Tell Me What Happeneddownload.e-bookshelf.de/download/0000/5795/16/L-G-0000579516... · Tell me what happened : structured investigative interviews of child victims and witnesses

FYX FYX

JWBK185-FM April 10, 2008 19:33 Char Count= 0

Tell Me What HappenedStructured Investigative Interviews of

Child Victims and Witnesses

Michael E. LambUniversity of Cambridge

Irit HershkowitzUniversity of Haifa

Yael OrbachNational Institute of Child Health and

Human Development

Phillip W. EsplinPrivate Practice, Phoenix, Arizona

iii

FYX FYX


xii

FYX FYX


Tell Me What Happened

i

FYX FYX


Wiley Series in

The Psychology of Crime, Policing and Law

Series Editors

Graham Davies and Ray BullUniversity of Leicester, UK

ii

FYX FYX


Tell Me What HappenedStructured Investigative Interviews of

Child Victims and Witnesses

Michael E. LambUniversity of Cambridge

Irit HershkowitzUniversity of Haifa

Yael OrbachNational Institute of Child Health and

Human Development

Phillip W. EsplinPrivate Practice, Phoenix, Arizona

iii

FYX FYX


Copyright C© 2008 John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester,West Sussex PO19 8SQ, England

Telephone (+44) 1243 779777

Email (for orders and customer service enquiries): [email protected] our Home Page on www.wiley.com

All Rights Reserved. No part of this publication may be reproduced, stored in a retrievalsystem or transmitted in any form or by any means, electronic, mechanical, photocopying,recording, scanning or otherwise, except under the terms of the Copyright, Designs andPatents Act 1988 or under the terms of a licence issued by the Copyright Licensing AgencyLtd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission in writing ofthe Publisher. Requests to the Publisher should be addressed to the Permissions Department,John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ,England, or emailed to [email protected], or faxed to (+44) 1243 770620.

Designations used by companies to distinguish their products are often claimed astrademarks. All brand names and product names used in this book are trade names, servicemarks, trademarks or registered trademarks of their respective owners. The Publisher is notassociated with any product or vendor mentioned in this book.

This publication is designed to provide accurate and authoritative information in regard tothe subject matter covered. It is sold on the understanding that the Publisher is not engagedin rendering professional services. If professional advice or other expert assistance isrequired, the services of a competent professional should be sought.

Other Wiley Editorial Offices

John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA

Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA

Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany

John Wiley & Sons Australia Ltd, 42 McDougall Street, Milton, Queensland 4064, Australia

John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore129809

John Wiley & Sons Canada Ltd, 6045 Freemont Blvd, Mississauga, ONT, L5R 4J3, Canada

Wiley also publishes its books in a variety of electronic formats. Some content that appears inprint may not be available in electronic books.

Library of Congress Cataloging-in-Publication Data

Lamb, Michael E., 1953-Tell me what happened : structured investigative interviews of child victims

and witnesses / Michael E.Lamb, Irit Hershkowitz, Yael Orbach.

p. cm.Includes bibliographical references and index.ISBN 978-0-470-51865-6 (cloth : alk. paper) – ISBN 978-0-470-51866-3 (pbk. : alk. paper)

1. Child abuse–Investigation. 2. Interviewing in child abuse. 3. Abused children.4. Child witnesses. I. Hershkowitz, Irit. II. Orbach, Yael. III. Title.

HV8079.C46L36 2008362.76′65–dc22 2008007649

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 978-0-470-51865-6 (hbk) 978-0-470-51866-3 (pbk)

Typeset in 10/12pt Century Schoolbook by Aptara Inc., New Delhi, IndiaPrinted and bound in Great Britain byThis book is printed on acid-free paper responsibly manufactured from sustainable forestry inwhich at least two trees are planted for each one used for paper production.

iv

http://www.wiley.com

FYX FYX


Contents

About the Authors vii

Series Preface ix

Acknowledgements xiii

1 Interviewing Children About Abuse: An Overview andIntroduction 1

2 Factors Affecting the Capacities and Limitations of YoungWitnesses 19

3 How do Investigators Typically Interview Alleged Victims? 63

4 The NICHD Investigative Interview Protocols for YoungVictims and Witnesses 83

5 Does Use of the Protocol Affect the Way Investigators Inter-view Alleged Victims and Witnesses? 103

6 Interviewing Suspected Victims Under Six Years of Age 137

7 The Effects of the Protocol on the Broader InvestigativeProcess 165

8 Interviewing Reluctant Suspected Victims and Suspects 185

9 Interviewing Children with Intellectual and Communica-tive Difficulties 243

FYX FYX


vi Contents

10 Promoting and Maintaining Developmentally-AppropriateInterviewing by Training Interviewers 253

11 What Has Been Achieved: What Else Needs to Be Done? 267

Appendix 1: Investigative Interview Protocol 283

Appendix 2: Focused Questions about Tactile Contact [Touching] 301

Appendix 3: Interview Guide for Youthful Suspects 307

References 317

Index 355

FYX FYX


About the Authors

Michael E. Lamb headed the Section on Social and Emotional De-velopment at the US National Institute of Child Health and HumanDevelopment before becoming Professor of Psychology at the Univer-sity of Cambridge. He received his PhD from Yale University, honorarydegrees from the Universities of Goteborg and East Anglia, and theJames McKeen Cattell Award from the Association for PsychologicalScience for Lifetime Contributions to Applied Psychological Research.He has authored or edited more than 40 books, including InvestigativeInterviews of Children and Child Sexual Abuse: Disclosure, delay, anddenial, as well as about 500 professional publications.

