Keywords

Overview

The child ’s account of what happened is often the only evidence available in investigations of child sexual abuse . Whenever there is a lack of physical or corroborating evidence, the child’s testimony and ensuing legal decisions can have far-reaching consequences for all those involved. Even when there is physical evidence of abuse , it often remains important to have an account from the child so that what happened and who was involved can be fully clarified. Obtaining and evaluating the evidence provided in these cases presents a challenge for many legal systems that were not designed to deal with the complexities of proceedings that involve children. Most importantly, those referred to simply as “children” actually represent a very broad categorization including infants, toddlers, young children, older children, and adolescents. Indeed, in many jurisdictions children and young persons are considered vulnerable until the age of 18 years. The accounts that children provide in forensic interview s must thus be viewed and assessed within the context of their social, emotional, and cognitive development by appropriately qualified experts. A substantial amount of scientific evidence guides our understanding and approach to many issues surrounding forensic interviews with children, informing us about how investigative interviews should be conducted and the pitfalls that should be avoided (e.g., Lamb, Hershkowitz, Orbach, & Esplin, 2008; Lamb, La Rooy, Malloy, & Katz, 2011; Poole & Lamb, 1998; and this volume).

In this chapter, we discuss key aspects of assessing the quality of forensic interview s conducted with child witnesses . We begin by discussing the quality of the case material, before reviewing the key components of investigative interviews and methods of assessing the quality of each of these components. The overall structure of the interview is discussed with reference to the use of “ground rules” and the effectiveness of rapport building and episodic memory training during the presubstantive phase of the interview. Issues surrounding the identification of appropriate and inappropriate questions and prompts are examined in relation to the strengths and weaknesses of memory. Additional issues that can influence the effectiveness of an interview, such as the use of repeated questions, the specific vs. generic nature of accounts, and the use of anatomical dolls, are also considered in light of the impact they may have on children’s responses.

Who Should Conduct Assessments of Interview Quality?

Requests for assessments of forensic interview s are often made by lawyers defending accused individuals in criminal and civil court cases to determine whether or not interviews were conducted fairly. An expert witness will often be asked to prepare a report for the parties involved that can inform fact finders of any concerns with the way the interviews were conducted, and thus affect the weight that should be placed on the interviews as evidence. Given what is currently known about the laws of memory and how they apply to child forensic interviewing , combined with the social, linguistic, and emotional factors that are involved, it is not surprising that professionals turn to psychologists to provide these assessments. Expert witness interview as evidence . Sometimes, however, there is heated debate about exactly who should be considered an expert, especially when issues to do with memory and suggestibility are to be considered.

In many jurisdictions judges themselves decide who should be considered an expert , but there are also professional standards and ethical considerations, about which psychologists should be particularly aware. Some experts are so considered because they are recognized by their peers as scholars in the field. The easiest way to assess this “recognition” is through publications in peer-reviewed journals. The peer-review process functions to improve the quality and coherence of scientific research and demonstrates that the expert’s reasoning is consistent with that of scholars who have had the opportunity to evaluate their work. Although qualified experts may still have differences of opinion, selecting suitably qualified expert witnesses decreases the chance that there will be “battles of the experts” in court. In reality, many professionals are willing to put themselves forward as memory experts based on their professional experience and training rather than an understanding of the dynamics of memory and scientific research about interviewing children. Thus, disagreements between experts are often better understood with reference to individual training and qualifications. For this reason, whenever experts provide court reports it is advisable to provide up-to-date curricula vitae so that professional qualifications are transparent.

Case Material to be Assessed

There is widespread agreement that both electronic records and verbatim transcripts should be examined in order to appropriately assess the quality of forensic interview s (La Rooy & Block, 2013; Lamb, Orbach, Sternberg, Hershkowitz, & Horowitz, 2000; Warren & Woodall, 1999). The advantage of having electronic recordings of interviews is that children’s evidence is accurately preserved so that systematic analyses can take place. Electronic recordings and written transcripts are equally important, and thus, the first step is to have the recording fully transcribed so that all of the interviewer’s questions and child ’s responses can be easily examined.

