The Nichd Interview Protocol

The NICHD protocol begins with an introductory phase that promotes rapport; includes a "truth and lie ceremony" to communicate to children the importance of truth telling; and establishes the ground rules for the interview, such as the acceptability of saying "I don't know" and correcting the interviewer. Children are then asked for information about themselves, their families, and their schools, using open-ended questions. Following this, children are asked to describe two recent experiences (e.g., Christmas, Passover, or a recent birthday) and are prompted for further elaboration using open-ended prompts, followed by more focused prompts such as time segmentation (e.g., "Tell me everything that happened from the time you went home until you went to the store"; Sternberg et al., 2002, p. 420). The substantive portion of the interview is then introduced, with a series of open-ended prompts used, to encourage the child to discuss the alleged abuse (e.g., "Tell me why you came here today"). None of these prompts refer to the actions, perpetrator, or location of the alleged abuse. The NICHD protocol includes specific questions to be used if a child has not disclosed the abuse in response to more open-ended prompts. The questions were formulated to be minimally leading, should they be required to prompt the child to report the alleged abuse. The NICHD protocol recommends following any direct questions with open-ended prompts for further information ("pairing"), such as "Tell me everything about that," to minimize the risk of subsequent information being contaminated by leading or suggestive questioning. When children indicate that there were multiple instances of abuse, the NICHD protocol recommends eliciting a description of the most recent incident first, followed by the first incident, and then any other specific, well-remembered incidents (e.g., "the time in the bathroom," "the time when you were camping"). Again, prompting for further information involves open-ended prompts, followed by more focused questions if necessary. The interview then finishes with a discussion of a neutral event. One of the distinctive features of the protocol is that it uses children's responses as cues for further information, resulting in a child-directed rather than an interviewer-directed approach throughout all phases of the interview.

Field studies have compared interviews conducted by investigators before and after they have been trained to use the protocol. Interviewers using the protocol adhere better to recommended practices. In a study of 50 interviews conducted by Israeli youth investigators using the protocol, Orbach et al. (2000) found that interviewers offered more than five times as many open-ended invitations as they did in comparable interviews conducted before the structured protocol was introduced. The number of option-posing questions dropped by almost 50% as well, and much more of the information was obtained using free recall rather than investigator-directed recognition probes in the protocol-guided interviews.

Similar results were obtained with investigative interviews conducted by police officers in the western United States (Sternberg et al., 2002; Sternberg, Lamb, Orbach, et al., 2001). The proportion of invitations increased from 10% to approximately one-third, while option-posing and suggestive prompts decreased from 41% to 24%. The total amount of information elicited from free-recall memory also increased dramatically; whereas only 16% of the information was elicited using free-recall prompts in the pre-protocol interviews, about half of the information was obtained using free recall in the protocol interviews. Furthermore, this pattern of results was similar regardless of the children's age. Although younger children provided shorter and less detailed responses than did older children, analyses of interviews with 4- to 6-year-old children revealed that the interviewers relied heavily on invitations (34% of their questions) and succeeded in eliciting a substantial amount of information (49% of the total) using free-recall prompts. Recent research using a British English version of the protocol in England (Lamb et al., 2006) and a French version in Quebec (Cyr, Lamb, Pelletier, Leduc, & Perron, 2006) has confirmed that the protocol leads to comparably dramatic increases in the use and productiveness of open-ended prompts.

In interviews using the protocol, children thus provided substantially more details in response to open-ended invitations. Because these studies involved criminal events, the accuracy of the children's responses could not be established, but research has consistently demonstrated that responses to open-ended questions are more likely to be accurate (Lamb, Sternberg, Orbach, Hershkowitz, & Esplin, 1999). Orbach et al. (2000) showed that almost all of the children who made a disclosure provided a narrative account of the alleged abuse in response to the first invitation. Likewise, Sternberg, Lamb, Orbach, et al. (2001); Lamb et al. (2006); and

Cyr et al. (2006) showed major increases in the amounts of information elicited using invitations. In each case, the total amount of information elicited from free-recall memory increased from about one-seventh of the information in the pre-protocol interviews to about half in the protocol interviews.

There are conflicting views as to whether very young children require more direct interviewing strategies to provide the required level of support and scaffolding for recall and reporting or whether open-ended strategies are sufficient, just as with older children (e.g., Lamb, 1994; Lamb et al., in press; Perry & Wrightsman, 1991). To further clarify the ability of preschoolers to address open-ended questions, Lamb et al. (2003) studied protocol interviews of 130 4- to 8-year-olds and showed that children as young as 4 years were able to report substantial numbers of details about the alleged abuse in response to open-ended invitations. On average, one-half of the information provided by the children came in response to open-ended prompts.

