Keywords

Introduction

Provided in this chapter are estimates of the numbers of true sexual abuse cases in which children deny abuse, the reasons these children do not disclose, predictors of disclosure failure , and strategies that may facilitate disclosure of sexual abuse. The chapter relies upon both research and practice knowledge. The research is comprised of both quantitative and qualitative studies with a primary focus on field research rather than analogue studies. Much of the knowledge about denial of sexual abuse derives from literature on children and youth who eventually disclose their abuse. In the interest of parsimony, citations will not be exhaustive and will emphasize recent findings. Often, illustrative research, but not the entire body of supportive findings for an assertion or observation, will be presented.

Forensic Interviewers Should Be Aware the Child May Not Have Been Sexually Abused

As awareness of the phenomenon of child sexual abuse , its signs and symptoms, and the contexts in which there is risk for sexual abuse increase, more possible cases are being identified by parents and professionals. In this context, it is important to be ever-mindful that the child may not have been sexually abused. Moreover, advances in electronic access to sexual material render obsolete, or at least limited, some prior hallmarks of sexual abuse, for example, advanced sexual knowledge for the child’s developmental stage and certain sexualized behaviors.

What Can Evidence Tell Us About the Number of Sexually Abused Children Who Do Not Disclose?

The above points having been made, there is nevertheless research that indicates that a substantial minority of children, who have been sexually abused, do not disclose their abuse when forensically interviewed.

Pioneering work by the National Children’s Advocacy Center (NCAC) determined that approximately a fourth of children who were evaluated for child sexual abuse at NCAC did not disclose in a single interview . NCAC staff defined cases which warranted further evaluation as follows: (1) the child did not disclose but there was other compelling evidence of sexual abuse (e.g., medical findings, sexualized behaviors), (2) the child was not able to disclose the full extent of sexual abuse in a single interview, and (3) the allegations were still unresolved after a single interview. Based upon these observations, NCAC undertook several studies involving extended assessments as a method for resolving these cases, studies that will be discussed later in this chapter (Carnes, Wilson, & Nelson-Gardell, 1999, 2000; Carnes, Wilson, Nelson-Gardell, & Orgassa, 2001).

As noted in earlier chapters, the National Institute of Child Health and Human Development (NICHD) protocol is the most extensively researched child forensic interview protocol to date. Early studies focused on optimal strategies for eliciting narrative accounts from children who were willing and able to disclose their sexual abuse (e.g., Sternberg et al., 1997; Sternberg, Lamb, Esplin, & Baradaran, 1999). Disclosing children comprised about two-thirds of the children interviewed using the NICHD protocol (e.g., Sternberg, Lamb, Orbach, Esplin, & Mitchell, 2001). More recently, the research team studying the NICHD protocol has examined interviews of non-disclosing children (see Pipe, Lamb, Orbach, & Cederborg, 2007). Hershkowitz, Horowtiz, and Lamb (2005) report that, despite the demonstrated advantages of the NICHD protocol, about a third of children do not disclose in a single interview . They further examined cases when there was clear evidence that the children were sexually abused (Hershkowitz et al., 2006), again with about a third not reporting sexual abuse. Comparable findings were reported in a New Zealand case study involving four girls (8–15 years old at time of report) and eight adult men. In this study, there were photographs and audiotapes of the sexual abuse. The omission rate for documented sexual acts was of 36.9 % (Bidrose & Goodman, 2000).

Rates of non-disclosure have been found to be higher in specific situations. For example, disclosure rates for children with sexually transmitted diseases have been found to be 42–43 % in two studies (Lawson & Chaffin, 1992; Lyon, 2007). Similarly, an exploratory study involving a single offender who videotaped his abuse of ten victims (one girl and nine boys) found only half of children admitted to some of the videotaped acts when interviewed by police (Cederborg, Lamb, & Laurell, 2007; Sjoberg & Lindblad, 2002). These and other findings indicate that non-disclosing children represent a population whose interview needs warrant professional attention.