In some jurisdictions interviews are not routinely recorded and transcribed, and interviewers instead rely on contemporaneous or retrospectively written notes. It is difficult to understand why this practice persists because note taking is unlikely to result in an accurate record of the exact questions posed and the information elicited from children. Lamb and colleagues (2000) directly compared handwritten notes taken by experienced interviewers with electronic recordings of the same interviews. The results showed that more than half of the questions and prompts posed by the investigators were not recorded verbatim, and a quarter of the details reported by the interviewees were not recorded. Importantly, details of the children’s accounts were often incorrectly attributed to being elicited by higher quality interviewer prompts, (i.e., the interviewers were recording the interviews as being of a higher standard than they actually were). Notes from interviews often include the gist of what each person has said rather than the exact wording of the questions and answers. Therefore, analysis of interviewer notes is somewhat problematic, and this issue should be clearly identified for fact finders if a transcript is not available.

Following transcription, the interview must be carefully inspected to identify aspects of the interview that are conducted in accordance with best practice, as well as any areas of concern surrounding the nature and appropriateness of questioning. The transcript allows for a more accurate examination than the electronic recording alone because it is possible to move backward and forward through the interview quickly to determine where information was introduced, and whether it was introduced by the child or the interviewer. It is not possible to do this type of detailed analysis in “real time” by simply viewing a recording of the interview, and it is too time consuming and difficult to rewind and fast forward electronic recordings to verify this type of information with any degree of accuracy.

A video recording is ideal because it allows consideration of nonverbal cues provided by both the interviewer and the child . Children may nod in response to prompts or indicate body parts in a way that cannot be captured in a typed transcript; it is only through careful viewing of a video that one can clarify the children’s responses. Interviewers also use nonverbal cues, such as gestures, intonation, and pauses, which may affect the child’s responses, so examining a video may help to put interviewer prompts in context. For example, if a transcript states “child nods,” it is important to check whether this was indeed a clear gesture or whether the child appeared to be mimicking the interviewer. The video cannot, however, be used to make judgments about truthfulness based on the child’s demeanor. Although some experts may claim that they are proficient at detecting nonverbal cues to lying and truth telling, research shows that adults’ accuracy at detecting children’s lies is at chance levels (i.e., they are no better than guessing; see Vrij, 2008, for a comprehensive review of this issue).

Rapport Building

During investigative interviews children may feel apprehensive about discussing highly personal experiences with strangers (Saywitz, Goodman, Nicholas, & Moan, 1991). This is especially true when discussing experiences of sexual abuse , which are intimate and embarrassing. It is recommended that interviewers begin by building rapport with interviewees to put them at ease, make them feel more comfortable disclosing sensitive information, and set a supportive context that encourages children to trust interviewers (Collins, Lincoln, & Frank, 2002; Goodman, Bottoms, Schwartz-Kenney, & Rudy, 1991; Lamb, Orbach, Warren, Esplin, & Hershkowitz, 2007; Sternberg et al., 1997). This can be done by discussing neutral or positive personal topics, such as friends and family, school, hobbies, etc., to get to know the children better (Lamb, Sternberg, & Esplin, 1998; Poole & Lamb, 1998).

There is evidence that rapport building improves both the amount and accuracy of information that children provide about personal experiences, especially if it is established using open-ended prompts (Roberts, Lamb, & Sternberg, 2004; Sternberg et al., 1997). Rapport building is especially important for reluctant children (Wood, McClure, & Birch, 1996), making it more likely that they will disclose abuse and provide more details about their experiences (Hershkowitz, Lamb, & Katz, 2014; Hershkowitz, Lamb, Katz, & Malloy, 2013). Overall, high levels of interviewer supportiveness help to improve the accuracy of children’s reports (Almerigogna, Ost, Bull, & Akehurst, 2007; Carter, Bottoms, & Levine, 1996; Davis & Bottoms, 2002; Goodman et al., 1991; Quas & Lench, 2007), but in practice interviewers do not often use open-ended questions during the rapport building phase (Warren, Woodall, Hunt, & Perry, 1996), or they may not do enough rapport building (Sternberg, Lamb, Esplin, & Baradaran, 1999). If rapport building is not done effectively, the benefits for children’s testimony will not be evident as seen in laboratory studies.