Irit Hershkowitz is a senior lecturer and researcher in the School ofSocial Work at the University of Haifa, Israel from which she receivedher PhD before accepting a post-doctoral fellowship at the US NationalInstitute of Child Health and Human Development. She was involvedthere in development of the Protocol described in this book before re-turning to the University. Her research has been focused on children’sresponses to various investigative strategies, on the value of trainingfor child investigators, on the discrimination of plausible from implau-sible allegations by children, and on the development of techniques forinterviewing reluctant witnesses and those with special needs.

Yael Orbach is an Adjunct Scientist at the National Institute of ChildHealth and Human Development and a Senior Researcher at the Chil-dren’s Studies Program and Center, Brooklyn College, The City Uni-versity of New York. A Fellow of the Association for Psychological Sci-ence, Dr Orbach received her PhD from the University of Natal and hasfocused her research on the application of cognitive and developmen-tal psychological research to criminal investigations involving children.She was on the team that developed the protocol described in this book,has published many professional articles on forensic interviewing of

FYX FYX


viii About the Authors

children, and was co-editor of Child Sexual Abuse: Disclosure, delay,and denial (2007). Among other current studies, Dr Orbach is conduct-ing research funded by the US National Institute of Justice focusing onthe extent to which best practice interviews with child abuse victimsinfluence case outcomes in the justice system.

Phillip W. Esplin is a licensed psychologist in the State of Arizonawho has been in private practice, specialising in forensic psychology,since receiving his Ed D from the Northern Arizona University. For sixyears, he headed a project for the US Bureau of Indian Affairs focusingon evaluation and placement in Special Education, and was a SeniorResearch Consultant to the Child Witness Project at the US NationalInstitute of Child Health and Human Development from 1989 to 2006.He has conducted numerous national and international training semi-nars of proper interview/investigative techniques in child molestationcases, consulted and/or testified in a number of major sexual abusecases in the 1990s, and co-authored many of the articles cited in thisbook.

FYX FYX


Series Preface

The Wiley Series in the Psychology of Crime, Policing and Law pub-lishes both single and multi-authored monographs and edited reviewsof emerging areas of contemporary research. The purpose of this se-ries is not merely to present research findings in a clear and readableform, but also to bring out their implications for both practice and pol-icy. Books in this series are useful not only to psychologists, but also toall those involved in crime detection and prevention, child protection,policing and judicial processes.

One of the significant areas of concern within contemporary foren-sic psychology has been the status of child witnesses. Until recently,psychology and the law have tended to neglect the problems that chil-dren face in first disclosing information concerning physical or sexualabuse and later giving their evidence in court. These problems are attheir most acute within the adversarial system of justice as practicedin the United Kingdom, the USA and many Commonwealth countries,where evidence is collected and assessed through the process of cross-examination: a robust test as much of the witnesses as of the quality oftheir evidence which traditionally makes few concessions to vulnerabil-ity. Earlier volumes in this series (Dent & Flin, 1992; Westcott, Davies,& Bull, 2002) have described research which has led to changes in lawand legal procedures in many countries, designed to facilitate the gain-ing and giving of evidence by children, and this in turn, has increasedthe numbers of such cases coming before the courts. In many instances,this has ensured that adult defendants who would have otherwise es-caped conviction for crimes against children receive the proper sentenceof the court. As with all areas of evidence, however, there are also in-stances where children’s evidence has led to demonstrable miscarriagesof justice or where doubt continues to surround guilty verdicts. In thislatest contribution to the series, Lamb, Hershkowitz, Orbach, and Es-plin argue that the source of error in many such cases lies in poorinvestigative interviewing practices, which have induced in children a

FYX FYX


x Series Preface

belief that events have been experienced or observed when they havenot.

Lamb et al. report the results of analyses of the content of actualforensic interviews conducted by police forces and social workers acrossthree continents which illustrate a general tendency for investigators touse few open-ended questions, but lots of closed and specific questions.They argue that the overuse of the latter style of questions encouragesinterviewers to follow their own hypotheses and look to the childrento confirm their preconceptions, which the children then take over astheir own memories of the event. Greater use of open-ended questions,they argue, would allow children more opportunity to express their ownversion of events while providing the kind of detail which will enablethe courts to reach safer verdicts.

The book describes a new type of interviewing procedure developed byMichael Lamb, the late Kathleen Sternberg, and others at the NationalInstitute for Child Health and Development in the USA. As Lamb et al.emphasise, the NICHD Protocol builds on extensive recent knowledgeof children’s cognitive and social development. It embodies as a centralfeature an emphasis upon open-ended questioning and rigorous train-ing in its use: continually framing questions in an open-ended way is atechnique alien to most everyday discourse between children and adultsthat needs to be laboriously learned, practiced and maintained. Long-term collaborations with researchers and practitioners in four differentcountries have allowed Michael Lamb, Irit Hershkowitz, Yael Orbach,and Phillip Esplin to accumulate a large amount of information con-cerning the superior effectiveness of the NICHD Protocol relative toother widely used interviewing strategies in eliciting extended narra-tives from children, including the very young and those with learn-ing and communicative difficulties, groups who have traditionally beenthought to be developmentally incapable of such demands. The newProtocol promises to produce informationally-rich transcripts, of greatvalue to the courts in their difficult task of discriminating between chil-dren’s true and valid accounts of abuse and the minority of false, oftensincerely held, beliefs.

As the authors emphasise, the endorsement of open-ended question-ing is not new and the potential value of such procedures is widelyappreciated among professional investigators. The problem has alwaysbeen implementation of this principle in practice, which requires con-siderable attention to training and maintaining appropriate interviewpractices through continuous monitoring and feedback. As I have em-phasised elsewhere, there is no agreed investigator training programmein England and Wales covering all police forces, no external valida-tion of training and no system for accrediting and monitoring the

FYX FYX


Series Preface xi

performance of interviewers after training. The position in other coun-tries appears to be no better and frequently much worse (Davies,Marshall, & Robertson, 1998). This book underlines the importance ofa rigorous and systematic training regime in implementing and main-taining effective interviewing practices.