What Are the Reasons Children Fail to Disclose Sexual Abuse When Forensically Interviewed?

Both research and practice experience inform knowledge about why children do not disclose sexual abuse . Non-disclosing children with a history of sexual abuse fall into two general categories: (1) Children who do not know to disclose and (2) Children who do not want to disclose (Faller, 2007a).

Children Don’t Know to Tell

There are a variety of reasons why children do not know to disclose sexual abuse . Chief among them that children lack general knowledge about sex and a range of sexual activities, as well as specific knowledge related to sexual abuse.

Children lack sexual knowledge. Young children, in particular, may have no knowledge about sexual behavior (Bussey & Grimbeek, 1995; Cederborg et al., 2007). Even children who have been taught “how babies are made” may lack knowledge about fondling behaviors, oral sex, and anal sex. Because of this, they do not relate sexual abuse to the knowledge they have about conception. Moreover, caretakers may not explain to children the pleasurable and other motivational aspects of sexual behavior. Sexual abuse, therefore, may be perceived as somewhat bewildering activity that does not fit into the child ’s knowledge base.

Children don’t know sexual abuse is wrong. Even if children have sexual knowledge, they may not know that sexual behavior between an older person and a child is wrong (Schaeffer, Leventhal, & Asnes, 2011). For example, Sas and Cunningham (1995), who interviewed 138 children after their cases had been litigated, determined that 30 % of children did not know the abuse was wrong when they first experienced it. Efforts to educate children about “good touch/bad touch” may be inadequate because the sexual behavior is not experienced as touching.

Children don’t understand the expectations for a forensic interview . Children do not know to tell because of the anomaly of a forensic interview (Faller, 2007a, 2007b, 2007c). For children, a forensic interview may be perceived as an encounter with a friendly (or not so friendly) stranger who asks the child open-ended questions which are foreign to the way most adults engage in discourse with children. In children’s everyday encounters with adults, adults do most of the talking and expect short responses from children (Faller, 2007a). Especially for young children, open-ended questions may not trigger free recall of abusive events (Lyon, 2005).

Children don’t perceive the abuse as noteworthy. Children may not know to tell because the abusive event may not be salient to the child . Forensic interviewers may focus on sexual abuse that happened sometime in the past, was disguised as childcare behavior, or occurred in the context of other more upsetting events, for example, living in a crack house. The child’s perspective on saliency is often very different from the forensic interview er’s. As a consequence, open-ended prompts such as “Tell me the reason you came here today,” or even “I understand something may have happened to you, tell me about it from the beginning to the end,” (e.g., Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007) may not trigger the child’s recollection.

Children Don’t Want to Tell

Children who don’t want to disclose pose great challenges to forensic interview ers. There are a number of reasons children are reluctant to disclose sexual abuse .

Children try to avoid being distressed. A typical child response to an upsetting topic is avoidance: avoidance of thinking about the topic and avoidance of talking about it. This response is often found in situations of documented sexual abuse (e.g., Hershkowitz et al., 2006; Leander, 2010).

Children may feel complicit. Children may be reluctant to disclose because they may feel they were complicit. When sexual abuse is discovered, sexually naïve children, who may have enjoyed the attention and/or the physical pleasure, discover that acts they thought were somewhat strange, in fact were very “bad.” As a consequence, they come to believe they are bad for being involved in the abuse. These children do not want to admit that they were “bad.” They are afraid they will be in trouble because of their involvement in the sexual abuse (Hershkowitz, Lanes, & Lamb, 2007). Children who were groomed or who were bribed may regard themselves as participants rather than victims and therefore not disclose (Alaggia, 2004; Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003; Staller & Nelson-Gardell, 2005).

