment History, and Memory

Research suggests that trauma-related psychopathology and children’s maltreatment histories can affect how children attend to, interpret, consolidate, and recall traumatic details later. We briefly discuss each topic in turn while acknowledging that the evidence reported needs replication as well as experimental and longitudinal study designs.

Trauma-related psychopathology. Perhaps the most well-known trauma -related psychopathology is posttraumatic stress disorder (PTSD). Surprisingly, few eyewitness memory studies have explicitly focused on PTSD in children and adolescents; evidence from related studies appears to be mixed as to whether children and adolescents with PTSD have poorer or better memory abilities compared to children and adolescents without PTSD (Beers & De Bellis, 2002; de Decker, Hermans, Raes, & Eelen, 2003; Eisen et al., 2007; Moradi, Taghavi, Neshat-Doost, Yule, & Dalgleish, 2000; Ogle et al., 2013; Yasik, Saigh, Oberfield, & Halamandaris, 2007). Although some findings suggest that the children and adolescents with PTSD have difficulty in accessing the specific details of autobiographical memories (e.g., Moradi et al., 2008), other studies indicate the opposite pattern: Adolescents who reported child sexual abuse as their most traumatic life event and who evinced greater PTSD symptomatology were observed to have more specific memory reports than adolescents with lower PTSD symptomatology (Ogle et al., 2013). Moreover, at least by older adolescence, PTSD is associated with particularly accurate memory of child sexual abuse (Alexander et al., 2005). Nevertheless, symptoms may be present in children and adolescents with PTSD that affect interview responses, such as inattentiveness, impulsivity, and lower working memory abilities (Beers & De Bellis, 2002; Ogle et al., 2013).

When different standardized measures of trauma -related symptoms (e.g., PTSD, depression, dissociation) are combined to form a single dimension of psychopathology, memory for positive information seems to be more consistently affected adversely than is memory for negative information (Goodman et al., in press). That said, in a study of maltreated children’s memory, clinicians’ appraisals of children as having lower adaptive functioning were significantly correlated with the children’s memory errors for a stressful event (Eisen et al., 2007).

Maltreatment histories. As was just discussed, the psychopathology resulting from traumatic experiences, such as child maltreatment, can possibly affect children’s memories. Trauma-related psychopathology is not always differentiated from maltreatment history in existing studies (e.g., Carrion, Weems, & Reiss, 2007). Forensic interviewers should take into account possible effects that maltreatment histories could have on children’s cognitive and socioemotional functioning, as these effects may influence how maltreated children remember and recount traumatic events to others (see Cicchetti & Toth, 2005; Goodman, Quas, & Ogle, 2010).

Prolonged or chronically elevated levels of stress, like those experienced by some maltreated children, may lead to dysregulation of the Hypothalamic-Pituitary Adrenal (HPA) axis, the stress hormone system responsible for releasing cortisol (Cicchetti, Rogosch, Howe, & Toth, 2010; Sapolsky, 1996), and may eventually (e.g., in adulthood) have an effect on hippocampal functioning (a region of the brain that is important for memory ). However, there is conflicting evidence of HPA axis effects on children’s eyewitness memory (see Goodman et al., in press, for review). Maltreatment histories may affect, however, how children attend to and process trauma -related details. Some researchers have found that maltreated children are particularly attentive to negative stimuli (e.g., Masten et al., 2008; Pollak, Messner, Kistler, & Cohn, 2009), which could lead to stronger memory representations for negative events, such as abuse (see Goodman et al., 2010). Yet basic memory processes (e.g., associative memory) for maltreated and nonmaltreated children appear to be quite similar (e.g., Cicchetti et al., 2010; Howe, Cicchetti, & Toth, 2006), and both groups can report details accurately about positive and negative events (e.g., Eisen et al., 2007). In other words, it appears that children with maltreatment histories can be just as accurate when recounting details about their lives as children without maltreatment histories.

An important consideration, like the one we addressed in the PTSD discussion, is that children with maltreatment histories, on average, display delays or deficits in several cognitive processes other than memory , such as executive function, language abilities, and IQ compared to children without maltreatment histories (Goodman et al., 2010; Lyon & Evans, 2014; Porter, Lawson, & Bigler, 2005; Veltman & Browne, 2001). Deficits in these cognitive areas could have implications for maltreated children in reporting their abuse episodes. For example, deficits in language comprehension and production (e.g., Veltman & Browne, 2001) may interfere with maltreated children’s understanding of an interviewer’s questions or phrasing, impinge on children’s abilities to articulate and describe details, and decrease the ability to monitor reports for errors while recounting their abuse (e.g., Lyon & Evans, 2014).