Following a prolonged period at NICHD in the USA, Professor Lambhas returned to the United Kingdom to head the Department of So-cial and Developmental Psychology and the Faculty of Social and Po-litical Sciences at Cambridge University. In addition to his work onchildren’s interview protocols, he continues to publish widely on otherpolicy-related family issues, including the impact of day care, the roleof fathers in children’s development and the impact of early schoolingon children’s social and emotional development. Given his engagementwith societal issues, this book will be of value not just to researchers,practitioners, judges, and lawyers, who are involved in child protectionand deal daily with children’s testimony for the courts, but also to allthose who are interested in the application of psychological theory tocontemporary social problems.

GRAHAM M. DAVIESUniversity of Leicester

REFERENCES

Davies, G. M., Marshall, E., & Robertson, N. (1998) The training needs of officersinvestigating child sexual abuse. Police Research Series, no. 94. London: TheHome Office.

Dent, H. R. & Flin, R. H. (1992). Children as Witnesses. Chichester: Wiley.Westcott, H., Davies, G. M., & Bull, R. (eds.) (2002). Children’s Testimony: A

handbook of psychological research and forensic practice. Chichester: Wiley.

FYX FYX


xii

FYX FYX


Acknowledgements

The research described in this book would not have been possible with-out the assistance of many skilled and dedicated colleagues. We areespecially indebted to Kathleen Sternberg, who played a central rolein the development of the Protocol and the research on its utility. Un-fortunately, Kathy lost a long fight with cancer and died before thework was finished and before this book was conceived or written. InIsrael, Dvora Horowitz headed the Department of Youth Investigationfor several crucial years, making possible many of the field validationstudies described in this book. Her continued and energetic supportwas absolutely critical. In the US, we also benefited from a long andfruitful collaboration with the staff of the Children’s Justice Centrein Salt Lake City. Heather Stewart and Susanne Mitchell played par-ticularly important roles in allowing us to implement and fine-tunethe Protocol in the field and they continue to provide first rate train-ing to interviewers and investigators while collaborating with us onfurther research. Jan Aldridge and Lynn Bowler made it possible toimplement a demonstration project in the UK, while Mireille Cyr andAlain Perron spear-headed a validation study in Quebec. In addition,we have benefited over the last 15 years from the support and helpof many investigators, especially in Israel, the US, and the UK, whohave followed the Protocol in numerous interviews, thereby allowingus to identify its strengths and weaknesses. They are too numerous toname individually, but this work would certainly not have been possi-ble without their participation. Finally, our research has benefited fromthe advice and criticisms of many colleagues, especially Mel Pipe, whojoined our group at NICHD towards the end of this period of researchand provided unparalleled insights into the experimental literature,and Hana Shiloach, who coordinated coding and ensured inter-coderreliability in multiple languages. Many others, including post-doctoralfellows, research assistants, and interns also made invaluable contri-butions which we acknowledge both here, and in the footnotes of themany papers summarised in this book.

FYX FYX


xiv

FYX FYX

JWBK185-01 April 3, 2008 21:1 Char Count= 0

CHAPTER 1

Interviewing ChildrenAbout Abuse: An Overview

and Introduction

Kempe and his colleagues (1962) helped launch scholarly interest inchild abuse with their landmark paper nearly 50 years ago. In suc-ceeding years, professional (and popular) interest shifted from physi-cal to sexual abuse, largely in response to dramatic increases in thenumbers of reported cases, and awareness that many instances ofabuse might go unrecognised because the victims, who were the onlypossible sources of information, seldom gave much information to in-vestigators. As a result, researchers made considerable efforts to un-derstand how children’s testimony can be made as useful and reliableas possible. Since 1990, furthermore, highly publicised cases in theUnited States (California, Massachusetts, New Jersey, North Carolina,and Florida), Norway (Bergen), New Zealand (Christchurch), the UK(Cleveland and Newcastle), and Italy (Rignano Flaminio), among oth-ers, have drawn attention to the counterproductive ways in which al-leged victims of sexual abuse are sometimes interviewed. In many suchcases, inappropriate interview techniques appear to have compromisedand contaminated the children’s testimony, rendering it flawed and un-reliable (Bruck, 1999; Ceci & Bruck, 1995). As explained in this intro-ductory chapter, therefore, the book is designed to: 1) summarise theextant research on children’s memory, communicative skills, and socialtendencies; 2) describe the ways in which that research has been incor-porated into a specific structured interview technique; and 3) reviewresearch involving more than 40 000 alleged victims documenting the

1

FYX FYX


2 Tell Me What Happened

usefulness of that technique. As we explain in some detail, forensic in-terviews with children can be invaluable sources of information, butthey should always be recognised as parts of the forensic investigation,not seen as synonomous with the investigation as a whole.

THE BACKGROUND: INTERVIEWING ANDCHILD DEVELOPMENT

Prompted in part by widespread publicity about the infamous casesjust mentioned, research on children’s capacities to provide reliableand valid information about their past experiences burgeoned in thelast two decades, with many other researchers paying special attentionto children’s suggestibility (see reviews in the last decade by Jones,2003; Lamb, Orbach, Warren, Esplin, & Hershkowitz, 2007; Memon &Bull, 1999; Pipe, Lamb, Orbach, & Esplin, 2004; Poole & Lamb, 1998).Initially, most researchers conducted controlled studies in the labora-tory, but the ecological validity of these studies was often questioned(Doris, 1991; Lamb & Thierry, 2005) so interest in field research wasstimulated too. Later studies conducted in both field and laboratory cir-cumstances focused more narrowly on issues of particular relevance toforensic application and helped generate a remarkable consensus aboutchildren’s limitations and competencies.