Children feel they will label themselves “damaged goods.” A related reason for not telling is that children may feel stigmatized, embarrassed, or ashamed (Deblinger & Runyon, 2005; Staller & Nelson-Gardell, 2005). Suzanne Sgroi (1982), a pioneer in the child sexual abuse field, described this phenomenon as “damaged goods syndrome,” a common effect of sexual abuse . Older children may be very mindful of the stigma associated with having been a victim of sexual abuse and not want people to know (e.g., Staller & Nelson-Gardell, 2005).

Children are instructed not to talk to strangers. The expectation of professionals that children will disclose sexual abuse to a stranger in the context of a forensic interview is somewhat naïve. Many children are taught to be wary of strangers, which is who a forensic interviewer is. Older children may have had prior experience with the child welfare system and do not want to repeat this experience. In most instances, abused children harbor greater affinity for the offender than for the forensic interviewer.

Offenders manipulate children to keep the secret. Interviewers may fail to fully appreciate the impact on victims of having been admonished not to tell by the offender (Cederborg et al., 2007; Hershkowitz, 2006). Offender admonitions cover a wide spectrum (Malloy, Brubacher, & Lamb, 2011). Schaeffer et al. (2011) added questions about disclosure delays to the RATAC interview protocol (Vieth, 2006) and gathered information from 191 children who made disclosures. These researchers identified nine specific offender threats that were barriers to disclosure. Many researchers note that children may have been threatened with death, bodily hard, or harm to others, including caregivers, siblings, and pets (e.g., Faller, 2007a; Goodman-Brown et al., 2003; Schaeffer et al., 2011). In intrafamilial cases, the offender may have told the child that the family will break up, that he will go to jail, or that he won’t love the child anymore (Faller, 2007a).

Children don’t want to trouble the non-offending caregiver. Children may be influenced not to tell based upon their relationship with a non-offending caregiver. Research supports denial or delay in disclosure because of the lack of support of the non-offending caregiver (e.g., Malloy, Lyon, & Quas, 2007; Olafson & Lederman, 2006). On the other hand, children may fail to disclose because of reluctance to distress the non-offending parent. Schaeffer and colleagues (2011) found that children were fearful that the non-offending parent would be angry, harm the perpetrator , go crazy, be upset, be overwhelmed, be sad, or do something bad.

Children are apprehensive about the future if they tell. Finally, children may fail to disclose sexual abuse because they fear the unknown (Faller, 2007a). Although they are in an abusive situation, at least it is predictable. Even when they have not threatened with consequences if they tell, they simply do not know what might happen to them or to others if they tell (Goodman-Brown et al., 2003; Schaeffer et al., 2011).

Predictors of Non-disclosure

A number of studies have documented predictors of non-disclosure of sexual abuse during forensic interview s .

Gender. One predictor of denial of sexual abuse is gender. Disclosure rates for boys are generally lower than those for girls (e.g., DeVoe & Faller, 1999; Hershkowitz, Horowtiz, & Lamb, 2007; O’Leary & Barber, 2008). Because most offenders are male (e.g., Russell & Bolen, 2000), arguably male victims must overcome two taboos in order to disclose, being involved in sexual abuse and a same-gender sexual encounter (Finkelhor, 1984; Sorsoli, Kia-Keating, & Grossman, 2008). Moreover, the socialization of males to refrain from talking about their problems and vulnerabilities may contribute to their reluctance to disclose.

Proximity of the relationship with the offender. Many studies find that a close relationship between the child and the offender predicts denial of sexual abuse (e.g., Goodman-Brown et al., 2003; Schaeffer et al., 2011). Illustrative are findings from a very large sample study of interviews conducted by Israeli Youth Investigators, masters-level, trained forensic interview ers responsible for interviewing children with allegations of both physical and sexual abuse (Hershkowitz et al., 2005). The researchers report on data from over 25,000 interviews collected over a 5-year period, two-thirds involving parental figures. Although the overall disclosure rate for children alleged to have been sexually abused was 71 % (7812 sexual abuse disclosures), 8 % were of sexual abuse by parental figures and 92 % by non-parental figures. The researchers further examined 373 high certainty (cases with corroborating evidence ) non-disclosing cases. Parents or parent figures were the alleged offenders in 85.5 % of these cases.

