ADHERENCE TO THE STEP-WISE INTERVIEW PROTOCOL BY TRAINED RCMP INVESTIGATORS DURING FORENSIC SEXUAL ABUSE INTERVIEWS WITH CHILDREN by Maureen G. Hewlett Hon. B.A. , Lakehead University, 1994 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in PSYCHOLOGY ©Maureen G. Hewlett, 1999, 2000 UNIVERSITY OF NORTHERN BRITISH COLUMBIA March 17, 2000 All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author. ii Abstract The purpose of this study was to determine, using descriptive statistics, whether investigators who were trained to use the Step-Wise Interview during forensic interviews with children (a) adhered to the structure of that protocol , (b) used inappropriate interviewing techniques, and (c) used general and specific questions to elicit information. Fifteen videotaped interviews and their associated transcripts were analyzed . Eight RCMP members conducted 12 interviews with children under a-yearsold and 3 with children 8- to 12-years-old. Only initial interviews with children whose cases were closed were included in the sample. Results showed that coverage of the key elements prescribed in the protocol varied between the two age groups and between interview status (i.e., No Disclosure or Disclosure of sexual abuse). The key elements associated with opening the interviews were well covered; however, the rapport building key elements were covered by less than half of the investigators in the No Disclosure interviews, with higher coverage in the Disclosure interviews. Few attempts were made to elicit non-abuse life event free narratives from the younger children , and no such attempts were made with the older children. Both general and specific questions were used when introducing the topic of abuse, but specific questions were used more often with the younger children. When abuse was disclosed, the majority of investigators attempted to elicit an abuse-related narrative. Most investigators used a body diagram, and when there were discussions related to body parts and functioning, all of the investigators used the same terminology as the younger children, and the majority did so with the older children. The key elements associated with closing the interview were poorly covered. With the younger children, the percentage of time spent opening the interviews, building rapport, and discussing body iii parts and functioning was similar, regardless of interview status . In the No Disclosure interviews, more time was spent introducing the topic of abuse. There was time spent asking specific abuse-related questions in both types of interviews, but, as expected, in order to clarify and extend the abuse-related information, there was more time spent asking specific questions in the Disclosure interviews. In both types of interviews, the time spent on topics unrelated to the protocol was greater than the time spent in any of the prescribed interview steps, and the least amount of time was spent closing the interviews. In general, most of the interview steps were introduced in the correct StepWise Interview order. However, some investigators skipped some steps, some performed steps that they should not have, and some steps were introduced out of order. Misleading and leading statements and questions rarely occurred, which is consistent with the goal of eliciting uncontaminated testimony. Regardless of interview status, during the first and second halves of the interviews, investigators probed for information almost as often as they made some other kind of statement. In both types of interviews, very few general questions were used during either half of the interviews, and the specific questions used were primarily of the wh_ and yes/no types. In summary, adherence to the Step-Wise Interview protocol by the trained investigators was good in many respects. However, according to the literature on effective child interviewing, many of the areas in which adherence was low are areas in which adherence seems most crucial. Failure to cover some of the key elements of the protocol, coupled with the high use of specific questions, increases the risk of obtaining inaccurate testimony about alleged sexual abuse events from young children. iv TABLE OF CONTENTS Abstract ........................................................................................................................... ii Table of Contents ........................................................................................................... iv List of Tables ..................................................................................................................vii ... L.IS t 0 f F'1gures ...... .......... .. ........... ............ ....................................................................... VIII Background Literature .................................................. ........ ... .......... .............................. 1 Statement of Problem ........................................................................................... 1 Significance of Problem ........................................................................................ 3 Review of Literature .. ............ .... ........ .. ............................................................................. 4 History ................................................................................................................... 4 Current State of Affairs ......................................................................................... 5 Memory ...................................................................................................... 5 Suggestibility .............................................................................................. 7 Language and communicative abilities ...................................................... 9 Summary .................................................................................................. 10 Effective Child Interviewing ............................................................................................ 10 A Structured Child Interviewing Protocol: The Step-Wise Interview .................... 13 Purpose of This Study ................................................................................................... 17 General Research Question #1: Adherence to the Step-Wise Interview............. 19 General Research Question #2: Use of Inappropriate Techniques ..................... 19 General Research Question #3: Types of Statements and Questions ................ 20 Research Question #1: Adherence to the Step-Wise Interview .......................... 21 Specific question 1(a): Coverage of key elements of the protocol ........... 21 Specific question 1(b): Time in the interview step and on topics .............. 22 Specific question 1(c): Order in which the steps were introduced ............ 24 Research Question #2: Use of Inappropriate Techniques .................................. 25 Specific question 2(a): Misleading segments ........................................... 25 Specific question 2(b): Leading segments ............................................... 25 Research Question #3: Types of Statements and Questions ............................. 27 Specific question 3(a): The use of statements and probes ...................... 27 Specific question 3(b): The use of general and specific questions ........... 28 Specific question 3(c): Types of specific questions used ......................... 30 Method .......................................................................................................................... 33 Participants ...... ............ .. ..................................................................................... 33 Setting of the Interviews ...................................................................................... 37 Procedures ........................................................... ............................... ... ..... ....... 37 Observer training in the Step-Wise Interview ........................................... 37 Informed consent procedures ......... ..... .................................................... 38 v Security and confidentiality ...................................................................... 38 Development of the coding manual and establishing interobserver agreement ...................................................... 