A relatively new avenue of research has revealed that children may benefit from truth induction procedures prior to their interviews (e.g., taking a child -friendly version of an oath to tell the truth); these procedures seem to increase disclosure of information while not increasing false details or false reports for both maltreated and nonmaltreated children (Evans & Lee, 2010; Lyon & Evans, 2014). However, although research reveals that maltreated children benefit from truth induction procedures, interviewers may need to simplify such procedures due to possible language deficits exhibited by the children (Lyon & Evans, 2014).

The Interviewer

Child forensic interview ers play a crucial role in investigations of child sexual abuse . It is not an easy job. With increasing frequency, child forensic interviewers are expected to master and use science-based interview protocols; to obtain in a nonleading way (e.g., “So tell me why you are here today?”) the legally specified (and often quite precise) information needed by child protection workers, law enforcement officers, and/or district attorneys to pursue or drop a case; to question children while being videotaped, observed, and evaluated by professionals in an adjoining room; to remain calm and supportive, and yet unbiased, even when children are recounting horrific experiences; to build rapport with traumatized children within minutes of meeting them; to interview a wide array of children, from 3-year-olds, who have no idea why they are there and can barely sit in a chair for more than a few minutes, to 14-year-olds, who know exactly why they are there and would rather be anywhere else; to work quickly because the children’s attention may not last long and anyway, the professionals observing have little time before they have to leave to investigate other crimes; and to be able to defend their interviews in court. It’s a lot to ask of anyone, and yet, seasoned child forensic interviewers (often specially trained social workers) do their jobs with apparent ease, interviewing hundreds of children each year—often thousands of children across the interviewer’s career.

Perhaps the most important, and yet understudied, interviewer skill is the ability to build rapport with children quickly. Children who are intimidated are more likely to “shut down” and refuse to talk, or to be suggestible (e.g., Saywitz & Nathanson, 1993). Science-based protocols uniformly begin with a rapport-building phase, which often consists of having children recount everyday events. Yet the interpersonal interactions that take place during that recounting, not often the subject of research, are likely crucial for making the child feel at ease, open up, resist false suggestions, and trust the interviewer enough to disclose what may be highly personal, distressing, and embarrassing information. To date, research indicates that females and males can be equally proficient in obtaining accurate memory reports from children, as long as sufficient rapport is built (Schaaf, Alexander, & Goodman, 2008). In addition to a supportive stance by the interviewer, a supportive interview context (one where children feel safe and comfortable discussing traumatic pasts) has been shown to be helpful for children of all ages when recounting events, including when disclosing abuse details (Brubacher et al., 2014; Lamb, Orbach, Hershkowitz et al., 2007; Lamb, Orbach, Warren et al., 2007; Saywitz & Nathanson, 1993).

The age appropriateness of the language used by the interviewer can also influence children’s responses. Language that is too complex for a given age, such as using words or sentence constructions that children do not understand, can result in greater memory error , especially when combined with intimidation (Carter, Bottoms, & Levine, 1996). Forensic interviewers may be able to account for possible cognitive deficits, like language difficulties, in maltreated children by adapting questions and interview formats to accommodate the children’s language needs (e.g., asking open-ended questions, keeping questions simple). Similarly, interviewers may want to use specific labels that are spontaneously generated by children, after clarifying what the children mean by the terms (e.g., Brubacher et al., 2014).

Another important skill for interviewers to possess is the ability to maintain neutrality and not reinforce specific types of answers or specific types of content (Garven, Wood, Malpass, & Shaw, 1998; Saywitz, Lyon, & Goodman, 2010). Interviewers are advised to keep an open mind about what may have occurred and to test alternative hypotheses through their questioning, rather than assume child sexual abuse took place (Ceci & Bruck, 1993). Although interviewer factors are important, when Gilstrap and Ceci (2005) analyzed interviews of children conducted by professional interviewers, these researchers reported that “interviewers’ use of leading questions did not result in increased acquiescence as previously found… analyses showed that it was possible to predict directly from child-to-child behavior, effectively skipping the intervening adult behavior.” (p. 40). These results imply that individual differences among children in their willingness to acquiesce to or counter misleading questions were more important than the interviewer’s behavior, at least for the interviews studied, which were likely not highly leading.