Non-supportive, non-offending caregiver. Having a non-supportive, non-offending parent predicts lack of willingness to tell the parent (Elliott & Carnes, 2001; Faller, 1988). Further, if the child does disclose to a parent but is not supported, there is increased risk for disclosure failure in a forensic interview (e.g., Hershkowitz, Lanes & Lamb, 2007; Lawson & Chaffin, 1992; Malloy & Lyon, 2006). Lack of trust of the caretaker (Schönbucher, Maier, Mohler-Kuo, Schnyder, & Landolt, 2012), especially with adolescents, and fear of being blamed are barriers to disclosure (Staller & Nelson-Gardell, 2005). Concern that the caretaker will not believe may delay or prevent disclosure (Alaggia, 2004; Faller, 1988; Schaeffer et al., 2011). If the caretaker is non-supportive or ambivalent (Bolen & Lamb, 2004, 2007a, 2007b), when the sexual abuse is revealed, the caretaker response may lead to recantation (Malloy et al., 2007) or failure to disclose in a forensic interview .

Age of the victim. Findings regarding age as a predictor of disclosure failure vary. Nonetheless, it appears that being very young and being an adolescent both are associated with decreased likelihood of disclosure. In a number of studies, disclosure rates are lower for younger children (e.g., Hershkowitz et al., 2005; Keary & Fitzpatrick, 1994). Lower disclosure rates can be explained by lack of knowledge about sex and sexual abuse , lack of understanding of expectations for a forensic interview , and less developed communication skills. Other research, however, finds lower disclosure rates among older children (e.g., Goodman-Brown et al., 2003; Hershkowitz, Lanes & Lamb, 2007), arguably because they can anticipate negative consequences emanating from telling.

Race and ethnicity. There is general recognition among practitioners that children from non-dominant racial and ethnic groups are potentially less likely to disclose sexual abuse in a forensic interview (e.g., Fontes, 2008; Fontes & Faller, 2007; Paine & Hansen, 2002). In addition, forensic interviewers are usually Caucasian (Williams, Nelson-Gardell, Faller, Cordisco-Steele, & Tishelman, 2014), and their interviewees are disproportionately children of color and increasingly children who are ethnically different from interviewers (Fontes & Faller). Despite the importance of race and ethnicity, there is little research that addresses racial/ethnic barriers and the role of racial and ethnic difference in forensic interviews (e.g., Paine & Hansen). There are some relevant studies, but with inconsistent findings.

Dunkerley and Dalenberg (1999) undertook an analogue study involving 128 children, ages 6–11 years, including 51 Black children and 43 Caucasian children. All children observed a male research assistant of the child ’s race engage in either a positive activity (providing a box of candy) or a negative activity (hiding a purse), both which they were told to keep secret. Children were asked about the secret either by a female graduate student, matched for race or cross race. Only 20 % of African–American children revealed the negative secret when interviewed by a white female, but 60 % did so when interviewed by an African–American female. Findings were comparable for Caucasian children but differences were less extreme.

Faller and Nelson-Gardell (2010) examined predictors of disclosure of sexual abuse , using data from the NCAC multi-site study (to be described in greater detail under strategies that may facilitate disclosure). In this study, 22 interviewers provided data on 137 extended assessments. All but one of the interviewers were Caucasian (one was a Latina). Thirty-five (25 %) of the children were children of color, mostly African–American. In this study, children of color were less likely to disclose sexual abuse.

On the other hand, Springman, Wherry, and Notaro (2006) examined 220 archived forensic interview s for level of disclosure as it related to the race of the forensic interviewer. They reported that African–American children were more likely to disclose sexual abuse to a white interviewer, and white children were more likely to disclose sexual abuse to an African–American interviewer.