In brief, the research reviewed at greater length later in this bookshowed that, although children clearly can remember incidents theyhave experienced, the relationship between age and memory is com-plex, with a variety of factors influencing the quality of informationprovided. For our present purposes, perhaps the most important ofthese factors pertain to the interviewers’ ability to elicit informationand the child’s willingness and ability to express it, rather than thechild’s ability to remember it. Like adults, children can be informativewitnesses, and a variety of professional groups and experts have recog-nised this, offering recommendations regarding the most effective waysof conducting forensic or investigative interviews with children (e.g.,American Professional Society on the Abuse of Children (APSAC), 1990,1997; Jones, 2003; Lamb, 1994; Lamb, Sternberg, & Esplin, 1998; HomeOffice, 1992, 2002; Orbach, Hershkowitz, Lamb, Sternberg, Esplin, &Horowitz, 2000; Poole & Lamb, 1998; Sattler, 1998; Warren & McGough,1996). As Poole and Lamb (1998) pointed out, these books and arti-cles reveal a substantial degree of consensus regarding the ways inwhich investigative interviews should be conducted, and a remarkableconvergence with the conclusions suggested by a close review of theexperimental and empirical literature. Clearly, it is often possible to

FYX FYX


Interviewing Children About Abuse: An Overview and Introduction 3

obtain valuable information from children, but doing so requires care-ful investigative procedures as well as a realistic awareness of their ca-pacities and tendencies. Specifically, accounts elicited using open endedquestions (“Tell me what happened”) that tap recall rather than recog-nition memory are typically more accurate, regardless of the children’sages. The completeness of these initially brief accounts can be increasedwhen interviewers use the information provided by children in theirfirst spontaneous utterance as prompts for further elaboration (e.g.,“You said the man touched you, tell me more about that touching”)(Lamb et al., 2003). Unfortunately, however, forensic interviewers fre-quently ask very specific questions (“Did he touch you?”) that drawupon recognition rather than recall memory. Such questions typicallyelicit less accurate responses than open-ended prompts and may evencause erroneous information to be incorporated into children’s testi-mony. What we have learned about children’s memories and reportingcapacities, as well as the implications for forensic interviewers, are thefocus of the next chapter.

Unfortunately, the research-based and expert-endorsed recommen-dations are widely proclaimed but seldom followed. As discussed morefully in Chapter 3, descriptive studies of forensic interviews in var-ious parts of the United States, United Kingdom, Canada, Sweden,Finland, Norway, and Israel consistently show that forensic interview-ers use open-ended prompts quite rarely, even though such promptsreliably elicit more information than more focused prompts do and areuniversally recommended as the preferred means of eliciting informa-tion from young children (and, indeed, adults, too). To the distress oftrainers, interviewers, and administrators, furthermore, such devia-tions from “best practice” were evident even when the interviewers hadbeen trained extensively, were well-aware of the recommended prac-tices, and often believed that they were adhering to those recommen-dations! Both intensive and brief training programmes for investigativeinterviewers appear to impart knowledge about desirable practices buthave little if any effect on the actual behaviour of forensic investigators.

Because forensic interviewers often have difficulty adhering to recom-mended interview practices in the field, the authors and their colleaguesdeveloped a structured interview Protocol designed to translate profes-sional recommendations into operational guidelines that were first pub-lished as an appendix to a report by Orbach and her colleagues (2000).The structured Protocol featured in this book guides interviewers by il-lustrating techniques designed to maximise the amount and quality ofinformation elicited from alleged victims. As detailed in Chapter 4, theNICHD Protocol (named after the research institute where most of thedevelopers worked and from which they received financial support for

FYX FYX



their work) covers all phases of the investigative interview. In the intro-ductory phase of the interview, the interviewer introduces him/herself,clarifies the child’s task (the need to tell the truth and describe eventsin detail), and explains the ground rules and expectations (i.e., that thechild can and should say “I don’t remember”, “I don’t know”, “I don’tunderstand”, or correct the interviewer when appropriate). In many ju-risdictions, law enforcement agencies requested the inclusion of severalquestions designed to establish that children understood the differencebetween true and false statements.

The rapport-building phase that follows the introductory phase com-prises two sections. The first is designed to create a relaxed, supportiveenvironment for children and to establish rapport between children andinterviewers. In the second section, children are prompted to describe arecently experienced neutral event in detail. This “training” is designedto familiarise children with the open-ended investigative strategies andtechniques used in the substantive phase while demonstrating the spe-cific level of detail expected of them.

In a transitional part between the pre-substantive and the substan-tive phases of the interview, a series of prompts are used to identify thetarget event/s under investigation non-suggestively and with promptsthat are as open as possible. The interviewer only moves on to somecarefully worded and increasingly focused prompts (in sequence) if thechild fails to identify the target event/s.

If the child makes an allegation, the free recall phase begins withan invitation (“Tell me everything.”) and other free-recall prompts orinvitations are recommended. As soon as the first narrative is com-pleted, the interviewer prompts the child to indicate whether the inci-dent occurred “one time or more than one time” and then proceeds tosecure incident-specific information using follow up (“Then what hap-pened.”) and cued invitations (e.g., “Earlier you mentioned a [person/object/action]. Tell me everything about that”) making reference to de-tails mentioned by the child to elicit uncontaminated free-recall ac-counts of the alleged incident/s.