Although creating a supportive environment is an important goal of rapport building, interviewers must be cautious about how they express their support to children. Normal responses to the disclosure of abuse , such as shock or surprise, might be seen as attempts to overempathize and shape the testimony . Thus, empathetic responses, such as, “That must have hurt,” could be seen as examples of interviewers suggesting information that was not provided by the children. Research also shows that children are sensitive to positive reinforcement and praise for answers that they think interviewers want to hear, so even comments like, “Aren’t you wonderful?” can be seen as ways of shaping testimony. While it is undesirable that interviewers overempathize with children in interviews , support can be provided in a more neutral manner that does not attract criticism. Commenting on the effort a child is making by saying, “I can tell you are trying hard,” and small acts of kindness such as offering a glass of water or a tissue should not be considered attempts by interviewers to alter the content of the child’s testimony , but can nonetheless help an interviewer build rapport with a reluctant child.

Ground Rules

Interviewers often establish “ground rules” or interview instructions early on that indicate what is expected of children during the interviews and define the conversational rules. Although they are often communicated at the beginning of an interview , they may also be communicated after rapport building or at any time during the interview when appropriate. Investigative interviews are strange for children, who most often interact with more informed adults who ask questions of them to assess their knowledge (Lamb et al. 2007a; Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007b; Lyon, 2010). Children must understand that interviewers do not know what happened, and communication of the ground rules is designed to help children become aware that they are in control and should not feel pressured to answer questions (Lamb, Orbach, Hershkowitz et al., 2007; Lyon, 2010; Saywitz, Camparo, & Romanoff, 2010; Sternberg, Lamb, Esplin, Orbach, & Hershkowitz, 2002). Commonly recommended ground rules include (a) communicating that it’s important to tell the truth /promise to tell the truth (Evans & Lee, 2010; Lyon & Dorado, 2008; Talwar, Lee, Bala, & Lindsay, 2002); (b) asking the child to demonstrate that s/he understands the difference between telling the truth and telling a lie; (c) it is okay to tell the interviewer if you do not understand (Saywitz, Snyder, & Nathanson, 1999); (d) it is okay to say “I don’t know” (Gee, Gregory, & Pipe, 1999; Saywitz & Moan-Hardie, 1994; Waterman & Blades, 2011); (e) it is okay to correct the interviewer (Krackow & Lynn, 2010; Saywitz & Moan-Hardie, 1994); and (f) if questions are repeated, it does not mean that that the interviewer was unhappy with the previous answer.

The ground rules are considered an important part of the presubstantive phase of the interview because they are designed to remove implicit pressure on interviewees to guess if they are not sure about what happened, and/or to acquiesce to interviewer suggestions. If the ground rules are not explained, children may answer in the way they think interviewers want to hear in order to seem cooperative (Ceci & Bruck, 1993, 1995; Ceci, Kulkofsky, Klemfuss, Sweeney, & Bruck, 2007; Melnyk, Crossman, & Scullin, 2007). The ground rules are intended to reduce the effects of poor quality questions, such as misleading or closed questions, because children will feel comfortable indicating that they do not know the answer or do not agree with the interviewers’ suggestions (e.g., see Lamb et al., 2008, pp. 85–87). The ground rules help to discourage children from hazarding a guess when they are actually unsure about the answer and are also associated with how informative children are during the substantive phase (Teoh & Lamb, 2010).

In sum, the research makes clear that it is desirable to see ground rules communicated to children at the start of forensic interview s ; however, ground rules are not always communicated, and there is variation both within and across jurisdictions about how many are communicated and which ones are perceived to be most important. For example, testing a child ’s understanding of the truth and lies is a legal requirement in some jurisdictions, but not in others.