38 Observational Coding ......................................................................................... 39 Pass 1....................................................................................................... 40 Pass 2 .... .................................................................................................. 40 Pass 3 ...................................................................................................... 42 Interobserver agreement .......................................................................... 46 Results .......................................................................................................................... 49 lnterobserver Agreement .................................................................................... 49 Pass 2 key elements ................................................................................ 50 Pass 3 turn-by-turn officer segments ...................................................... ; 53 General Research Question #1: Adherence to Step-Wise Interview .................. 53 1(a): Coverage of the Key Elements with Younger and Older Children ... 56 Step 1 ............................................................................................ 56 Step 2 ....................... ..................................................................... 61 Step 3 ............................................................................................ 62 Step 4 ............................................................................................ 64 Step 6 ............................................................................................ 64 Body parts and functions ............................................................... 65 1(b): Time Spent in the Interview Steps and on Topics ............................ 65 Data preparation ............................................................................ 66 Time spent. .................................................................................... 67 1(c) Order in which the Six Steps were Introduced .................................. 69 Data preparation ............................................................................ 70 Order of step introduction .............................................................. 71 Summary of Adherence to the Structure of the Step-Wise Interview ....... 75 General Research Questions #2: Use of Inappropriate Techniques ................... 76 2(a) Misleading segments ........................................................................ 77 2(b) Leading segments ............................................................................ 78 General Research Question #3: Types of Statements and Probes .................... 79 Data preparation ...................................................................................... 79 3(a): Use of statements and probes ......................................................... 80 3(b): Use of general, specific and other questions ................................... 81 3(c): Types of specific questions used ..................................................... 84 Summary of Types of Statements and Questions .................................... 86 Discussion .. ................................................................................................................... 87 Implications ......................................................................................................... 96 Recommendations .............................................................................................. 98 Training and practicum experience .......................................................... 99 Constructive criticism and ongoing feedback ......................................... 100 Refresher courses .................................................................................. 101 Limitations of the Current Study ........................................................................ 102 Suggestions for Future Research ..................................................................... 103 Conclusions ...................................................................................................... 104 vi Reference .................................................................................................................... 106 Appendix A: RCMP Six Step Handout.. ...................................................................... 114 Appendix 8: Officer Consent Letter ............................................................................ 115 Appendix C: Victim Services Telephone Script .......................................................... 118 Appendix D: Parent/Guardian Consent Letter ............................................................ 120 Appendix E: Coding Manual used for this Thesis ....................................................... 122 Appendix F: Pass 3 Variable Coding Flow Chart ........................................................ 157 Appendix G: Officer Segment Variables and Collapsed Categories Flow Chart .............................................................................................. 158 Appendix H: Mean Percent and Standard Deviations for Time Spent in Interview Steps and on Topics ............................................................................... 159 Appendix 1: Standard Deviations for Officer Turn Types ............................................ 160 Appendix J: Standard Deviations for Officer Question Types ..................................... 161 Appendix K: Standard Deviations for Specific Question Types ................................... 162 vii LIST OF TABLES Table 1 The Steps and Goals of the Step-Wise Interview ..................................................... 15 2 Sample Distribution for Interviews by Children's Age, Gender and Interview Disclosure Status by Officer Number and Gender .................................................... 36 3 Percentage Agreement and Kappa Statistics for the Key Elements of the Step-Wise Interview .................................................................................................................... 51 4 Percent Agreement and Cohen's Kappa for Turn-by-Turn Officer Segments in Steps and on Topics ............................................................................................................ 54 5 Percentage of Officers who Covered the Step-Wise Interview Key Elements with Younger and Older Children by Interview Status ...................................................... 57 6 Order in which the Steps of the Step-Wise Interview were Introduced in No Disclosure Interviews with Younger Children ............................................................ 72 7 Order in which the Steps of the Step-Wise Interview were Introduced in Disclosure Interviews with Younger Children .............................................................................. 73 8 Percentage of Officer Turn Types for the First and Second Halves of the interviews with Younger Children by Interview Status ... ... ........................................... ............... 81 9 Percentage of Officer General, Specific and Other Question Types For First and Second Halves of Interviews with Younger Children ................................................. 83 10 Percentages of Officer Specific Question Types For First and Second Halves of Interview with Younger Children by Interview Status ................................................ 85 viii LIST OF FIGURES Figure 1 Percentage of Time Spent in Interview Steps by Disclosure Status (!1 = 12) .............. 