To date, there is recognition by practitioners that race matters (e.g., Fontes, 2008) as reflected in both practice guides (e.g., APSAC, 2012) and training of forensic interview ers (e.g., NCAC, 2014). More research is needed, however, to better understand the impact of race and ethnicity on non-disclosure of sexual abuse .

No prior disclosure . Finally, children who have not yet revealed their sexual abuse are less likely to do so in a formal forensic interview (e.g., Keary & Fitzpatrick, 1994; Olafson & Lederman, 2006; Paine & Hansen, 2002). For example, as noted earlier, children whose abuse is first identified by the presence of sexually transmitted disease (Lawson & Chaffin, 1992; Lyon, 2005) and children whose abuse is first identified by audiovisual evidence (Cederborg et al., 2007; Leander, 2010) have low rates of disclosure. Similarly, children thought to have been sexually abused because they exhibit sexualized behavior have lower disclosure rates (e.g., Olafson & Lederman, 2006).

Strategies that May Facilitate Disclosure

A number of strategies may facilitate disclosures from reluctant or non-disclosing children who have been sexually abused. These strategies derive from both research and practice.

Use the revised NICHD protocol. In their highly influential 2011 book, Lamb, La Rooy, Malloy, and Katz provide a revised NICHD protocol aimed at facilitating disclosure in reluctant and non-disclosing children. The NICHD protocol is a linear interview protocol; the revised protocol changes the order of rules and rapport by placing building rapport before providing interview rules. Hallmarks of the revised NICHD protocol are instructing the interviewer to present a friendly, supportive demeanor, to use the child ’s name frequently during the interview, to acknowledge the child’s feelings, but not to interpret them, and to provide non-contingent positive reinforcement. With regard to demeanor, Lamb and colleagues encourage smiling, leaning forward, and making eye contact. Because forensic interview ers are admonished to be neutral, they may be at risk for presenting as cold and unfriendly. Similarly, neutrality could result in failure to acknowledge the child’s feelings. In the revised NICHD protocol, interviewers can acknowledge the child’s feeling related to the interview process. The interviewer can say, for example, “I see you are upset.” The interviewer may also say, “You are really doing a good job,” but take care not to provide this feedback only when the child discloses abuse . Hershkowitz, Lamb, Katz, and Malloy (2013) report on a study in which 199 suspected victims of intrafamilial sexual abuse were randomly assigned to either the revised NICHD protocol or the standard NICHD protocol. The revised NICHD protocol elicited more disclosures (60 % compared to 50 %), disclosures to more open-ended prompts, reduced omissions (e.g., no answer, don’t know, not sure), and fewer denials.

Extend the rapport-building phase of the interview . Extending the rapport-building portion of the interview with children who present as uncooperative in the early stages of the interview can increase the likelihood of disclosure . In a study that matched 50 high certainty, non-disclosing and 50 high certainty, disclosing children, Hershkowitz et al. (2006) report that non-disclosing children communicated less in the rapport-building phase of the interview, which continued into the abuse-related--> portion of the interview. Interviewers used fewer open-ended prompts and fewer supportive comments with these non-communicative children during rapport-building, suggesting an interactional phenomenon. That is, the less the child communicates, the fewer opportunities to offer support and the more the interviewer uses close-ended probes. One recommendation from this study was, when interviewers note a non-communicative pattern, interviewers extend rapport-building rather than moving on to the abuse-related part of the interview. Similarly, Hershkowitz (2011) offers this advice and provides a catalogue of non-suggestive, supportive comments that interviewers might use and examples. These include welcoming the child, expression of personal interest in the child, expression of care about the child’s well-being, checking on the child’s feelings during the interview process, reinforcement that the child is helping the interviewer understand, gestures of goodwill such as “Are you cold?” or “Here is a glass of water,” and thanks at the end of the interview.