The Practice Interview

Interviewers commonly extend rapport in the presubstantive phase of the interview by conducting what is known as a “practice interview” (sometimes referred to in academic writing as “episodic memory training”; e.g., Orbach et al., 2000). The practice interview ideally involves interviewers using open prompts to elicit detailed accounts of neutral, specific, real experiences from interviewees. The purpose of this phase is to provide children with practice at remembering specific experiences in detail rather than recalling the gist of what happened. Importantly, this is when interviewers should introduce open prompts and continue building rapport. Children thus have a chance to feel successful at providing information and feel in control. Practice interviews also provide good opportunities for interviewers to better understand the cognitive abilities and communicative styles of the children they are interviewing (Roberts, Brubacher, Price, & Powell, 2011).

Research on the benefits of practice interviews has shown that, after practice phases, children’s reports in the substantive phase are longer and contain more details (Brubacher, Roberts, & Powell, 2011; Hershkowitz, 2009; Price, Roberts, & Collins, 2013; Sternberg et al., 1997), and are also more accurate (Roberts et al., 2004). Even a low quality practice interview is better than not conducting a practice interview at all, but in particular, a practice interview containing open-ended questions is more likely to increase the amount of information provided during the substantive phase of the interview than one using closed questions (Anderson, Anderson, & Gilgun, 2014; Price et al., 2013; Sternberg et al., 1997).

While practice interviews are desirable features of forensic interview s , there may be circumstances in which the practice interview is omitted. For example, if a child seems eager to disclose or discuss why they came to the interview, an interviewer might not include aspects of the presubstantive phase and instead allow the child to discuss substantive events. It is important to remember that the absence of a practice interview does not mean that information elicited using open prompts in the substantive phase is less likely to be accurate (Brubacher et al., 2011).

When assessing forensic interview s it is important to be aware that interviewers often think that they have conducted practice interviews when they have not done so. Interviewers sometimes ask children for descriptions of fantasy-based topics such as television shows, movies, and video games. This is undesirable because it could adversely affect the child ’s testimony (based on the literature on suggestibility and false memory , discussed below) or the credibility of the witness , whose ability to distinguish between fantasy and reality may be questioned (Woolley & Ghosaini, 2013). Interviewers may also ask for general descriptions of repeated or highly scripted events (e.g., dance lessons, sports, or a birthday party). Though children are describing real events, it does not provide them with the opportunity to practice discussing specific episodic memories, as they may instead report general details about what usually happens in such events. In neither case would interviewers be using the practice interview in the way that it is intended and supported by research.

Transition Phase

Building rapport, explaining ground rules, and conducting practice interviews are all things that can be done in the presubstantive phase to prepare children for their roles as informative witnesses. Once these phases are complete , interviewers should introduce the substantive topic non-suggestively. It is crucial to assess the transition from the presubstantive to the substantive phase of the interview . How was the allegation elicited? Did the child make an allegation free from any pressure, or was it necessary for the interviewer to use direct questions and/or to introduce the allegation? Most research-based protocols recommend that the transition be initiated by the interviewer using a neutral prompt such as, “I’d like to talk about why you came to see me today.” As vague as this seems, research shows that most children who have already made a prior disclosure will make an allegation when given this opportunity (Hershkowitz, Fisher, Lamb, & Horowitz, 2007).

Some children do not provide any information about alleged events in response to these initial open prompts. When this happens, depending on the case characteristics, it may be necessary for the interviewer to ask a very direct or suggestive question like, “Did somebody hit you?” On one hand there may be physical evidence and/or strong suspicions that abuse has occurred that appear to justify such direct suggestions. On the other hand, the suggestibility of child witnesses and evidence that even neutrally posed suggestions can elicit inaccurate information are causes for concern (Ceci et al., 2007). In some cases, interviewers’ suggestions have resulted in descriptions of entire false events (e.g., Bruck, Ceci, & Hembrooke, 2002; Ceci, Huffman, Smith, & Loftus, 1994; Ceci, Loftus, Leichtman, & Bruck, 1994; Quas et al., 2007; Strange, Garry, & Sutherland, 2003). However, in these cases interviews were considered to be very suggestive, and in the vast majority of forensic interview s with children, issues surrounding false allegations and false memory surface rarely.