More recently, research using the NICHD protocol has been focused on understanding special groups of witnesses—those who are reluctant or unwilling to disclose, and those who have learning disabilities. Although much attention has been given to the dangers of eliciting false reports of abuse from children, little has been directed toward an equally serious issue—children who have experienced abuse but do not disclose. In Israel, for example, approximately one-third of the suspected child victims interviewed do not disclose abuse during forensic interviews, despite suspicion that abuse might have occurred (Hershkowitz, Horowitz, & Lamb, in press). In a smaller-scale study of interviews conducted using the NICHD protocol, Pipe et al. (in press) reported that 23% of the 294 children interviewed did not make a disclosure, despite sound reasons to suspect that abuse had occurred (e.g., suspect confession). The likelihood of disclosure increased with age and decreased if the suspected perpetrator was a close family member, especially for the younger children in the sample.

Children may be reluctant to disclose abuse for many reasons, including the desire to protect familiar perpetrators, especially family members (Paine & Hansen, 2002; Yuille, Tymofievich, & Marxsen, 1995), or because they have been coerced into secrecy (DeYoung, 1988; Goodman-Brown, 1995; Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003), assume some responsibility or blame (Lyon, 2002; Sjoberg & Lindblad,

2002), feel ashamed or embarrassed (Lyon, 1995; Saywitz et al., 1991), or fear threatened or imagined negative outcomes (Berliner & Conte, 1995; DeYoung, 1988; Paine & Hansen, 2002; Palmer, Brown, Rae-Grant, & Loughlin, 1999). In addition, young victims may not understand that they have been abused and may have failed to encode or remember experiences that did not appear salient to them (Cederborg, Lamb, & Laurel, in press).

Three new studies (Hershkowitz et al., in press; Hershkowitz, Horowitz, & Lamb, in preparation; Orbach, Shiloach, & Lamb, in press) show that children who are reluctant to disclose or who do not disclose at all, despite substantial evidence that abuse occurred, differ from children who are willing to disclose abuse even in the early presubstantive phases of the interview. Hershkowitz et al. (in press; in preparation) examined the transcripts of forensic interviews conducted with 70 children (4 to 13 years old) about whom there was compelling evidence that abuse had indeed occurred. Half of the children disclosed abuse and half did not, with children who did not disclose any abuse being more uncooperative and providing fewer details and more uninformative responses to the interviewers' prompts during the presubstantive rapport-building and episodic-memory practice segments of the interview. During the substantive phase, interviewers posed fewer free-recall prompts and made fewer supportive comments when interviewing children who did not disclose. Hershkowitz et al. concluded that the presubstantive portion of the interview may be important for identifying reluctant witnesses, and that a premature focus on getting an allegation may result in nondisclosure from these children. Similar conclusions were reached by Orbach et al. (in press), who explored the relation between the type of prompt needed to elicit allegations of abuse and the amount of information subsequently disclosed by the alleged victims. The 70 4- to 12-year-old children studied were classified as either nonreluctant disclosers (those who made allegations in response to open free-recall prompts) or reluctant disclosers (those children who did not make an allegation until focused prompts were used). Nonreluctant disclosers reported more forensically relevant information in response to free-recall prompts throughout the interview. Furthermore, the amount of information reported during the presubstan-tive segment of the interview was positively correlated with the amount reported during the substantive segment, indicating that reluctant witnesses are less communicative even during the nonsubstantive segments of the interview and continue to report less information following disclosure. Both studies thus highlight the importance of understanding the motivations and factors that may increase or decrease the likelihood that children will make genuine disclosures of abuse, particularly against the backdrop of statistics suggesting high rates of nonreporting (e.g., Hersh-kowitz et al., in press; London, Bruck, Ceci, & Shuman, 2005; Pipe et al., in press). The identification of such variables may pave the way for the development of strategies to support reluctant or nondisclosing witnesses, and thus enhance our ability to protect these children from continued exposure to abusive situations.