Ask the accompanying adult to give the child permission to talk. Interviewers can attempt to preempt reluctance and non-disclosure by asking the caregiver who brings the child to the forensic interview to give the child permission to answer the interviewer’s questions. The adult gives this permission in the presence of the forensic interviewer. This strategy may be especially helpful in cases where the alleged offender is someone close to the child and in cases where the child is concerned about the impact of disclosure on the caregiver (Faller, 2007a).

Explain the expectations of a forensic interview . Because the forensic interview context is anomalous for most children, explaining the process is advised (e.g., Saywitz, Goodman, & Lyon, 2002). Setting the stage for the forensic interview has been shown in analogue research to result in increased free recall (e.g., Dorado & Saywitz, 2001). The interviewer should provide information about his or her role, about the method of recording (notes, audiotape, or videotape), and about the interview rules. A typical rule is a statement that the interviewer will be asking the child lots of questions and, if the child knows the answers, the child should answer the questions; if the child does not, the child should say, “I don’t know.” Additional rules are that if the child doesn’t understand the question, the child should say so, and the interviewer will ask the question in a better way, and the child should talk about what really happened and not about pretend.

In addition, getting the child to promise to tell the truth and reassuring the child he/she will not get in trouble for telling the truth have been found to increase the likelihood of disclosure . Lyon and Dorado (2008) conducted analogue studies with latency-aged children who were under the jurisdiction of the dependency court. Children were interviewed about minor transgressions involving themselves and an adult confederate. Children in both the truth induction and the reassurance conditions were more likely to disclose the transgressions than children in the control condition.

Use facilitative strategies when the child has acknowledged something happened. There are a series of strategies derived from practice that may be useful in situations where the child acknowledges something happened but does not want to talk or says he/she does not remember much.

Focus on the context of abuse first. The interviewer may gather context information first. Contextual details include where the abuse happened, where others were at the time, when it happened, what the child and the offender were wearing, and what, if anything was said during the abusive encounter. When the interviewer has exhausted the child’s report about the context, he/she then asks the child about the sexual acts (what?) and the perpetrator (who?) (Faller, 2007, Chapter 13).

Address reasons for denial. Another strategy, again if the child acknowledges something happened, the interviewer can explore reasons for non-disclosure and then address them. This can include reassurance that the child did nothing wrong and is not in trouble, but should not include promises that everything will be fine if the child discloses, because likely, things will not be fine (Faller, 2007a).

Attempt to motivate disclosure . A related strategy, again in circumstances in which the interviewer has information that the child has, in fact, been abused, is attempting to motivate disclosure. Children may be persuaded to disclose because they want the abuse to stop, they want to protect other vulnerable children, or they want the offender to suffer some consequences (e.g., Schaeffer et al., 2011). A question like “what do you think should happen?” may be appropriate (Faller, 2007a).

Normalize the disclosure process. Sometimes children can be motivated to tell by normalizing the disclosure process. The interviewer can say, “I talk to lots of kids when things have happened to them,” or “This is a safe place for kids to talk.”

Offer other media for disclosure . Finally, when children are reluctant to talk about abuse , the interviewer may suggest other modes of communication. The interviewer may present an anatomical drawing for the child to use to indicate where on the body abuse happened or what body part the offender used (Faller, 2007b). The interviewer may ask the child to draw a picture of the abusive event (Faller, 2007b). Older children may be given the option of writing their responses. In research on the NICHD protocol, both human figure drawings (Aldridge et al., 2004; Teoh, Yang, Lamb, & Larsson, 2010) and asking the child to draw a picture of the abuse (Katz & Hershkowitz, 2010) have been demonstrated to increase disclosures about sexual abuse. Interviewers can also employ dolls, including anatomical dolls to facilitate disclosure (e.g., Faller, 2007b; Goodman, Quas, Batterman-Faunce, Riddelsberger, & Kuhn, 1997; Saywitz, Goodman, Nicholas, & Moan, 1991).