The Substantive Phase: Assessing the Questions and Prompts

A consistent and clear message that has emerged from research on investigative interviews with children is that the manner in which children are questioned, and the types of questions that they are asked, can dramatically affect the accuracy of the information they provide. Therefore, those tasked with assessing the quality of interviews must be able to identify different types of interviewer utterances and be familiar with research informing us about their strengths and weaknesses. Although different researchers/protocols sometimes use different terminology to identify the types of prompts and questions that interviewers ask, there is strong agreement based on both memory and linguistic factors about the types of questions that are the safest to use and those that are more risky (Kassin, Tubb, Hosch, & Memon, 2001).

Open prompts. Memory experts agree that children ought to be allowed to describe events in their own words, free from pressure and any suggestive influence. For this reason, interviewers are advised to use as many open prompts as possible (see Lamb et al., 2008 for a review). The most common example of an open prompt is, “Tell me what happened.” Other examples of open prompts that are also desirable in forensic interview s are, “Tell me more about that,” “Tell me everything about that,” “Then what happened?,” and “What happened next?” When reviewing a forensic interview a useful rule to use is, “Does the prompt allow the child to respond using a narrative, rather than just a few words?” If the answer is “yes” then the interviewer utterance is likely to be open.

It is also acceptable to use open prompts to enquire about specific information, especially relating to elaboration of information that the child has already provided (Lamb et al., 2003). One way this can be achieved is using anchor points in time as components of open prompts. For example, “Tell me what happened from the moment he came in to your room to the moment that he left your room,” is an acceptable way of refocusing a child to provide more details about a critical element of the account. Similarly, combining information that the child has provided with open prompts is another method of increasing the specificity of a question in a non-suggestive manner; e.g., “You said he hit you, tell me more about that.” Because information obtained using open prompts comes from free recall memory it is more likely to be accurate than information elicited using other question types (e.g., Dent & Stephenson, 1979; Hutcheson, Baxter, Telfer, & Warden, 1995; Lamb & Fauchier, 2001; Oates & Shrimpton, 1991).

Focused questions. Prompts for specific details using “wh-” questions (who, when, where, etc.) need to be considered carefully given the known risks associated with them. Ideally, these types of questions are intended to refocus the child ’s attention on topics that have already been mentioned and request specific additional details; for example, “What time did that happen?” Questions of this nature are often answered using only a few words and often contain concepts and words that children have emerging abilities to understand (e.g., times, dates, and numbers representing ages, event frequency, etc.; Evans, Lee, & Lyon, 2009; Walker & Kenniston, 2013; Zajac & Hayne, 2003; Friedman, 1991, 1993; Orbach & Lamb, 2007; Sharman, Powell, & Roberts, 2011; Wandrey, Lyon, Quas, & Friedman, 2012).

Those assessing the quality of interviews need to be mindful that young children’s vocabularies are limited (Dale, 1976; de Villiers & de Villiers, 1999). It is normal for children to use words before they understand their adult meanings, and there is a risk that they will not indicate when they have not understood questions that are asked of them. Moreover, some children may not realize that they did not understand questions and thus fail to see the need to ask interviewers for clarification. Because children sometimes try to answer questions they do not understand (Gee et al., 1999; Waterman, Blades, & Spencer, 2000, 2001, 2004), responses to these types of questions need to be viewed in light of the cognitive abilities of the children concerned. As mentioned above, the ground rules may help with this issue by emphasizing that it is okay to say “I don’t know” and seek clarification.

Some interviewers use focused question as “memory tests,” attempting to determine whether events really happened by identifying contradictions and gaps in knowledge that could be evidence of fabrication. Psychological research is able to explain that contradictory and apparently “missing” information are common when these types of questions are asked, regardless of the truthfulness of the report. When children are asked focused questions, the questions themselves may be enquiring about information that was not encoded in memory.