68 1 Adherence to the Step-Wise Interview Protocol by Trained RCMP Investigators during Forensic Sexual Abuse Interviews with Children Background of Research Problem Statement of the Problem The National Centre on Child Abuse and Neglect found that reports of sexual abuse of children increased more than 300% between 1980 and 1986 (as cited in Leippe, Manion, & Romanczyk, 1993). Poole and Lamb (1998) reported that in 1993, the 2.9 million reports of child maltreatment in the United States of America (USA) "prompted approximately 1.6 million investigations, most of which were conducted without adequate resources to respond effectively" (p. 13). Thus, as reports of abuse against children increased (Lamb, Sternberg, & Esplin, 1994), so too did concern over the plight of the children who were required to give evidence (Davies, Westcott, & Horan, in press) and the methods used by interviewers to obtain such evidence (Poole & Lamb) . Those phenomena played key roles in promoting the demand for research relevant to child eyewitness testimony and child interviewing techniques (Ceci & Bruck, 1993a, 1995; Faller, 1988; Poole & Lamb). Often the young victim and the perpetrator are the only witnesses to the sexual offence. The credibility of the child's testimony becomes a very important issue because (a) there are no third party witnesses, and (b) physical evidence to support the allegations is seldom available (Poole & Lamb, 1998). When an allegation has been made, the offender may be motivated to misrepresent his or her behavior because of the penalties associated with a criminal conviction. Therefore, it is not surprising that 2 lawyers, judges, and researchers have focused their attention on children's eyewitness memory and testimony (e.g., Ceci & Bruck, 1995; Ceci, Leichtman, & Putnick, 1992; Doris, 1991; Fivush, Hamond, Harsch, Singer, & Wolf, 1991; Fivush & Hudson, 1990; Goodman & Bottoms 1993; Perry & Wrightsman, 1991; Saywitz, 1995; Zaragoza, Graham, Gordon , Hirschman, & Ben-Porath, 1995). Psycholegal researchers have also spent much time and energy studying the strengths and weaknesses of various interviewing techniques as they relate to children (e.g., Ceci & Bruck, 1993a, 1995; Geiselman, Saywitz, & Bornstein, 1993; Lamb et al., 1996; McGough & Warren, 1994; Yuille, Hunter, Joffe, & Zaparniuk, 1993). Over the past few years, however, there has been vigorous debate about the research designs and results of studies on children's eyewitness memory and testimony as it relates to sexual abuse (see Doris, 1991; Goodman & Bottoms 1993; Goodman, Rudy, Bottoms, & Aman, 1990). Researchers have expressed serious concerns about the ecological validity of results obtained from laboratory studies on children's abilities to recall and report information (Ceci, 1991; Goodman & Bottoms; Zaragoza et al., 1995). Goodman et al. (1990) and Yuille (1988) argued that data collected in less ecologically valid studies were limited in their applications to legal contexts. Ceci stated that most studies of children's event recollections had not considered the motivational forces or the forensic context that are part of the aftermath of sexual abuse. As a result of the ongoing debate, research designs are now beginning to take the 'real world' issues surrounding child sexual abuse into account. The knowledge gained from past studies has provided researchers and law enforcement professionals with insights into the complexities involved in maximizing the accuracy and breadth of information that can be obtained from children. It must be remembered, however, that the children who 3 participated in past studies were not recruited from the 'abused population.' Thus, fieldtesting is required to investigate the practicality of conducting interviews with sexually abused child ren based on the currently prescribed methods. Significance of the Problem Police investigators and/or child protection authorities attempt to interview children who allege sexual abuse soon after the initial disclosure of abuse has been made. The sensitive nature of sexual abuse allegations, and the age of the alleged victims, often presents a difficult and emotional environment in which to conduct the interview. The desire to protect the well-being and emotional state of the child combined with a need to elicit accurate and reliable testimony about a sensitive topic often results in conflicting objectives. As Yuille, Marxsen, and Menard (1993) stated, "An investigative interview cannot also be a therapeutic interview. Attempting to combine investigation and therapy is near impossible. Any such hybrid interview tends to be both poor investigation and poor therapy" (p. 15). Several structured child interviewing protocols have been developed in an attempt both to assist interviewers in maintaining an environment that is as stress-free as possible and to elicit the most reliable and accurate testimony possible (see Davies, Marshall, & Robertson, 1998; Geiselman & Fisher, 1988; Geiselman et al., 1984; Yuille, Hunter, et al., 1993). Although these child interviewing techniques vary slightly in their methods, they possess many fundamental similarities (Poole & Lamb, 1998). For example, the various interviewing techniques follow a structured format that begins with rapport building, which leads to the introduction of the topic of abuse, and recommendations are made to obtain free narrative accounts from the children during both of those interview phases. Furthermore, recommendations are made with regards 4 to the use of specific questions; they are to be used to clarify or extend information that the child initially provided in his or her abuse related free narrative. When ending the session, the investigators are to close the interview in a supportive manner and ensure that any questions the child may have about the investigative process are answered. However, researchers have observed and reported that child protection workers (Warren, Woodall, Hunt, & Perry, 1996; Wood, McClure, & Birch, 1996) and trained police detectives (Davies, Westcott, et al. , in press; Geiselman et al., 1993) often do not follow the prescribed protocols when interviewing children. Not following a structured interview protocol that incorporates effective child interviewing techniques can create legal difficulties. For example, if the police charge the alleged perpetrator with sexual abuse, the testimony elicited from a child during the investigative interview may be used as evidence in the courts. If a defense lawyer can demonstrate that the interview methods were inappropriate, or that the information elicited from a child was coerced or influenced in any way, the case could be dismissed. On the other hand, the use of inappropriate questioning could contribute to the child making or confirming a false disclosure of sexual abuse that could have dire consequences for an innocent person . Therefore, reliable and valid interviewing techniques are crucial for protecting the emotional state of any child who has alleged sexual abuse, as well as being necessary for providing accurate and reliable evidence in cases where the allegations are well founded. Review of Literature History Interest in children's competence as eyewitnesses and their ability to provide accurate testimony has historically reflected specific judicial events, the structure of the 5 judicial system, and the general social and legal conditions of the times (Ceci & Bruck, 1993b; Poole & Lamb, 1998). Since early in the 20th century, the reliability of children's testimony has been systematically researched in Europe. However, in the USA, the legal profession historically rejected such research. Thus, in America, little empirical research was conducted on children's ability to provide accurate testimony until late in the 1970s. Since that time, American researchers have made a substantial contribution to the extant literature on children's suggestibility and their eyewitness reports (Ceci & Bruck, 1993b). Current State of Affairs This section summarizes some of the current literature on the roles of memory, suggestibility, and language and communication with regards to child eyewitness testimony. The objective of this review is to outline the types of developmental considerations that affect children's level of competency with regards to providing testimony. Information is provided on various research findings that have contributed to, and/or provided the foundation for, the development of the structured interviewing protocols that are now considered to be effective with children. Memory. Researchers have demonstrated that with age and development, there are changes in both the quality and quantity of information individuals remember. In response to general, open-ended questions (e.g., "Tell me everything that happened at your birthday party") young children (e.g., 11 years of age and under; Davies, Westcott, et al., in press) provide less information about events than older children (e.g., 