For example, Goodman and colleagues (1997) studied reports of forty-six 3–10-year-old children who had experienced voiding cystourethrogram fluoroscopy (VCUG), an intrusive medical procedure for determining the source of urinary track problems . Most children explicitly revealed genital contact as part of the procedure using anatomical dolls but not in free recall. That said, because the research findings on anatomical dolls are mixed (e.g., Bruck, Ceci, & Francoeur, 2000; Bruck, Ceci, Francoeur, & Renick, 1995), forensic interview ers should be mindful that the use of anatomical dolls may be challenged.

Use externally derived information. If the interviewer has other information that supports sexual abuse , the interviewer can judiciously present this information to the child and ask the child to explain the inconsistency between the child’s denial and the supportive information. Situations in which the child has made a prior disclosure or there is corroborating evidence such as medical findings, a confession, an eye witness , or physical evidence (e.g., video) are possible examples.

Employ more than a single interview . There is increasing support for conducting more than a single forensic interview with children whose sexual abuse allegations cannot be resolved in a single interview (e.g., Hershkowitz et al., 2006; La Rooy, Katz, Malloy, & Lamb, 2010; La Rooy, Lamb, & Pipe, 2009; Patterson & Pipe, 2009). Williams and colleagues conducted a web-based survey of professionals from the lists of the American Society on the Abuse of Children, the National Children’s Advocacy Center, and the National Children’s Alliance, on the need for extended assessments (also called extended forensic evaluations and extended forensic interviews ) (Williams et al., 2014; Williams, Nelson-Gardell, Faller, Tishelman, & Cordisco-Steele, 2013). The 1294 child maltreatment professionals who responded to the survey reported that a not insignificant percentage of their caseloads (mean = 20 %, mode = 10 %; median = 10 %) during the past year could have benefited from an extended assessment. In terms of numbers of cases that could benefit, the respondents indicated a mean of 13, mode of 10, and a median of 6 children. Among the types of cases that could benefit from an extended assessment, 93.1 % of respondents identified “children/adolescents who deny sexual abuse when there is other persuasive evidence (medical indicators, audio or video evidence, offender confession).” There is an emerging body of research that supports both the value of more than a single interview and extended assessments.

Support for more than one interview . There are studies employing the NICHD protocol, with disclosing children who receive a second interview closely following the first (Hershkowitz & Terner, 2007; Katz & Hershkowitz, 2013). These studies found substantial additional information elicited in the second interview and only modest overlap in information from the first and second interviews . For example, Hershkowitz and Terner studied interviews with 30 children (ages 6–13) which occurred a half an hour apart. One-fourth of the details about sexual abuse derived from the second interview, and there was only a 47 % overlap in information from the first and second interviews. Moreover, in the second interview, the interviewers asked more open-ended questions, and the children’s narratives were better organized.

Additional support for multiple interviews is provided by Leander (2010), who examined police interviews with 27 children whose sexual abuse was corroborated by video or photographs. Children were interviewed three times. Although children were avoidant and denied documented abuse in their initial interview , they provided twice as many new details about the abuse in second and third interviews than in the first and fewer denials than in their initial interview. Altogether these children provided 45 sexual details in the first interview, 100 in the second interview, and 103 in the third interview.