Sometimes contradictions are provided in response to poorly conceived questions, and thus, questions need to be carefully examined to illuminate apparent contradictions. For example, in a case reported by Jones and Krugman (1986), a 3-year-old child reported that the perpetrator ’s car was black when it was actually orange. In fact, the child may have been accurately describing the color of the upholstery, as opposed to the outside of the car as intended by the questioner. Questions like these often need to be considered carefully to determine whether there is potential for misunderstanding. Importantly, inaccurate answers in response to focused questions are not diagnostic of the accuracy of information elicited freely in response to open prompts.

Option posing and yes/no questions. These types of prompts focus the child ’s attention on details that the child has not previously mentioned, asking the child to select an interviewer-given option, or to answer by saying “Yes” or “No.” This type of question can normally be answered using one word or only a few words. Option-posing questions are sometimes referred to as leading questions. Similar to focused questions, they may also contain concepts and words that children have emerging abilities to understand.

When very specific yes/no questions are asked, there is a risk that children will make acquiescence errors , tending to agree or “go along” with what is being said by answering “yes” when they do not really remember what happened; this tendency increases at long recall delays and is more problematic for younger children (5 years and younger) than older children (Ahern, Lyon, & Quas, 2011; Fivush, Peterson, & Schwarzmueller, 2002; Peterson, Dowden, & Tobin, 1999). The more closed a question is, the riskier it is because there is more potential for misunderstanding to occur or for interviewers to introduce inaccurate information.

Suggestive questions. It is vitally important to assess the impact of suggestive questions in forensic interview s with children, and for experts to be able to identify both subtle and overt influences on children’s responses. By definition, suggestive questions are stated in such a way that the interviewer communicates what response is expected, assumes details that have not already been provided by the child , or introduces information that has not been provided by the child (Ceci & Bruck, 1993). When questioned suggestively, children may go along with the interviewers’ suggestions and so it is important to identify the impact on children’s reports. For example, the child may incorporate words introduced by the interviewer into subsequent narratives, and in some cases children have been shown to provide elaborate descriptions of events that have not actually happened (e.g., Bruck et al., 2002; Ceci, Huffman et al., 1994; Ceci, Loftus et al., 1994; Strange et al., 2003). Alternatively, children may simply disagree with interviewers’ suggestions; thus, it is important to identify the effect that suggestive questions might have had on the interviews as a whole.

Repeated Questions

Another well-studied aspect of focused questions such as those discussed above revolves around the potential effects of repeating questions (for reviews, see Fivush & Schwarzmueller, 1995; Poole & White, 1993). It has been relatively easy to study the effects of repeated questions in laboratory studies because questions are precisely constructed and read to children word for word (e.g., Fivush & Schwarzmueller, 1995). When assessing forensic interview s , questions are rarely repeated verbatim so it is necessary to broaden the definition of “repeated questions” to include all interviewer questions and prompts that refocus children on their previous responses in the same way that repeated questions would. For example, “When did it happen?” followed by, “When did you say it happened?” should be considered repeated questions. In contrast, interviewer prompts that are identical should not always be considered repetitions. For example, if an interviewer asks, “What did he do?” and a child replies, “He didn’t do anything it was my brother,” the next question, “What did he do?” ought not to be considered a repetition for obvious reasons. Ultimately, decisions about repeated questions should focus on the content of interviewer prompts, rather than the exact language used.

In some cases repeated questions are necessary, and the reasons for repeating questions may be clear from examining the interview ; for example, a need to refocus children on their previous responses because initial answers were incomplete or unclear, to summarize and check details about topics already discussed, or after reassuring reluctant witnesses that it is safe to disclose information. However, generally speaking, interviewers are trained not to repeat focused questions in interviews because doing so may pressure children to change their answers and these inconsistencies may reduce their credibility as witnesses (Andrews & Lamb, 2014; Brock, Fisher, & Cutler, 1999; Bruck, Ceci, & Hembrooke, 1998; Gilbert & Fisher, 2006; Poole & Lamb, 1998; Poole & White, 1993). Therefore, it is important to assess whether contradictions within a child ’s statement arose due to question repetition. As a safeguard against inadvertently suggesting that children’s previous responses are incorrect, professional guidelines often recommend that children be told why questions may be repeated in order to minimize the risk that they feel pressured to change their responses.