Only after exhaustive free-recall prompting do interviewers proceedto directive questions (focused recall questions that address detailspreviously mentioned by the child and request information withinspecific categories (e.g., time, appearance) such as “When did it hap-pen?” or “What colour was that [mentioned] car?” If crucial detailsare still missing, interviewers then ask limited option-posing ques-tions (mostly yes/no or choice questions referencing new issues thatthe child failed to address previously). Suggestive utterances, whichcommunicate to the child what response is expected, are stronglydiscouraged.

FYX FYX



EVALUATING THE STRUCTURED PROTOCOL

When we developed the structured Protocol, we expected that its im-plementation would improve the organisation and quality of interviewswith children of all ages so that interviewers using the Protocol woulduse more open-ended utterances and fewer option-posing and sugges-tive utterances and would postpone option-posing questions until laterstages of the interview. Because children interviewed using the Proto-col practiced responding to open-ended questions in the pre-substantivephase of the interview, furthermore, we predicted that they would pro-vide absolutely and proportionally more details in response to the firstfree-recall open-ended substantive prompt and more details per open-ended utterances than children interviewed by investigators not guidedby the Protocol. Because interviewers using the Protocol should offermore open-ended prompts, we also predicted that children interviewedin that way would provide absolutely and proportionally more detailsabout the alleged abuse in response to the open-ended questions andfewer in response to option-posing and suggestive questions than chil-dren in comparison groups would.

As discussed in Chapter 5, independent field studies in four differ-ent countries (Orbach, Hershkowitz, Lamb, Sternberg, & Esplin et al.,2000; Sternberg, Lamb, Orbach, Esplin, & Mitchell, 2001; Cyr, Lamb,Pelletier, Leduc, & Perron, 2006; Lamb, Sternberg, et al., 2006) demon-strate convincingly that when forensic investigators employ recom-mended interview procedures by following the structured Protocol, theyenhance the quality of information elicited from alleged victims. Inter-viewers employing the Protocol use at least three times more open-ended and approximately half as many option-posing and suggestiveprompts as they do when exploring comparable incidents, involvingchildren of the same age, without the Protocol. In each study, about halfof the informative and forensically relevant details and more than 80%of the initial disclosures of sexual abuse were provided by preschool-ers in response to free-recall prompts. Such findings suggest that thelikely accuracy of information provided by alleged victims is enhancedwhen interviewers use free-recall prompts exhaustively before turn-ing to more focused prompts. These findings also indicate that cued-invitations should be exhausted before ‘wh’ prompts are introduced be-cause cued-invitations are input-free and thus foster retrieval of free-recall information without limiting responses to investigator-specifiedcategories. Non-suggestive yes/no and choice questions, in which in-terviewers by definition introduce information, should be used onlyif essential information is still missing after free-recall and directiveprompts have been exhausted, because these riskier alternatives are

FYX FYX



more likely to elicit inaccurate information and their introduction maycontaminate subsequent information. When priority was given to open-ended strategies and techniques, there were also significant increasesin the number of facilitators and other supportive comments addressedto child witnesses (Hershkowitz, Orbach, Lamb, Sternberg, & Horowitz,2006); this further enhanced the recall and reporting of information byencouraging children to be more cooperative.

Interviewers using the Protocol also introduce option-posing and sug-gestive questions later in the interview process than do peers not usingthe Protocol. Because option-posing and suggestive questions by defi-nition involve the introduction of information by the investigator, theyhave the potential to contaminate later phases of the child’s report, es-pecially when younger children are involved and thus their delayed util-isation is forensically important. Clearly, forensic interviewers shouldprovide children with opportunities to recall information in response toopen-ended prompts before assuming that more risky interview tech-niques are needed. We have also shown that versions of the Protocol canbe used when interviewing witnesses who are not also victims (Lamb,Sternberg, Orbach, Hershkowitz, & Horowitz, 2003) as well as youthfulsuspects (Hershkowitz, Horowitz, Lamb, Orbach, & Sternberg, 2004).These developments are also discussed in Chapter 5.

The Cognitive Interview (Fisher, Brennan, & McCauley, 2002), whichhas also been popular, especially in the United Kingdom, draws on manyof the same cognitive principles as the NICHD Protocol, and it has beenshown to help interviewers elicit more detailed and accurate informa-tion from children about staged events than ‘standard’ interview proce-dures do (Kohnken, Milne, Memon, & Bull, 1999). Although the Cogni-tive Interview has not been evaluated systematically in the field, somecomponents, like Mental Context Reinstatement, have been shown toenhance the effectiveness of the Protocol (Hershkowitz, Orbach, Lamb,Sternberg, & Horowitz, 2001), and it is possible that other componentsmight be similarly useful.

IS THE PROTOCOL SUITABLE FOR INTERVIEWSWITH YOUNG CHILDREN?

Clearly, as discussed more fully in Chapter 6, there are important dif-ferences between the autobiographical memory retrieval strategies andcapacities of preschoolers and those of older children (Schneider &Bjorklund, 1998). Younger children tend to remember less informa-tion and to provide briefer accounts of their experiences than older

FYX FYX



children do. In addition, young children are more likely than older chil-dren both to respond erroneously to suggestive questions about theirexperiences and to select erroneous options when responding to forced-choice questions. On the other hand, although young children tend toremember less information and provide briefer accounts of their experi-ences than older children do, their reports are no less accurate. Despitethis, some practitioners (e.g., Bourg, Broderick, Flagor, Kelly, Ervin,& Butler, 1999; Hewitt, 1999; Lyon, 1999; Saywitz & Goodman, 1996)have claimed that open-ended questions usually fail to elicit forensi-cally valuable information from young children, especially preschool-ers, even though the inadequacies and capacities of preschoolers hadnot been examined closely in forensic contexts.