12 yearsof age and over; Davies, Westcott, et al.). It is therefore often necessary to use specific questions (e.g., "Did you play games at your party?") to obtain details from younger children. The use of specific questions, however, is problematic because such cues 6 often dramatically increase the amount of inaccurate information that young children report (Poole & Lamb, 1998). Young children, relative to older children and adults, are at risk for adding wrong details based on script memories when they are recounting specific events. Anderson (1990) explained that information about the world is organized in the brain by interconnecting items (schemata) which tend to be remembered in conjunction with one another. Within these schemata, familiar and routine events are often organized into "scripts" (Baddeley, 1990). Scripts are representations of "averaged" or "typical" events rather than memories of particular incidents. Both children and adults may add erroneous information based on script memory to accounts of specific events, although the tendency to do so generally declines with age (Nelson, 1986). Children appear to forget at faster rates than adults and this applies to all types of information, including the relevant, irrelevant, central and peripheral details of an event (McGough & Warren , 1994). However, in addition to forgetting information, young children may also simply forget to report information they do have in memory, or not realize they are supposed to report the information. Saywitz and Snyder (1993) showed that 7- to 11-year-olds who received narrative elaboration training prior to recalling a target event had superior memory performance relative to their counterparts who did not receive such training. Because of the impact of age and development on memory recall and rates of forgetting, investigative interviews with children who allege that they have been sexually abused should be conducted as soon as possible after the initial disclosure has been made (Poole & Lamb, 1998). Early interviewing will not only allow the investigator to tap the child's memories while they are fresh, but the process itself may provide an 7 inoculation against further forgetting (Brainerd & Ornstein, 1991 ). General, open-ended questions are recommended for initial queries during investigative interviews because they typically result in the most accurate memory of the witnesses' experience(s) (Poole & Lamb; Yuille, Hunter, et al., 1993). However, Goodman and Saywitz (1994) cautioned that in response to open-ended questions, some preschool-aged children in their studies recalled events other than the one of interest to the interviewer. Thus, after obtaining a free narrative response to an open-ended question, it is important that investigators clarify and extend information provided by young children. Such clarification and extension can be accomplished by asking specific questions. Suggestibility. Quicker and greater forgetting may put children at risk for susceptibility to misleading information. Weak memory traces, whether weakened by time or by insufficient original encoding, may be less resistant to integration with, or overwriting by, post-event information (Brainerd & Ornstein, 1991 ). There are, however, a number of other ways in which children might be suggestible. It is possible that suggested information may merely supplement and/or embellish information already in memory without actually impairing the child's ability to remember originally stored information (Zaragoza, 1991 ). Suggestibility may also be unrelated to memory. For example, because of their desire to please, children could conform or comply with the suggestions provided by adult authority figures; or they may simply trust information provided by adults more than their own memory (Ceci & Bruck, 1995). Nevertheless, it has been demonstrated that adults sometimes elicit false allegations from children when they use misleading questions, suggestive questioning, or apply social pressure to obtain responses (Poole & Lamb, 1998). 8 Laboratory research concerning children's suggestibility has revealed a mixed and confusing picture (Ceci, 1991; Ceci, Ross, & Toglia, 1987; Goodman & Bottoms, 1993; Zaragoza et al., 1995). Goodman and her colleagues reported that 4 and ?-yearold children who either played with a confederate clown, or observed the clown and a youngster at play, were seldom misled by sexually relevant questions such as "Did he kiss you?" and "Did he kiss the other child?" (cited in Goodman & Clarke-Stewart, 1991 ). However, Steller (1991) noted that the suggestive questions in Goodman et al's. studies "were unrealistic in content and had nothing to do with the event observed by the children prior to the interview" (p. 107). Other researchers have shown that preschool children were susceptible to suggestion (Ceci et al., 1987; Doris, 1991 ). For example, Ceci and Bruck (1993a) examined the influence of postevent suggestions on preschool children's reports about a pediatric visit that included an inoculation (see also, Bruck, Ceci, Francoeur, & Barr, 1995). Some of the children were provided with pain-affirming feedback (hurt condition) about their inoculation behavior, whereas others were given pain-denying behavior feedback (no-hurt condition). A third group of children were simply told that the shot was over (neutral condition). Results from phase two of the study, which took place a year after the pediatric visit, showed that the initial behavior feedback, and later experimenter attempts to mislead children about persons who performed various actions during the examination, produced significant suggestibility effects. Poole and Lamb (1998) reviewed the literature on the effects of memory and suggestibility on children's eyewitness reporting and concluded that it is not the case that children's event reports are generally distorted and unreliable, nor is it the case that children cannot be prompted to falsely report 9 events that might be considered abusive. Rather, the quality of children's testimony is a joint product of their cognitive and social maturity, their experiences outside formal interviews, and the interviewing context. (p. 69) Because the interviewing techniques used with children play a key role in the quality and quantity of information obtained from them, it is vital that the appropriate methods are implemented during interview sessions. Not doing so could contribute to the need for multiple interviews with a child to obtain information, and could also increase the likelihood of the child's statement being questioned during the legal proceedings if the alleged perpetrator were charged with the crime. Language and communicative abilities. Young children, relative to older children and adults, have limited and less descriptive vocabularies. Nevertheless, there are large individual differences among children and developmental changes occur rapidly. The challenge confronting investigators is to obtain accounts that are sufficiently rich in descriptive detail to permit an understanding of a child's testimony. The more impoverished the child's language, the greater the likelihood that his or her statements will be misinterpreted and that the child will misinterpret the interviewer's questions and purposes (King & Yuille, 1987; McGough & Warren, 1994; Perry & Wrightsman, 1991; Walker, 1993). Furthermore, the linguistic style of the investigator may influence the accuracy of children's accounts. For example, Walker highlighted three ways in which interviewers can influence children's apparent communicative competence: (a) by using age-inappropriate words and expressions, (b) by constructing syntactically complex sentences, and (c) by being ambiguous. To minimize confusion interviewers should use developmentally appropriate language and encourage children to use their own 10 words when describing or explaining what they have witnessed or experienced (Dent, 1991; Geiselman et al., 1993). Summary. Children's ability to accurately and completely recount past events and their ability to monitor their listener's comprehension and identify misunderstandings are heavily taxed during forensic interviews. Researchers such as Goodman and Saywitz (1994) and Poole and Lamb (1998) have reported that forensic interviewers must be responsive