In any case, interviewers should keep in mind that they are not usually the first people to interview the alleged child victim. Especially with young children, the first person to suspect sexual abuse and to whom the child has disclosed is likely to be the parent, particularly the mother. Research indicates that children are less suggestible about abuse when interviewed by their mothers versus strangers, and that children can maintain accuracy in later forensic-like interviews even when parents are wrongly suspicious that something bad happened (Goodman, Sharma, Thomas, & Considine, 1995; McWilliams, 2014). That said, with sufficient pressure, coaching, or misinformation, some children may clam up or provide false information (e.g., Bottoms, Goodman, Schwartz-Kenney, & Thomas, 2002; Poole & Lindsay, 1995). Interviewers often ask children if anyone told the children to say something or not to say something, in an effort to sort out possible contamination. Although there is little research to validate this practice, asking children such questions may help fact finders feel more comfortable evaluating children’s statements.

It is incumbent on child forensic interview ers to be conversant with the laws governing their counties, state, and countries. Increasingly in the United States, the courts are considering child forensic interviews to fall within the realm of law enforcement and for the interviews therefore to be considered “testimonial.” This means that, at trial in criminal proceedings, it is likely that the child victim will have to testify face-to-face with the defendant in order for the videotaped forensic interview to be entered into evidence and shown to the jury, as the videotape is considered a form of “hearsay” (Myers, 2011). In other types of hearings (and in other countries), however, the videotape might be more easily shown even without the child appearing. When shown, the interviewer and the interview will likely be scrutinized as to whether proper rapport was established, whether the language used was age appropriate, and so forth—but perhaps especially, the types of instructions given, the interview questions asked, and interview methods used. We turn to these topics next.

The Interview

The cognitive and emotional demands of the interview conducted with child witnesses must be considered when assessing their reports. Laboratory and field studies reveal that young children are particularly sensitive to interview conditions that overburden their cognitive abilities, introduce social pressures that promote lack of disclosure or tainted testimony , and neglect their socioemotional needs and concerns (Brubacher et al., 2014; Lamb, Orbach, Hershkowitz et al., 2007; Lamb, Orbach, Warren et al., 2007; Lyon & Evans, 2014). We focus on three components of the interview that may influence children’s eyewitness abilities: interview instructions, interview questions, and interview props.

Interview Instructions

For children, the expectations and conversational rules governing child forensic interview s differ from those of natural conversations with adults. For example, in their day-to-day lives, children are accustomed to being questioned by parents and teachers who often already know the answers to questions; and, in some cultures, children are not supposed to disagree with or correct adults. Given the uniqueness of child forensic interviews , children need instructions on how to proceed. Fortunately, researchers have developed instructions to aid children’s performance.

Based on the research, science-based forensic interview protocol s typically provide a set of instructions that interviewers are told to state toward the start of the interview . For example, in Lyon’s Ten Step Interview protocol, instructions to be provided include the following: the Don’t Know Instruction (e.g., “If I ask you a question and you don’t know the answer, then just say, ‘I don’t know’.”); the Don’t Understand Instruction (e.g., “If I ask you a question and you don’t know what I mean or what I’m saying, you can say, ‘I don’t know what you mean’.”); and the You’re Wrong Instruction (e.g., “Sometimes I make mistakes or say the wrong thing. When I do, you can tell me that I am wrong.”). To be effective, such instructions need to be provided with children being able to practice each one, as specified in the protocols. Even young children can profit from such instructions (e.g., Cordon, Goodman, & Saetermoe, 2005).

Interview Questions

One of the most widely studied topics on child forensic interview ing concerns the influence of question type on children’s memory reports. Question format can influence both the accuracy and the amount of information that children provide (Bjorklund, Bjorklund, Brown, & Cassel, 1998; Lamb, Orbach, Hershkowitz et al., 2007; Lamb, Orbach, Warren et al., 2007). Open-ended questions and prompts such as “I heard you saw a policeman last week, tell me more about that” (Lyon, 2005) are widely recognized as the most ideal format for children, typically resulting in more accurate responses than specific closed-ended questions (Eisen et al., 2007; Poole & Lindsay, 1995). Open-ended questions are preferred because they do not limit children’s options for a response and are typically not leading (or at least are often less leading than other types of questions). Instead, open-ended questions allow children to give their own account without significant amounts of information being provided or implied by the interviewer (Lamb, Orbach, Hershkowitz et al., 2007; Lamb, Orbach, Warren et al., 2007; Saywitz et al., 1991).

Unfortunately for forensic interview ers, children, especially preschool-aged children, do not always offer a significant amount of information when solely asked open-ended questions (Goodman et al., 1994). Often, children need memory cues to recall certain information, or children may assume adults already know the information and thus the children fail to include it in their reports. To obtain the amount of information needed for most forensic situations, open-ended questions must be followed up with direct questioning. Direct questioning can be a beneficial strategy when working with younger witnesses; however, interviewers must be careful when posing direct questions to young children. Direct questions, when formatted in certain ways, can be leading and introduce misinformation (Peterson & Biggs, 1997).