We expected that older children would provide more details thanyounger children, but that use of the Protocol would increase theamount of information retrieved by recall from all alleged victims, in-cluding the youngest children. Indeed, because interviewers guided bythe Protocol should use more open-ended prompts regardless of thechildren’s ages, we predicted that use of the Protocol would especiallyenhance the performance of the younger children, ensuring smaller dif-ferences between preschoolers and older children than would otherwisebe the case.

As expected (see Chapter 6), Lamb et al. (2003), found that childrenas young as four years of age can indeed provide substantial amountsof forensically important information about alleged abuse in responseto free-recall prompts. On average, almost one-half of the informationprovided by the children came in response to free-recall prompts, re-gardless of age. Older children reported more details in total and intheir average responses to invitations than the younger children did,but the proportion of details elicited using free-recall prompts did not in-crease with age. Moreover, our study showed that very young childrenare capable of providing most of the information (e.g., time, location,participants) needed by forensic investigators in response to free-recallprompts, thereby reducing reliance on the more risky (potentially con-taminating) yes/no and forced-choice questions. Cued invitations, par-ticularly those that remind children of actions they have previouslymentioned, constitute effective ways of triggering the recall of informa-tion that is more likely to be accurate than information elicited usingrisky forced-choice and yes/no questions from alleged victims as youngas four years of age. Interestingly, action-based cues (e.g., “Tell me moreabout the touching.”) were consistently more effective than all othertypes of cues, regardless of age.

These compelling findings indicated that forensic interviewers needto provide children of all ages with opportunities to recall information

FYX FYX



in response to free-recall prompts before assuming that more risky in-terview techniques are needed, especially because risky questions areeven riskier when addressed to children aged six and under, and thusthat forensic investigators need to make special efforts to maximise theamounts of information elicited from 4- to 6-year-olds using less risky,free-recall prompts.

Because use of the Protocol enhances the quality and informativenessof forensic interviews with alleged victims, it should enhance the valueand conclusiveness of investigations into suspected incidents of sexualabuse by making it easier for interviewers to judge whether victimsare telling the truth (because the children provide more informationin a narrative form which is more amenable to credibility assessment)and by helping investigators to elicit more clues that may guide theirsearch for corroborative evidence. These issues are explored more fullyin Chapter 7.

One relevant study was designed to explore whether the credibilityof children’s statements regarding their alleged experiences of childsexual abuse could be assessed in a more valid and reliable way wheninvestigative interviews were conducted using the Protocol rather thanin an unstructured manner (Hershkowitz, Fisher, Lamb, & Horowitz,2007). In many laboratory analogue studies, children are asked to lieabout events that are not salient or emotionally meaningful, so thegeneralisation of findings to the assessment of credibility in forensiccontexts is obviously problematic, whether or not efforts are made toinclude repeated suggestive questions about body contact, or to avoid in-troducing information not reported by the child. Hershkowitz et al. thusexamined credible and incredible allegations of sexual abuse providedby children in the course of forensic investigations conducted in Israelby the professional youth investigators who have been required since1998 to use the Protocol. Half of the interviews studied were conductedbefore and half were conducted by the same professionals after use ofthe Protocol became mandatory. The cases were individually matchedwith respect to the children’s ages, the types of allegations, and thestrength of the validating evidence.

Forty-two experienced youth investigators each assessed the cred-ibility of allegations of sexual abuse made by alleged victims of sex-ual abuse when interviewed either with or without the Protocol. Halfof the alleged incidents were judged likely to have happened (“plausi-ble”) on the basis of independent evidence, while half were deemed un-likely to have happened (“implausible”). Subsequent analyses showedthat more non-Protocol than Protocol interviews were rated as “Nojudgement possible” rather than as either credible or incredible.Allegations made in Protocol interviews were more accurately rated as

FYX FYX



credible or incredible when they were either plausible or implausible,respectively, than those made in non-Protocol statements. Levels ofinter-rater reliability were also higher when Protocol interviews wererated. Such findings suggested that use of the Protocol facilitated the as-sessment of credibility by child investigators although incredible allega-tions (those describing incidents that were unlikely to have happened)remained difficult to detect, even when the Protocol was used. Again,it is important to recognise that forensic interviews are only part of theoverall investigation, with information provided by child witnesses pro-viding some of the information needed to understand what might havehappened.

INTERVIEWING RELUCTANT ANDNON-COMPLIANT WITNESSES

Most of the published research on forensic interviewing has focusedon interviews with cooperative alleged victims who were ready to dis-close, had often made specific allegations of abuse prior to the formalinvestigation, and were especially responsive to open-ended prompts.However, there is ample evidence that many victims of abuse reportthe abuse belatedly, if at all, with many denying or failing to report theabuse even when they are directly asked or formally interviewed. Theexact numbers cannot be calculated because an unknown number ofvictims never disclose their victimisation and because some proportionof those who initially offer denials and later make allegations may bedoing so falsely, perhaps in response to repeated suggestive questioning.Debate about the relative sizes of the false positive and false negativegroups is intense (London, Bruck, Ceci, & Shuman, 2005; Lyon, 2007),but there is consensus that many abuse victims cannot be protected orhelped because they never disclose their experiences or do so belatedly.In one study, Hershkowitz, Horowitz, and Lamb (2005, 2007) examinedall suspected cases of physical and sexual abuse investigated in thestate of Israel between 1998 and 2002. All investigative interviews wereconducted using a single standardised Protocol, the Protocol discussedin this book. Overall, 65% of the 26 446 children made allegations wheninterviewed, but rates of disclosure were greater in the case of sexual(71%) than physical (61%) abuse. Children of all ages were less likelyto disclose or allege abuse when a parent was the suspected perpetra-tor. Rates of disclosure/allegation increased as children grew older, with50% of the 3- to 6-year-olds, 67% of the 7- to 10-year-olds, and 74% ofthe 11- to 14-year-olds disclosing abuse when questioned.