Episodic vs. Generic Language

Child abuse is often a repeated experience (e.g., Connolly & Read, 2006), but children may be required to describe specific instances of abuse (i.e., particularization) in order for charges to be laid and so that defendants have a fair opportunity to challenge allegations (Guadagno, Powell, & Wright, 2006). When children experience events repeatedly, they create a script, or general representation of what usually happens during these types of events. Scripts can guide memory recall, leading to a general account of the gist of events, rather than a specific account of one instance. Identifying and describing one instance of a repeated event requires source monitoring, which is difficult for young children (see Roberts, 2002, for a review).

If it has been established that the alleged abuse has occurred multiple times, it is recommended that interviewers ask about the first time, the last time, and another time in order to elicit accounts of individual instances. Another strategy that interviewers can use to assist children in describing specific instances is to use episodic language (asking about what happened on specific times) rather than generic language (asking about what usually happens; Brubacher, Malloy, Lamb, & Roberts, 2013; Powell, Roberts, & Guadagno, 2007). Children tend to respond in kind when faced with episodic or generic questions; therefore, using more episodic prompts encourages children to use more episodic language and describe specific episodic memories.

Information About the Initial Disclosure

Forensic investigations begin following children’s statements that cause alarm to the adults in their lives. Children initially disclose their experiences in contexts over which we have no control; prior to a formal investigation , children may often have conversations with adults who ask closed or suggestive questions about the alleged events. This cannot be prevented, so the best strategy is to conduct open interviews to clarify what has happened. Research-based interviewing protocols suggest asking children about the conversations they have had about the abuse , how the allegations were disclosed, and to whom. Asking about the initial disclosure provides more information about the potential abuse and also about whether children have been coached by adults about what to say.

Anatomically Detailed Dolls and Other Props

The use of anatomically detailed dolls was initially thought to be a promising means of facilitating the communication of children’s experiences by allowing them to demonstrate what occurred. However, anatomically detailed dolls have caused significant controversy and their use has been, and continues to be, heavily criticized (e.g., Poole, Bruck, & Pipe, 2011). One issue relates to the possibility that the dolls themselves are inherently suggestive because they inadvertently encourage certain types of play. A second concern is that they might simply encourage “make believe” play that is interpreted as actual experience. Given that there are no specific behaviors that reliably diagnose sexual abuse (Bridges, Faust, & Ahern, 2009) it is widely agreed that the risk of inaccurate conclusions being drawn from observations of children’s interactions with anatomically detailed dolls is too high to support their use in investigations of child abuse. Importantly, research investigating the use of anatomically detailed dolls has shown adverse effects on the accuracy of the information obtained, even within a single interview (Bruck, Ceci, Francoeur, & Renick, 1995; Goodman & Aman, 1990). There is no evidence that the content of children’s interactions or play with dolls provides a reliable diagnostic indicator of abuse. Evidence for the use of anatomically detailed dolls comes largely from practitioners who are not aware of the risks.

Although in theory the value of anatomical dolls seems plausible, it is extremely difficult to assess their use and impact in interviews . An important requirement is that the interactions with the dolls be recorded in minute detail. In practice, this is not typically achieved and many of the interactions are “off camera,” or happen too quickly, which prevents any assurance about what actually happened and the potential for suggestion. There is a risk that interviewers could shape interactions by preempting the actions of the child . The requirement to have a video recording and to be able to determine the sequence of child actions and interviewer utterances makes assessment of this practice very time consuming. Using interview aids such as human figure drawings and props create similar problems when assessing forensic interview s .