Researchers have theorized that exposure to misinformation and suggestion can alter children’s actual memories, rather than just the children’s responses. One such theory specifies that memory errors result from misinformation when children fail to engage in source monitoring efforts to differentiate an experienced event from what was implied in the misleading question (Ceci, Loftus, Leichtmen, & Bruck, 1994). However, misinformation and suggestion may not alter children’s actual memories but rather create social pressures that influence children’s responses. During forensic interview s , children are questioned about sensitive information, usually by authority figures. Some children acquiesce to questioning in an attempt to please the interviewer, rather than reporting their true memory. Awareness of how misinformation and suggestion can be inadvertently introduced during a forensic interview may permit interviewers to avoid such influence on children’s testimony and on children’s credibility down the legal road.

Two types of direct questions that should be avoided are yes/no (“Did your dad touch you?”) and forced choice questions (“Was the person who hurt you your mom or your dad?”). These types of questions can be problematic for young children for a variety of reasons. First, yes/no and forced choice questions limit the options children have for a response . Children could interpret these forms of questioning to indicate that adults already know the correct answer. As a result, children may falsely affirm or choose one of the options, respectively, due to social pressure rather than relying on their own memory . Second, to pose a yes/no or forced choice question, an interviewer may provide false information in the question itself, which could introduce misinformation.

To avoid introducing suggestion into an interview with direct questioning, some researchers have recommended that forensic interview ers attempt to utilize a “Wh-” question format for their specific questions (Peterson, Dowden, & Tobin, 1999). For example, after a child discloses, “My dad touched me,” a useful “Wh-” questions would be, “Where did your dad touch you?” The former structure of the question allows interviewers to probe for more specific details during their investigations, while avoiding many of the problems elicited by yes/no and forced choice questions (Peterson & Biggs, 1997). These specific questions should be followed up with open-ended questions, such as “Tell me more” or “Then what happened?” However, many researchers recognize that avoiding all specific (and therefore potentially leading) questions during forensic interviews with children can be difficult, and structured forensic interview protocol s have been developed to address this issue, permitting use of some specific questioning (e.g., NICHD Protocol, Ten Step Investigative Interview; a discussion of these protocols is provided in a later chapter of this book). Moreover, misleading questions do not necessarily led to error in children’s later reports (Peterson, Parsons, & Dean, 2004).

The frequency with which an interviewer asks a child about specific information is also important to consider, although the influence of repeated questioning on children’s memory reports is complex (Goodman & Quas, 2008). Repeated questioning presents a risk, within or across interviews , if children interpret these repetitions as insinuations that their initial accurate reports were not desired responses. However, repeated questioning also has the potential to benefit memory. Questions that are repeated in sequential interviews can give children the opportunity to rehearse memories which can potentially strengthen their memory traces and representations, and keep such memories from fading over time (e.g., Quas et al., 2007). In some instances, repeated questioning can also lead to hypermnesia, where additional details of traumatic memories are recalled after a delay (La Rooy, Pipe, & Murray, 2005; Payne, 1987).

Over time, memories are subject to forgetting and distortion as delay increases between when events were experienced and when events are described; in this regard, it is important to conduct interviews as soon as possible. Laboratory studies reveal that the completeness of children’s memory reports may remain constant over time; however, the amount of detail has been shown to change, accuracy decreases, and children have increased risk of exposure to misinformation (e.g., La Rooy et al., 2005; McWilliams et al., 2013; Waterman & Blades, 2013; but see Quas et al., 1999). Additionally, delays have been linked to children’s willingness to guess in response to unanswerable questions with longer delays associated with increased willingness to guess (e.g., Waterman & Blades, 2013), which could lead to the possibility of errors in a forensic situation. To minimize this risk, ideally interviews should be conducted as soon as possible to tap into children’s memory while it is still robust. It is also, possible, however, that with development, some children will be better able to recount events (e.g., have the words, retrieval strategies, and concepts) when they are older than when they were younger.

Interview Props

Research on the use of forensic interview ing props, such as anatomically detailed dolls, body diagrams, and comfort drawings, has revealed mixed results. Although such tools have been used by investigators to interview children of all ages in nearly half of child sexual abuse cases (Hlavka, Olinger, & Lashley, 2010), the effectiveness of these props at eliciting accurate details and minimizing false reports remains in question (e.g., Goodman & Melinder, 2007). Professionals reportedly prefer these tools as they are believed to help children clarify prior disclosures, overcome communication barriers, and support coping strategies to help children discuss traumatic details (e.g., “You said the man touched you; Point to where he touched you.”; Hlavka et al.), although their effectiveness may not supersede that of other empirically based interviewing methods (e.g., Salmon, Pipe, Malloy, & Mackay, 2012).