FYX FYX



A diverse array of factors, including veiled disclosure to non-professionals (e.g., family members and teachers) or to professionals(e.g., medical doctors, CPS workers or police officers), as well as suspi-cions that the child was abused, may trigger formal investigative inter-views with children who are unwilling to disclose. Unlike cooperativeinformants, children who are reluctant to disclose may be less respon-sive to open-ended prompts and may require more guidance and morefocused prompts before making allegations of abuse. As a result, thoseinterviewing them face an inevitable tension between the desire to initi-ate the disclosure of information about what actually happened and theneed to avoid contaminating the memories by suggestively implantinginformation (even prompting false allegations) by using leading andsuggestive prompts. Aiming to minimise the amount of informationprovided by the interviewer, rather than the child, especially duringthe crucial early stages of the interview, recent work has focused onidentifying techniques that might profitably be used when interview-ing reluctant witnesses (Chapter 8).

In another study, Pipe, Sternberg and their colleagues (2007) focusedon the numbers of children who disclosed abuse when formally in-terviewed. The younger children were not only less likely than olderchildren to make allegations when formally interviewed, but they werealso less likely to do so following a prior disclosure. Of course, the priordisclosures were reported by other people, and the reliability of theirsecond hand reports may be questioned, especially when the reporterswere not “disinterested”. It appears, however, that if the person to whomthe child had reportedly made the prior disclosure was an immediatefamily member, presumably those most likely to have a strong inter-est, children were no less (or more) likely to make an allegation in theformal interview.

Although the suspect confessed to the abusive incident(s) in less thana third of all cases, confessions were not always associated with an al-legation. Somewhat surprisingly, several of the older children did notmake an allegation in the interview, when the suspect’s confession hadbeen triggered suspicion in the first place. More detailed examinationshowed, however, that in these cases the abusive incident(s) had oc-curred several years earlier, and/or the nature of the abuse was suchthat the child might not have interpreted it as abuse at the time, asdiscussed by Cederborg, Lamb, and Laurell (2007). Nonetheless, to theextent that suspect confession is corroborative evidence, we can con-clude that there were children in all age groups who had been abused,but did not report the abuse. The reasons for the non-disclosure aremany and varied, and likely to differ developmentally, as a function ofthe nature of the abuse and the circumstances surrounding it.

FYX FYX



In the first field study to explore the dynamics of forensic interviewswith non-disclosing victims (Hershkowitz, Orbach, Lamb, Sternberg,& Horowitz, 2006; Hershkowitz, Orbach et al., 2007), we compared 50children who did not disclose abuse in the course of forensic interviews,despite strong evidence that abuse occurred, with 50 children who dis-closed abuse. Hershkowitz and her colleagues showed that forensic in-terviews which yielded allegations of abuse were characterised by quitedifferent dynamics than interviews with children who seemed equiva-lently likely to have been abused but did not make allegations duringthe interview. When interviewing non-disclosers, interviewers madeless frequent use of free recall prompts and offered fewer supportivecomments than when interviewing children who made allegations ofabuse. Children who did not disclose abuse were somewhat uncoopera-tive, offered fewer details, and gave more uninformative responses, evenat the very beginning of the interview, before the interviewers focusedon substantive issues and before the interviewers themselves began tobehave differently. These findings suggested that premature focus onsubstantive issues may prevent children who are not responsive in theepisodic memory training phase from disclosing abuse. Identifying re-luctant disclosers and making more extensive efforts to build rapportbefore substantive issues are broached, or interviewing such childrenin more than one session, may help suspected victims disclose theirexperiences.

Orbach, Shiloach, and Lamb (2007) also sought to determine whetherthere is a relationship between the type of prompting needed to elicitallegations of abuse and the amount of information disclosed by allegedvictims during investigative interviews. All interviews were conductedby British or American police officers using the Protocol. Non-reluctantdisclosers who made allegations in response to open-ended, free-recall,prompts provided significantly more forensically relevant informationoverall in response to free-recall prompts than a matched group of reluc-tant disclosers who made their initial allegations in response to focused(option-posing or suggestive) prompts. Positive correlations were foundbetween the amount of information provided by children in the pre-substantive and the substantive phases of the interview. The findingsdemonstrated that reluctant witnesses are less communicative even innon-substantive portions of the interview, and continue to be reluctantand provide less information following disclosure.

Hershkowitz, Lanes, and Lamb (2007) focused on the ways in whichchildren disclosed sexual abuse by alleged perpetrators who were notfamily members. Thirty alleged victims of sexual abuse were inter-viewed using the Protocol by six experienced youth investigators. Thesame principles were followed when the parents were asked to describe

FYX FYX



in detail what had happened since the abusive incidents. The state-ments made by the children and parents were then content-analysed.Major characteristics of the children’s and parents’ reported behaviourswere identified by two independent raters. More than half (53%) of thechildren delayed disclosure for between one week and two years, fewerthan half first disclosed to their parents, and over 40% did not dis-close spontaneously but did so only after they were prompted; 50%of the children reported feeling ashamed or afraid of their parents’responses, and their parents indeed tended to blame the children oract angrily. The disclosure process varied depending on the children’sages, the severity and frequency of abuse, the parents’ expected re-actions, the suspects’ identities, and the strategies they had used tofoster secrecy. The children’s willingness to disclose abuse to theirparents promptly and spontaneously thus decreased when they ex-pected negative reactions, especially when the abuse was more serious.A strong correlation between predicted and actual parental reactionssuggested that the children anticipated their parents’ likely reactionsvery well.

