Studies in two countries find continued support for the use of the NICHD protocol in interviewing young alleged victims of child sexual abuse

Date Published: 
03/10/2011
Source: 

(1) Lamb, M., Sternberg, J., Aldridge, J., Pearson, S., Stewart, H., Esplin, P. & Bowler, L. (2009). Use of a structured investigative protocol enhances the quality of investigative interviews with alleged victims of child sexual abuse in Britain. Applied Cognitive Psychology, 23(4), 449-467.

(2) Cyr, M. & Lamb, M. (2009). Assessing the effectiveness of the NICHD investigative interview protocol when interviewing French-speaking alleged victims of child sexual abuse in Québec. Child Abuse & Neglect, 33(5), 257-268.

Reviewed by: 
Lise Milne
Mireille De La Sablonnière-Griffin
Summary: 

Over the last 20 years researchers have asserted that information retrieved from memory using open-ended, free recall processes is more likely to be accurate than that retrieved using recognition processes such as directive, option-posing and suggestive questions. The National Institute of Child Health and Human Development (NICHD ) structured interview protocol (Orbach et al., 2000) emphasizes free-recall invitations and provides increased opportunities for children to practice responding to open-ended questions in the earlier stages of the interview before delving into the abuse-related portion of the interview. A number of controlled studies have demonstrated the improved quality of investigative interviews carried out with alleged child sexual abuse victims when following the NICHD protocol (see review by Lamb, Orbach, Hershkowitz, Eslpin & Horowitz, 2007).

Two recent studies were undertaken to further assess the effectiveness of the NICHD protocol. In a British study by Lamb, Sternberg, Aldridge, Pearson, Stewart et al. (2009), the NICHD protocol was compared with the widely recognized Memorandum of Good Practice (“standard protocol”) currently in use in Britain. The goal of the study was to determine whether the use of the NICHD protocol would yield further improvements with respect to the amount of quality information from alleged victims of child sexual abuse acquired in interviews. In a 2009 Québec study by Cyr & Lamb, the authors aimed to provide evidence of the applicability of the NICHD protocol to French-speaking children and the replication of findings when interviewers were not trained and supervised by the protocol developers.

Both studies used a group of NICHD protocol guided interviews matched with either a group of non-protocol guided interviews (Québec study) or standard protocol guided interviews (British study). In the Québec study, 83 NICHD protocol guided interviews carried out by police officers and social workers were matched with the same number of non-protocol guided interviews for children aged 3-13 referred to police or child protection services following an allegation of sexual abuse. In the British study, 50 NICHD protocol guided interviews carried out by police officers were matched with a sample of 50 standard protocol guided interviews for alleged child victims of sexual abuse aged 4-13.

Both studies reveal that when the NICHD protocol was used, there were significantly fewer substantive interviewer utterances, approximately three times more open invitations, half the number of directive questions, half the number of risky recognition prompts (option-posing and suggestive prompts), and more recall prompts than the standard interviews. In both studies the NICHD protocol yielded more detailed information from the children: approximately four times the number of details through invitations and approximately half through directive questions.

Findings from both studies reaffirm convincingly the utility of the NICHD protocol for interviewing young alleged victims of sexual abuse. The British study concluded that the NICHD protocol facilitates best-practice interview procedures that lead to better quality information from alleged victims than the oft-touted Memorandum of Good Practice used as the standard protocol. The authors argue that this may be due to the NICHD protocol’s greater specificity, concreteness, emphasis on on-going supervision and feedback, and the pre-substantive part of the interview that allows children to practice responses to open-ended questions. However, authors of both studies assert the importance of extensive training and direct, immediate and ongoing feedback for interviewers using the NICHD protocol.

Methodological notes: 

The two studies were based on a design that matched interviews prior to and after training in the NICHD protocol based on the children’s age, the child-perpetrator relationship, and the type and frequency of abuse. The final sample size was 100 interviews in the British study (n=50 prior; n=50 post; age 4-13) and 166 interviews in the Québec study (n=83 prior; n=83 post; age 3-13).

In the British study, the interviewers were police officers who had less than a year of experience in conducting investigative interviews. In contrast, the Québec study involved police officers and social workers with over 3 years of experience. Both studies had a portion of the interviews coded by two raters, with excellent intra-class coefficient for inter-rater agreement on the classification of utterances (British study: 0.96; Québec study: 0.99).

Limitations of the studies include the short time span between the training and the interviews, the exclusion of interviews that did not follow the NICHD protocol, and the lack of specificity regarding the blindness of the raters in the Québec study.

The Canadian study is considered noteworthy in that it is the first to include interviewers who were not trained and supervised by the developers of the NICHD protocol tool; therefore the authors assert its applicability in other parts of Québec and in other French-speaking countries. Authors of the British study contend that while previous studies demonstrate the value of the NICHD protocol, this study is the strongest test of the NICHD protocol’s ability to improve upon a universally recognized protocol in widespread use at the time.