to developmental and individual differences in memory, suggestibility, communicative competence, and socioemotional concerns (e.g., intimidation and/or embarrassment). To enhance the eyewitness performance of children in the forensic context, techniques to bolster children's memory strengths continue to be investigated. Effective Child Interviewing Although many questions about children's eyewitness testimonies have yet to be answered, psychologists , legal professionals, and child-protection specialists have drafted general guidelines and protocols, based on empirical research findings, for developmentally appropriate forensic interviews with children. For example, effective child interviewing guidelines have been developed by the American Professional Society on the Abuse of Children (see Poole & Lamb, 1998) and the Home Office in London (see Davies, Marshall, et. al. , 1998; Poole & Lamb), whereas child interviewing protocols have been developed by independent researchers such as Geiselman and Fisher (1988) and Yuille, Hunter, et al. (1993). The foregoing guidelines and protocols all share two very important goals: (1) to increase children's understanding of the interview process and (2) to emphasize practices that maximize children's accuracy (Poole & Lamb). To accomplish those goals, the guidelines and protocols cited above 11 adhere to the following principles: (a) Children should be inteNiewed as soon as possible after the alleged events have been disclosed or discovered (b) inteNiews should begin with a settling-in period so children can adjust to the inteNiewing environment, build rapport with the inteNiewer, receive instructions about the rules of the inteNiew, and practice being informative, (c) inteNiewers should obtain as much information as possible by initially using general, open-ended questions and when specific probes are required to clarify details, inteNiewers should use questions that provide the most options for responding- specific probes should be followed with general prompts for information, (d) inteNiewers should remain neutral and be open to multiple interpretations of children's statements, and finally, (e) when closing inteNiews, inteNiewers should review and clarify reported information, provide information on how they can be contacted later, and discuss neutral topics that end the session with a supportive tone. Readers are referred to Poole and Lamb for a more complete oveNiew of the goals and principles related to child inteNiewing protocols. In addition to the foregoing guidelines for effective child inteNiewing, Goodman and Saywitz (1994) provided more precise descriptions of inteNiewing techniques that, according to the literature, will assist in bolstering the reliability of eyewitness reports when inteNiewing young children: (1) Questions should be short, grammatical constructions should be simple, and vocabulary should be familiar. (2) Accuracy is facilitated when questions concern salient and meaningful events. (3) Hesitant preschoolers should not be pressured, coerced, or bullied into answering questions by authority figures. (4) Suggestibility may be reduced when inteNiewers are neutral or supportive of children's efforts but do not overly praise or intimidate children into recounting 12 specific content. (5) Interviewers should take an objective, nonjudgmental stance in tone of voice, facial expression, and wording of questions (p. 654). Taken together, the guidelines and protocols reviewed in this section recommend that the initial focus of the interview should be on providing the child with an opportunity to become familiar with the interviewer and the interview process (Wood et al., 1996; Yuille, Hunter, et al., 1993). This process is accomplished through rapport building and orienting the child to the types of responses that are expected. Effective rapport building and orientation includes the use of general, open-ended questions. The use of such techniques assist in establishing an atmosphere of trust and cooperation as well as in encouraging the child to engage freely in communication (Wood et al.). Additionally, it is important that the investigator not mislead the child by misrepresenting the purpose of the interview or by making promises that cannot be kept (Poole & Lamb, 1998; Yuille, Hunter, et al. ; Yuille, Marxsen, et al., 1993). The use of such tactics can interfere with the child's sense of trust with the interviewer and the investigative process. It is highly recommended that before introducing the topic of abuse, interviewers attempt to obtain a narrative about some non-abuse life event (Warren et al., 1996; Wood et al., 1996; Yuille, Hunter, et al., 1993). Doing so provides the investigator with an opportunity to evaluate the child's cognitive developmental level. Once the child's linguistic and cognitive skills have been gauged, the interviewer should use developmentally sensitive and age-appropriate language (McGough & Warren, 1994; Wood et al.; Yuille, Hunter, et al.). During rapport building, it is recommended that the interviewer cover details such as (a) telling the child to relate only what he or she truly remembers about the event(s), (b) telling the child that he or she is not limited to 13 answering questions with "yes" or "no" responses, and (c) discussing the difference between telling the truth and telling a lie. Once the investigator has established a rapport with the child and discussed the expectations for the interview, he or she is to introduce the reason for the interview. Researchers recommend that the topic of abuse be introduced using general, openended questions such as "What happened to cause you to be here to talk to me today?" (Davies, Westcott, et al., in press; Warren et al., 1996; Wood et al., 1996; Yuille, Hunter, et al., 1993). According to Staff Sergeant Roberts (personal communication, February 27, 1996), a body diagram can be used at this point of the interview to either assist the child in providing a disclosure or to help clarify disclosed abuse-related information. However, interviewers must avoid contaminating the interview by making leading statements or by introducing terminology that is unfamiliar to the child (Poole & Lamb, 1998). Rephrasing the child's terminology for body parts or functions could confuse the child and introduce age-inappropriate language into the investigative process (Yuille, Hunter, et al.). Interviewers are to elicit as much information as possible from the children by using general questions and narrative prompts for free recall. Specific questions should be used only when absolutely necessary and with the intent of clarifying or extending information that was provided in narrative reports (McGough & Warren 1994; Wood et al.; Yuille, Hunter, et al.). A Structured Child Interviewing Protocol: The Step-Wise Interview Yuille collaborated with psychologists, social workers, prosecutors, and police in Canada, the USA, the United Kingdom and Germany while developing the standardized procedures prescribed in the Step-Wise Interview protocol. The Step-Wise Interview is the standard protocol for child abuse interviews in most provinces of Canada, several 14 states in the USA, the U.S. Army, and in England and Wales (J. Yuille, personal communication, June 2, 1998). The Step-Wise lnteNiew incorporates key elements that are associated with effective child inteNiewing techniques. It is believed that covering the key elements will assist the inteNiewer in establishing the type of communication dynamics required to minimize both the trauma of the investigation for the child and the contaminating effects of the inteNiew on the child's memory of the event(s) (Yuille Hunter, et al., 1993; Yuille Marxsen, et al., 1993). According to Yuille and his colleagues, following the Step-Wise lnteNiew protocol also maximizes the amount of information obtained from the child about the alleged events(s). The main steps and goals of the Step-Wise lnteNiew protocol, as outlined for training for Royal Canadian Mounted Police (RCMP) in British Columbia, are provided in Table 1. lnteNiewers are trained to start the inteNiew by obtaining legally relevant information such as the child's name, age, and address. Investigators then begin the rapport building phase of the inteNiew, which involves attempts to elicit a non-abuse