On another front, lawyers, forensic interviewers, child psychiatrists, and psychologists are increasingly calling for research exploring the abilities and vulnerabilities of witnesses with mental retardation and other learning disabilities (e.g., Jones, 2003; Kebbell & Davies, 2003; O'Kelly, Kebbell, Hatton, & Johnson, 2003; Stobbs & Kebbell, 2003). Children with mental retardation are much more likely than their typically developing counterparts to either witness or experience abuse, but they are less likely to have their complaints investigated and/or prosecuted because their capacity to provide reliable evidence is often doubted (Hershkow-itz et al., in preparation; Westcott & Jones, 1999). Hershkowitz et al. (in preparation) examined the extent to which children with mental retardation were vulnerable to different types of abuse, characteristics that may be associated with abuse of these children, and reluctance to disclose as measured by denials of abuse and delayed disclosure by analyzing records of forensic interviews conducted in Israel between 1998 and 2004 (n = 40,430) with children between 3 and 14 years of age who had been classified as slightly disabled (11%), severely disabled (1.2%), or not disabled (87.8%). Children with disabilities (both slight and severe) were more likely than those without disabilities to experience sexual, but not physical, abuse. They were also less likely to disclose, and disclosure was more likely to occur after a delay; they were more likely to experience serious offenses, and the perpetrators were more likely to be family members. Thus, these children were at particular risk of serious abuse within the family but were less likely to make disclosures that would have allowed their complaints to be investigated.

The limited research on eyewitness memory in children with mental retardation suggests that these children are as capable of providing reliable information about their experiences as children matched for developmental level (mental age) when interviewed with optimal techniques (free recall using open questions), although they may perform worse than children of the same chronological ages (Agnew & Powell, 2004; Dent, 1986; Henry & Gudjonsson, 2003; Michel, Gordon, Orn-stein, & Simpson, 2000), with suggestive and/or specific questions being especially problematic (e.g., Agnew & Powell, 2004; Henry & Gudjonsson, 2003). In an ongoing program of research, we are exploring the ability of children with mental retardation (CWMR) to provide meaningful accounts of a personally experienced event when interviewed with the NICHD interview protocol compared to typically developing children matched for both mental age (MA) and chronological age (CA). In addition to examining the memorial and narrative abilities of CWMR and the effectiveness of different levels of interviewer prompting within the NICHD protocol, we are exploring the impact of the severity of mental retardation (mild versus moderate), the effects of delay between the experience and the interview, the effect of repeated interviews, and children's ability to answer suggestive questions that vary in structure (open versus closed) and content (central versus peripheral detail). Preliminary analyses from a subset of the sample (CWMR, n = 17; MA, n = 15; CA, n = 12) interviewed after a short delay (1 week) revealed no significant differences in the completeness of the accounts provided by the three groups of children or in the total amount of correct, incorrect, or ambiguous information reported. CWMR and their MA matches were, however, less accurate than the CA matches. In terms of the amount of support required from the interviewer to elicit the information, some interesting patterns emerged. CWMR required more prompts to orient them to the event of interest than CA matches. When the proportion of all information reported was considered, CWMR and MA children reported significantly less information to free-recall and open prompts and more information in response to focused prompts than CA children. There were no significant differences in the accuracy of children's responses to the suggestive questions, and no differences in the numbers of correct, incorrect or "don't know" answers between the CWMR, CA, and MA groups (Brown, Lewis, Stephens, Lunn, & Lamb, 2006).

In a small study of forensic interviews conducted in Sweden with alleged victims of abuse (6 to 22 years old) who had a learning disability (n = 11), Cederborg and Lamb (in preparation) showed that interviewers relied heavily on focused questions, although children were able to answer open questions when given the opportunity to do so and provided many forensically relevant details. The interviewers spoke more than the children, used questions that were developmentally inappropriate, and did not demonstrate an awareness of the needs and capacities of the children they were interviewing. Such findings are consistent with two recent studies demonstrating that the manner in which lawyers and judges interact with witnesses who have learning disabilities may also limit their participation in court and compromise the likelihood of just verdicts. Cederborg and Lamb (2006) examined court transcripts of Swedish cases where the alleged victims had learning disabilities and found that little or no information was provided to the court regarding the children's disabilities or their possible impact on the children's ability to give testimony. Judges expected them to provide reports much the same as those provided by typically developing children. Similarly, O'Kelly et al. (2003) examined court transcripts involving adults with learning disabilities and found that lawyers asked the same types of questions of them as of nondisabled adults and that judges did not differentially intervene. It is thus clear that research-based guidelines are needed to inform professionals (e.g., social workers, forensic investigators, police, lawyers, judges, and jurors) about the capacities of children with learning disabilities to recall and communicate their experiences and strategies that may enhance or detract from their capabilities.

Was this article helpful?

0 0
Single Parenting Becoming the Best Parent For Your Child

Single Parenting Becoming the Best Parent For Your Child

Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.

Get My Free Ebook


  • Mario
    Are direct questions more productive with all ages of children NICHD protocol?
    6 years ago

Post a comment