Support for extended assessments. Research on extended assessments conducted by the National Children’s Advocacy Center (NCAC) demonstrates that they can resolve a substantial portion of cases when a single forensic interview cannot (Carnes et al., 1999, 2000, 2001). NCAC conducted a pilot study in which children, whose allegations had not been resolved in a single interview , received an eight session extended forensic evaluation (Carnes et al., 1999). Approximately half of these children’s allegations were resolved with a conclusion the children had been sexually abused, about a fifth not sexually abused, and about a fourth still unresolved. The National Children’s Advocacy Center then conducted a multi-site study (20 sites; 147 cases) comparing a four session protocol to an eight session protocol (Carnes et al., 2001). Later analysis of the data from the multi-site study demonstrated that the eight session protocol resulted in 56.6 % of cases being classified as credible disclosures, but the four session protocol only resulted in 29.5 % credible disclosures (Faller & Nelson-Gardell, 2010). Additional examination of disclosures in the eight session condition determined that 95 % of disclosures occurred before the seventh session, suggesting an extended assessment should consist of up to six sessions (Faller & Nelson-Gardell).

An extended assessment allows more time for rapport building and assessing the particular child ’s functioning; it adapts the pace and the structure of the inquiry to the child and gives the child more than a single chance to tell (Faller & Cordisco-Steele, 2014). Thus, it accommodates children who do not know to tell and children who don’t want to tell. There are several models for extended assessments (Faller & Cordisco-Steele; J. N. Anderson, personal communication, May 15, 2014). Nevertheless, there is a need for more detailed articulation of the components of an extended assessment and research to document model efficacy.

Despite the support for extended assessments, there are two legitimate reasons for caution about them. First is expense in a context of scarce forensic interview resources. Even if extended assessments are used judiciously and with a small minority of allegations , Children’s Advocacy Centers do not normally have funds (L. Cordisco-Steele, personal communication, Jan. 26, 2014) and federal child welfare funds are not easily accessible for this type of intervention. The second concern is that interviewers may, by virtue of “not taking no for an answer,” induce children who have not been sexually abused to say that they have been. Research, however, demonstrates that the danger of inducing false positives results from leading and suggestive questioning in the context of multiple interviews , rather than from multiple interviews per se (Lyon, 1999).

Refer the child to abuse -focused therapy. A final strategy to address disclosure failures is abuse-focused treatment. Although non-disclosing children are routinely referred for treatment by child welfare professionals, this treatment has scant coverage in the literature. James, Everson, and Friedrich (n.d.) proposed an abuse-focused treatment model in which the clinician is mindful that the child may have been sexually abused and consciously switches from therapeutic to forensic mode if the child begins to disclose abuse. They advise this model for very young children, children who are reluctant to disclose, and children for whom the likelihood of sexual abuse is uncertain.

Best practice for abuse -focused treatment is for the clinician to use open-ended inquiry about possible abuse, provide opportunities for disclosure by structuring the treatment (e.g., using media or other materials that might trigger disclosure), and to avoid suggestive interpretations of the child ’s statements and behavior in treatment. If the child begins to disclose, the clinician gathers information using open-ended methods. Verbatim documentation of any disclosures and the clinician’s mode of inquiry are important. Depending upon the characteristics of the case (e.g., relationship to the alleged offender, coherence of the child’s disclosure, and safety issues), the clinician may prepare the child and then refer the child for a formal forensic interview .

Conclusions

Research and practice document that non-disclosing children who have likely been sexually abused comprise a significant minority of children and youth who come to professional attention because of sexual abuse allegations . There appear to be two distinct categories of denying children: (1) those who don’t know to tell and (2) those who don’t want to tell. Strategies to address these two categories of children can vary substantially. Research indicates there are factors that inhibit children’s disclosures, including gender, relationship with the offender, having a non-supportive caretaker, age, race/ethnicity, and having not made an outcry before a formal forensic interview . Happily, both clinicians and researchers have begun to turn their attention to denying children and generate strategies that can assist them. Among these are the revised, more child -friendly NICHD protocol, extended rapport-building, practice-based strategies to be used within the interview itself, extended assessments, and abuse-focused treatment.

Nevertheless, there are no foolproof methods for eliciting information about abuse from denying children. Moreover, child abuse professionals must balance concerns about false negatives (children who have been victimized but do not disclose) and false positives (children who, because of interview strategies, falsely confirm sexual victimization).