life event free narrative from the child, and setting the parameters with regards to expectations around testimony. Once a rapport is established, the officer introduces the topic of abuse in an attempt to obtain confirmation or denial of the alleged abuse from the child. If the child discloses that abuse has occurred, the investigator is to obtain a free narrative account of the abuse-related events. When the child's abuse-related narrative is complete, the officer, if necessary, asks specific questions to clarify or extend that information. After the officer has obtained sufficient information from the child, he or she closes the inteNiew and thanks the child for cooperating. During training, officers are taught to identify and avoid the use of leading questions and they 15 Table 1 The Steps and Goals of the Step-Wise Interview as Outlined for RCMP Training Steps Goals 1) Introduction • Cover interview identification procedure • Explain professional role • Relax child • Assess developmental level • Cover truth/lie • Discuss "I don't know" answers • Elicit life event narrative • Focus child's attention on reason for the 2) Rapport Building 3) Introduce Topic (Abuse) interview 4) Free Narrative (Abuse) • Provide context for abuse narrative • Child to provide narrative about abuse (uninterrupted) (table continues) 16 Table 1 (cont'd) Steps Goals 5) Specific Questioning • Clarify/extend abuse-related information • Follow-up on inconsistencies • Answer child's questions • Explain what will happen next • Thank and reassure child 6) Closure Note. Adapted from the RCMP Step-Wise Interview Training Session Handout. Unpublished materials presented at the February 26 and 27, 1996 training session in Chetwynd B.C. Original handout is included as Appendix A. 17 are educated in the use of a sexually inexplicit body diagram and appropriate use of developmentally sensitive terminology for body parts and functions. The body diagram, if introduced during the interview, is often used to assist the child with the initial disclosure of abuse or to clarify abuse-related information after the disclosure has been made. Purpose of this Study The Step-Wise Interview protocol was developed to incorporate the effective child interviewing techniques that are espoused in the extant child eyewitness and testimony literature. However, as noted by Poole and Lamb (1998), " ... few attempts have been made to examine the structure and yield of investigative interviews conducted with children using the Step-Wise [Interview] procedures" (p. 98). Therefore, to add to the literature on the use of the Step-Wise Interview, this study was designed to evaluate whether trained RCMP investigators adhered to the structure of that protocol and used the recommended questioning techniques when they interviewed children in actual cases of alleged sexual abuse. This thesis is intended both to add to the existing knowledge on the use of child interviewing procedures and to make a unique contribution to the sparse research on the use of the Step-Wise Interview in the forensic context. As reported earlier in this thesis, researchers have stated that investigators often do not follow prescribed protocols when interviewing children in sexual abuse cases. To examine whether trained RCMP investigators adhered to the Step-Wise Interview protocol when interviewing children in actual cases of alleged sexual abuse, the investigators' interviewing behaviors were observed and analyzed. The sample consisted of 15 videotaped interviews and their associated transcripts. Although the 18 main focus was on interviews with young children (i.e., 8 years of age and under), 3 interviews with older children (i.e., ages 8- to 12-years) who made a disclosure of sexual abuse were included for partial analysis. The interviews with the older children were included to determine whether the investigators' adherence to the prescribed key elements and steps of the protocol was similar with younger and older children. The Step-Wise Interview does not make a distinction between prescriptions for use with younger and older children. However, in viewing the initial batch of videotaped interviews provided by the RCMP, it was noted that the younger children, relative to their older counterparts, appeared to provide the investigators with more interviewing challenges. Twelve interviews with children who were under 8 years of age were included in all of the analyses. Of those 12 interviews, 5 children did not make a disclosure of sexual abuse and 7 children did disclose that such abuse had occurred. The interviews with the younger children were divided by interview status (i.e., No Disclosure and Disclosure of sexual abuse) to determine whether the investigators' interviewing behaviors were similar in both types of interviews. The division by interview status was also made because some prescriptions of the Step-Wise Interview protocol are relevant only to interviews where children make a disclosure that abuse has occurred. A review of the relevant literature lead to the development of three general research questions for this thesis. Specific research questions, which are refined and more precise questions that address the general questions, were developed while observing the interviewing behaviours of the investigators in the videotapes that were used for observational and coding training during the development of the coding manual. 19 General Research Question #1: Adherence to the Step-Wise Interview Protocol The first general research question was developed to evaluate how well the interviewers in this study followed the overall structure of the Step-Wise Interview protocol. As noted by Poole and Lamb (1998), little research has been conducted on the structure and yield of interviews that follow the procedures set out in the Step-Wise Interview. Thus, the first general research question is "Did the RCMP investigators who were trained to use the Step-Wise Interview when questioning children adhere to the prescribed structure of protocol?" To investigate the first general research question, three specific research questions were posed. Those specific questions are outlined in the research question #1 section following the general questions. General Research Question #2: Use of Inappropriate Interviewing Techniques The second general research question was developed to evaluate whether the trained investigators in this study used techniques that are deemed inappropriate in the child interviewing literature. For example, the child's confidence in the interviewer and the interview process, and the integrity of the investigative process, could be compromised if interviewers misrepresent the intentions of the interview or make promises that cannot be kept (Poole & Lamb, 1998; Yuille, Hunter, et al., 1993). Furthermore, the use of leading questions raises the issue of children's suggestibility and the potential for such questions to contaminate the child's testimony (Ceci & Bruck, 1995, Doris, 1991; Goodman & Bottoms, 1993; Poole & Lamb). The rationale for the development of the second general research question was the need for further knowledge about the strengths and weaknesses of the Step-Wise interview protocol. Thus, the second general research question was stated as "When interviewing children under 8 years of age, did the trained RCMP investigators use inappropriate interviewing 20 techniques when using the Step-Wise Interview protocol?" To evaluate this general research question, two specific research questions were posed. Those specific questions are outlined in research question #2 section following the general questions. General Research Question #3: Types of Statements and Questions Used There is much discussion in the literature on child eyewitness testimony about the effects of question types and questioning techniques on the quality and quantity of information obtained from children (e.g., Ceci & Bruck, 1993a, 1993b, 1995; Davies, Westcott, et al., in press; Doris, 1991; Geiselman et al., 1993; Goodman & Bottoms, 1993; Lamb et al., 1996; McGough & Warren, 1994; Poole & Lamb, 1998; Powell & Thomson, 1994; Sternberg et al., 1996; Warren et al., 1996; Wood et al., 1996; Yuille, Hunter, et al., 1993; Zaragoza et al., 1995). However, the overall strengths and weaknesses of the Step-Wise Interview