Just as special techniques may be needed when interviewing childrenwho are too scared or confused to talk, special techniques may be neededwhen interviewing children and adults with learning, communicative,or intellectual difficulties. Development of these techniques is espe-cially timely because these individuals are at substantially increasedrisk of maltreatment (Crosse, Kaye, & Ratnofsky, 1993; Hershkowitz,Horowitz et al., 2007; Sullivan & Knutson, 2000) and have less access toa criminal justice system that is often insensitive to their capacities andlimitations (Cederborg & Lamb, 2007; Westcott & Jones, 1999). From aconversational perspective, we might expect children with learning dis-abilities to be even more reliant on their adult interlocutors to providestructure and support to enable them to participate than their typi-cally developing counterparts. There have been relatively few studiesthat explore the ability of children with learning disabilities to pro-vide complete and accurate accounts of personally experienced events,however. When interviewed using the kinds of questions advocated fornon-learning disabled children, however, children with learning dis-abilities are able to give reliable accounts of brief witnessed or experi-enced interactions, although their performance relative to chronologi-cally age-matched and mental age-matched controls has varied acrossstudies. The special considerations that need to be addressed by in-vestigators exploring the possible victimisation of children with learn-ing, communicative, and mental difficulties are explored more fully inChapter 9.

FYX FYX



IMPORTANCE OF TRAINING

As mentioned earlier, interviewer training depressingly often yieldsimprovement in trainees’ knowledge but no meaningful changes in theways in which they actually interview alleged victims. Recognising this,training in use of the Protocol has always been accompanied by effortsto provide continued support, guidance, and feedback on interviewer be-haviour in interviews conducted after starting to use the Protocol. Theincremental value of verbal and written feedback during the course oftraining had been experimentally demonstrated previously in individ-ual and group contexts, but only the NICHD training model includesfeedback beyond the training period (i.e., in post training investiga-tive interviews as well). Research on effective training strategies isdiscussed more fully in Chapter 10.

The importance of continuing quality control and feedback was ini-tially assessed by comparing the effectiveness of four different trainingmodels designed to help interviewers implement recommended inter-viewing practices (Lamb, Sternberg, Orbach, Hershkowitz, Horowitz, &Esplin, 2002). In all training conditions, interviewers were first pro-vided with a theoretical framework to help them understand howthe recommended practices were consistent with basic research onchildren’s memorial, linguistic, communicative, and social develop-ment and the performance of the interviewers was compared withthat of the same interviewers conducting interviews with childrenof comparable age and circumstances in the six months prior to thetraining.

Meaningful long-term improvement in the quality of informationobtained from young alleged victims of sexual abuse were observed onlywhen well-established principles were operationalised in a clear andconcrete fashion and when training was distributed over time, ratherthan provided in the form of a single initial session, however inten-sive. Didactic workshops and instruction in the utilisation of highlystructured pre-substantive interview procedures thus had little effecton the number of open-ended prompts used to elicit information or onthe amount of information elicited in this way, whereas intensive train-ing in the use of a highly structured interview Protocol, followed bycontinuing supervision, monthly day-long seminars, and feedback onall field interviews, yielded dramatic improvements on these measuresof the interviews.

In a related study, furthermore, Lamb, Sternberg, Orbach, Esplin,and Mitchell (2002) showed the adverse effects of the terminationof supervision and feedback on investigators’ performance. Forensic

FYX FYX



interviews conducted by trained investigative interviewers who re-ceived close and continuing supervision and intensive individual feed-back were compared with interviews conducted by the same interview-ers in the six months immediately following the completion of trainingand the termination of the supervision-and-feedback. As predicted, thequality of the later interviews was inferior to that of the earlier inter-views, as indexed by: 1) declines in the use of open-ended prompts; 2)corresponding increases in reliance on more focused prompts; and 3)the earlier introduction of focused prompts. The expected changes inthe interviewers’ questioning style were accompanied by decreases inthe amount of information elicited using free-recall prompts.

These reports have important implications for those attempting touse the results of basic research in the real world. Clearly, it is possibleto improve the quality of information elicited from alleged victims ofchild abuse, but these benefits are obtained only when extensive effortsare made not only to train interviewers to adopt recommended prac-tices, but to ensure the maintenance of these practices as well. Regard-less of their skilfulness, interviewers continue to maintain or improvetheir skills only when they regularly review their own and others’ inter-views closely, discussing their strategies, successes and mistakes withother interviewers.

OUTLINE OF THE BOOK

As summarised above and detailed later in the book, intensive sys-tematic research on both children’s suggestibility and their capacitiesto provide reliable and valid information about past experiences hashelped generate a remarkable consensus about children’s limitationsand competencies. In brief, although children clearly can remember in-cidents they have experienced, the relationship between age and mem-ory is complex, with a variety of factors (including the interviewer’sskills) influencing the quality of information provided. Like adults, chil-dren can be informative witnesses, and a variety of professional groupsand experts have offered recommendations regarding the most effectiveways of conducting forensic or investigative interviews with children.The book begins (Chapter 2) with a review of the relevant experimentaland field research underlying the international consensus regarding theways in which investigative interviews should be conducted. Clearly, itis often possible to obtain valuable information from children, but do-ing so requires careful investigative procedures as well as a realisticawareness of their capacities and tendencies.

Documents

Tell Me What Happeneddownload.e-bookshelf.de/download/0000/5795/16/L-G-0000579516... · Tell me what happened : structured investigative interviews of child victims and witnesses