Recall Delay

We should not be surprised that very young children can provide clear descriptions of their experiences when they have occurred recently, yet forget those details when questioned months and years later. Researchers have studied the effects of recall delay on children’s memory in great detail. For example, Jones and Pipe (2002) documented the rate at which memory declines over time by asking different groups of 5- and 6-year-old children about a visit to a “friendly pirate” either immediately, 1 day, 1 week, 1 month, or 6 months later. When the results were graphed, it became apparent that forgetting is most rapid soon after the event; as more time passes, the amount of forgetting decreases, until there is very little further forgetting. Children not only remember less over time, but there are also increases in the number of errors in their reports (Bruck et al., 2002; La Rooy, Pipe, & Murray, 2007; Melnyk & Bruck, 2004). Younger children forget event details more quickly than older children do (Brainerd, Reyna, Howe, & Kingma, 1990). Because forgetting is most drastic early on, every attempt should be made to conduct interviews in a timely fashion and not delay interviewing witnesses.

The Issue of Psychometrics

Interview protocols are not subject to traditional psychometric testing; each interview has unique questions and the quality of interviews cannot be assessed using a scale. It would be ideal if there were psychometrically valid tests to determine whether or not children have been abused or whether or not they are telling the truth , but these do not exist. Because researchers do not know what has happened to children who have been interviewed, they cannot tell if children’s responses are “valid” (in the traditional way that the concept of validity is described).

One tool that was initially thought to be potentially useful for this purpose is Criteria-Based Content Analysis (CBCA) (Raskin & Esplin, 1991; Steller & Koehnken, 1989). Lamb et al. (1997) used the CBCA procedure to assess the testimonies of 98 children, some of whom were known to have been describing incidents that were improbable, while others described events for which there was strong corroborating evidence . CBCA scores significantly differentiated between the plausible and implausible accounts, but there was considerable overlap between the scores and the technique was clearly not precise enough to be used in forensic contexts. Lamb et al. (1997) noted that most of the testimonies included few narratives, making it difficult for raters to identify the crucial criteria. In a later study, Hershkowitz, Fisher, Lamb, and Horowitz (2007) showed that investigators assessed credibility more accurately when the children provided more narratives and the interviews had been conducted in accordance with best-practice guidelines, although, as in previous studies, the raters correctly identified plausible statements much more accurately than they identified implausible ones. Though this may be the closest researchers have come to developing a measure for the validity of children’s reports, it is not accurate enough to be used to determine the outcomes of real cases involving children’s interviews.

Therefore, the only way to assess interviews is to focus not on whether the children are telling the truth , but on what the interviewers have done during the interviews, and whether the conditions are right for obtaining accurate statements. There is extensive evidence that interviewers have profound effects on what children say, so experts can examine what interviewers have done to determine if that may have affected children’s statements in any way, positively or negatively.

Conclusion

Determining whether or not abuse happened is the job of the jury and the judge, not an expert witness . The question that expert witnesses can answer is, “Was the interview itself conducted in a way that meets the standards that are agreed upon in the scientific literature?” Experts should assess whether the interviews were conducted appropriately and explain the possible impact of the practices followed so that fact finders can come to reasoned decisions about the evidence . For example, if no ground rules were laid out, no open prompts were used, and suggestive questions abounded, experts need to make clear that the evidence is of poor quality because well-established research-based guidelines were not followed.

High quality interviewing is not an art form; it is a science. The evidence about interview quality is highly consistent and it is important to follow these guidelines in order to increase the ability of interviewers, experts, and fact finders to evaluate the usefulness of the information elicited. Good reports about interview practices can also help to prevent unnecessary court exposure for children when cases are resolved out of court and/or more quickly.

As explained above, there are many interviewing principles that are consistently agreed upon and supported by scientific evidence . If interviews meet these agreed upon criteria, they should be taken seriously because they are likely to yield accurate accounts of what has happened. The expert ’s role is to describe the issues about which there is a consensus and to address whether the interviews in question were conducted in accordance with relevant interviewing guidelines. If child witnesses are interviewed under ideal conditions, we can have reasonable faith in what they have said; their testimony ought to convince the public, and may also convince accused persons that the testimony will be persuasive in court.