protocol are poorly understood because of the relative absence of reports on its use (Poole & Lamb). Because the Step-Wise Interview is the standard protocol for child abuse interviews in most Canadian provinces, several states in the USA, and in England and Wales (J. Yuille, personal communication, June 2, 1998), it is apparent that further research on the use of the protocol is necessary. The need for baseline research on the use of the Step-Wise Interview when investigating child sexual abuse and obtaining testimony from young children provided the rationale for the third general research question. For this general research question, the interviews were divided in half because Yuille, Hunter, et al. (1993) prescribe that general questions should be used during Steps 1, 2, 3 and 4 (if applicable) and that specific questions, if necessary, should only be used to clarify and extend abuse-related information after disclosure has been made. Thus, the rationale for the division of the interviews was to evaluate the general 21 questioning techniques in each half of the No Disclosure and Disclosure interviews. The third general research question was stated: "When interviewing children under 8 years of age, what percentage of the trained RCMP investigators' segments, in first and second halves of the interviews, were statements versus questions, and what types of questions were used most often in each half of the interviews?" To investigate the third general research question, three specific research questions were posed and those questions are outlined below under research question #3. As stated, the general research questions arose from the literature on child interviewing techniques. The next research step involved developing coding categories that tapped the relevant dimensions of the interviews. That process involved watching the videotaped interviews and developing and modifying the coding definitions until the definitions captured the relevant aspects of the interviews. Then, building on those definitions, the researcher was able to construct the following specific research questions. Research Question #1: Adherence to the Step-Wise Interview Protocol Specific question 1(a): Coverage of key elements of the protocol. Warren et al. (1996) compared 42 transcripts of sexual abuse interviews conducted by child protective services personnel and found that the interviewers "rarely conducted practice interviews regarding past, neutral events, and rarely informed children that 'I don't know,' 'I don't understand,' and 'I don't remember' are acceptable answers to questions" (p. 231 ). Wood et al. (1996) noted during the rapport building phase of the interviews they reviewed, the interviewers often used "stereotyped" questions that required brief answers (e.g., "What is your favorite school subject?" or "Who lives in your house?"). Warren et al. reported that fewer than half of the interviewers in their sample introduced 22 the topic of abuse using general, open-ended questions that were intended to elicit narrative responses. Wood et al., like Warren et al., reported that interviewers in their sample seldom used general, open-ended questions to introduce the topic of abuse, and furthermore, they rarely used general questions to encourage children to provide narrative accounts once they began to disclose the abuse events. The child interviewing literature recommends the use of general questions early in the session to assist in making the child feel comfortable, to help the interviewer assess the child's developmental level, and to model how questioning will proceed later in the interview once the topic of abuse has been introduced. Findings such as those reported above indicate that interviewers often do not follow recommended child interviewing protocols and, thus, those findings provided the rationale for the first specific research question about adherence to the Step-Wise interview protocol: "Did the RCMP investigators, who were trained in the use of the Step-Wise Interview when questioning children, cover the key elements associated with Steps 1, 2, 3, 4, and 6 of the protocol?" Note: There were no key elements associated with Step 5 of the interview protocol. Specific question 1(b): Time spent in the interview steps and topics. A search for publications specifically addressing the optimal length of time a forensic child interviewing session should last produced few results and the absence of research on the optimal length of rapport sessions was noted by Warren et al. (1996). However, in the 1992 Memorandum of Good Practice for interviewing children (published by the Home Office in London), it was recommended that the duration of such interviews should be less than one hour (cited in Davies, Westcott, et al., in press). During the development of that Memorandum, the inclusion of the 'one hour rule' was bitterly 23 contested and it will likely not be included in the Memorandum once the revisions that are currently underway are completed (G. Davies, personal communication, May 14, 1999). Some researchers have reported the length of time interviews last. For instance, Underwager and Wakefield (1990) reviewed 9 videotaped sexual abuse interviews with children and found that the duration of those interviews ranged from 15 to 50 minutes (M = 36 min). Davies, Westcott, et al. (in press) reported that of the 36 interviews they analyzed, the majority (!1 = 28) lasted under an hour (average time was between 40 and 49 min), and that the minimum and maximum amount of time for the entire sample was 20 and 90 minutes, respectively. Davies, Westcott, et al. also reported that in the majority of the interviews (!1 = 28), less than 10 minutes was spent in the rapport building stage. Warren et al. (1996) found that when interviewers in their sample attempted to build rapport, the average proportion of an interview spent on rapport building was 15.0% (based on total words for interviewer and child). Wood et al. (1996) reported that when interviewers in their sample had difficulty establishing rapport, they usually moved on and spent a great deal of time attempting to get the children to talk about any abuse events that may have taken place. They also pointed out that "in general, interviewers seem to regard rapport building as a formality that must be observed, before getting down to the real business of talking about abuse" (Wood et al., p. 223). Davies, Westcott, et al. stated that further research is required to provide positive guidance on how to best handle rapport building and furthermore, that perhaps for the first time, their results demonstrated that rapport can have a powerful influence on the course of the interview. Because of the lack of knowledge with regards to appropriate rapport building with children during forensic interviews and the amount 24 of time to spend in that phase of the interview, the time spent in the various steps of the interviews observed for this study were recorded. Thus, the rationale for the second specific research question was based on the fact that limited research has been conducted on the time spent in the various stages of the recommended protocols. The question posed was "When interviewing children under 8 years of age, what percentage of time did the trained RCMP investigators spend in each of Steps 1 through 6 of the Step-Wise Interview protocol, what percentage of time was spent discussing body terms, and what percentage of time was spent on topics not related to the specific steps of the protocol?" Specific question 1(c): Order in which the steps were introduced. No empirical research could be located that specifically addressed the order in which interviewers proceeded through each of the various steps of the recommended child interviewing protocols. However, adherence to the order of the steps as prescribed by the StepWise Interview protocol is important because, while each step possesses unique key elements, the key elements of each step provide the foundation for the subsequent steps. For example, rapport building with general questions provides children with the opportunity to practice providing narrative responses and, therefore, that step is necessary because it sets the tone for the type of narrative responding that is expected during the abuse-related step of the interview. The lack of literature on adherence to the prescribed step order of the Step-Wise Interview protocol provided the rationale of the following research question: "When interviewing children under 8 years of age, did the trained RCMP investigators follow the recommended step order of the Step-Wise Interview protocol?" 