Particularly for younger children (3- to 5-year-olds) who may lack cognitive abilities to understand the dual-representational function of dolls and diagrams, interviewing tools such as these may be ineffective at eliciting additional correct information (e.g., Salmon et al., 2012). Children (approximately 6 years and older) tend not to report false details with the use of dolls or diagrams; however, the use of these props does not always result in children providing additional accurate details when compared to the use of verbal prompts (e.g., “You mentioned ‘X’, Tell me more about ‘X’.”; Salmon et al., 2012). For child forensic interview ers, verbal prompts (e.g., “Tell me more”) and effective interview strategies (e.g., NICHD protocol, Ten Step Interview, Narrative Elaboration Technique; Saywitz & Comparo, 2013) may be preferred over use of dolls and diagrams given the lack of empirical validation of props; if used, it is often recommended that body diagrams should be presented after a disclosure (e.g., at the end of the interview) to provide clarification of prior disclosures and with cognitively competent children.

Evidence does support, however, use of comfort drawings (i.e., allowing children to draw freely while they answer questions) as this may serve, in effect, as a coping strategy to provide comfort while disclosing traumatic details and also by keeping children engaged, facilitating more opportunities for interviewers to ask nonleading open-ended questions (Katz, Barnetz, & Hershkowitz, 2014; Patterson & Hayne, 2011). Comfort drawing has been shown to increase the amount of information reported (although it may not increase the number of accurate details), but of importance, it did not increase errors (e.g., Patterson & Hayne, 2011); children are merely comforted while disclosing. Props that do not adversely affect the accuracy of children’s reports but provide emotional support could be of interest because children feel more comfortable disclosing traumatic details when emotionally supported (Katz et al., 2014; Malloy et al., 2011). If comfort drawings are used, interviewers are cautioned against interpreting meaning of details within the drawing (e.g., colors chosen to represent specific people) as these details are not necessarily related to the emotional content or memory for the event (Crawford, Gross, Patterson, & Hayne, 2012).

Conclusion

It is challenging to conduct forensic interview s with children, especially young children or children who are hesitant to disclose. Although scientific researchers continue to examine and identify effective ways to interview children in child sexual abuse cases, many factors can influence children’s memories and reports, some of which are still in need of research and others of which will always be difficult for researchers to study ethically. For example, there is little scientifically sound research on children’s memory and suggestibility regarding their parents’ actions, regarding situations about which children feel shame or fear retribution, regarding reporting of events with anatomical dolls alone (apart from other props) or human figure drawings that show naked bodies, and so forth. There is still much to learn.

Nevertheless, our review, although not exhaustive, illustrates some of the main issues that should be considered, based on research at this point in time, when interviewing children about sexual abuse in relation to the interviewee, interviewer, and interview . A list of basic principles interviewers might review prior to interviewing children include:

  • The amount of information children provide increases with age.

  • Preschool children on average tend to report less information on their own in free recall and also to be more suggestible than older children. For this and other reasons, very young children (e.g., 3-year-olds) can be especially difficult to interview , even with scientifically based protocols.

  • Carefully consider both cognitive and socioemotional abilities and needs of children, and that these needs change with development.

  • Individual characteristics of each child should be considered, including for young children their level of cognitive control and verbal ability.

  • Psychopathology factors may result in less accurate reports (e.g., for positive information).

Nevertheless, children with certain symptoms of psychopathology (e.g., PTSD) are likely to have intact memory abilities, yet potential deficits related to other domains (e.g., inattentiveness, impulsivity), which could impinge on answering questions accurately.

  • Maltreated children’s basic memory abilities are similar to that of nonmaltreated children, but effects of maltreatment may influence the content of memory and other domains (e.g., language).

  • Interviewers should ensure they can develop rapport with children quickly.

  • Interviewers should remain neutral and unbiased in their questioning of children and be cognizant of current laws regarding child forensic interview s and child sexual abuse in their jurisdictions.

  • Open-ended, nonsuggestive questions can help circumvent misinformation effects and augment accurate reports; children’s credibility is also less likely to be questioned if the interview is not overly leading.

  • Interview props do not necessarily increase reporting of accurate details over empirically supported interviewing methods, particularly for preschool-aged children.