25 Research Question #2: Use of Inappropriate Interviewing Techniques Specific question 2(a): Misleading segments. Poole and Lamb (1998), Yuille, Hunter, et al. (1993), and Yuille, Marxsen, et al. (1993) asserted that it is inappropriate for investigators to mislead children by misrepresenting the purpose of the interview, or by making promises that cannot be kept. It is obvious that the use of such techniques could play a role in undermining the child's sense of trust in the investigative process and, thus, jeopardize its integrity. However, the empirical research reviewed for this · thesis did not specifically address the extent to which forensic interviewers used such techniques while conducting interviews with children. The lack of knowledge with regards to the extent of use of misleading statements in general, and more specifically, the use of such statements while using the Step-Wise Interview protocol, provided the rationale for the following specific research question: "When interviewing children under 8 years of age, did the trained RCMP investigators misrepresent the intention of the interview or make promises they could not keep?" Specific question 2(b): Leading segments. The literature on suggestibility and the use of leading questions with children is extensive (e.g., Ceci & Bruck, 1993b, 1995; Doris, 1991; Goodman & Bottoms, 1993; Poole & Lamb, 1998; Zaragoza et al., 1995). Although authors debate the degree to which intrinsic and extrinsic factors influence children's susceptibility to suggestive and leading questions (see Ceci & Bruck, 1993b; Poole & Lamb), there is consensus in the effective child interviewing literature that leading questions can contaminate children's testimony and, therefore, should be avoided. However, a review of the literature on suggestibility and questioning style revealed that there is no standardized operational definition for what constitutes a leading question and, thus, definitions do vary among research studies. For example, 26 Lamb and his colleagues (Lamb et al., 1996; Sternberg et al., 1996) explored the use of various investigative utterance types in forensic interviews with children and they defined leading utterances as those which "focus the child's attention on details or aspects of the account that the child has not previously mentioned, but do not imply that a particular response is expected" (Lamb et al., 1996, p. 631; Sternberg et al., p. 443). However, because of children's tendency to acquiesce to yes/no questions, researchers are reporting that yes/no questions can be considered leading. For instance, Poole and Lamb noted that "Even yes-no questions are considered leading by many psychologists, particularly if the child is young or the interviewer does not reiterate the child's right to say 'no'." (p. 147). Because of the variation in operational definitions for what constitutes a leading question, researchers must be cautious when reviewing the literature on the use of such questioning techniques. The findings reported in this section are limited to studies in which questions were considered leading if the operational definition was similar to the one used for this thesis (i.e., segments which contained the answer or a choice of answers, named the suspected offender before the child had done so, contained explicit details of the alleged offence to which the child had not previously referred, or contained the interviewer's assumptions about the alleged abuse events). Lamb et al. (1996) and Sternberg et al. (1996) analyzed the use of leading utterances in forensic interviews conducted with children in Israel and the USA, respectively. Lamb et al. reported that 25.4% of the interviewers' utterances in the Israeli sample (N =22 interviews) were leading, whereas Sternberg et al. reported that approximately 40.0% of utterances in the USA sample (N =45 interviews) were leading (40.8% in interviews where abuse was a single incident and 38.8% when abuse events 27 were multiple). Warren et al. (1996) found that 93.9% of the interviewers in their study (N = 42 interviews) introduced new and potentially leading information (i.e., information the child had not already disclosed). However, in a recent child interviewing study conducted by Davies, Westcott, et al. (in press), only 3.0% of all questions asked by the police officers who were trained to use a structured interview protocol that is similar to the Step-Wise Interview, were judged to be leading (!i = 36 interviews). The rationale for the following specific research question rests on the reportedly contaminating effect of the inappropriate use of leading questions on testimonial statements obtained from young children in cases of alleged sexual abuse. What is not known, however, is whether the use of the Step-Wise Interview protocol can be effective in assisting interviewers to avoid the use of such leading questions. Thus, the research question was stated: "When interviewing children under 8 years of age, did the RCMP investigators, who were trained in the use of the Step-Wise Interview, use leading statements or ask leading questions?" Research Question #3: Types of Statements and Questions Used Specific question 3(a): The use of statements and probes. Because of the relative absence of reported research findings on the use of the Stepwise Interview, the rationale for the first specific question was to obtain overall baseline measures for the use of various types of interviewer statements and probing questions throughout the interview sessions. The specific research question was stated as "When interviewing children under 8 years of age, what percentage of the trained RCMP investigators' segments were plain statements, tag questions, probes for information, repetitions of the child's pervious statement, and acknowledgements?" 28 Specific question 3(b): The use of general, specific and other question types. The child interviewing literature emphasizes the benefits of using general, open-ended questions to obtain free narrative accounts from children (e.g., Davies, Westcott, et al., in press; Poole & Lamb 1998; Yuille, Hunter, et al., 1993). Free narrative reports from children 3-years-old and onward are highly accurate. However, the amount or completeness of the information provided is age sensitive, with younger children providing fewer details (King & Yuille, 1989; Powell & Thomson, 1994). Thus, during rapport building, general, open-ended questions serve dual purposes. They provide the interviewer with the opportunity to evaluate the child's cognitive developmental level, and they provide the child with the opportunity to practice providing narrative reports (Poole & Lamb; Warren et al., 1996; Wood et al., 1996; Yuille, Hunter, et al.). During the abuse-related step of the interview, general, open-ended questions provide the child with the opportunity to disclose the abuse-related information in a narrative and, therefore, the disclosure testimony is likely to be a highly accurate account of the remembered events. In the literature on interviewing children, the terms 'general questions' and 'open-ended questions' are often used interchangeably and refer to questions that encourage multiple-word responses (Poole & Lamb; Warren et al.). In the Step-Wise Interview protocol, Yuille, Hunter, et al. (1993) prescribe the use of specific questions to clarify and extend information or to follow-up on inconsistencies in information that was previously disclosed by the child during the abuse-related free narrative step of the interview session. King and Yuille (1989) caution that "interviewers should avoid specific questioning of children, particularly during the initial phases of the interview" (p. 192). However, operational definitions for what constitutes a specific question vary in the child interviewing literature (Peterson & Briggs, 1997). For 29 example, Warren et al. (1996) considered questions such as "Can you tell me what you remember?" and "Do you know what that is called?" specific yes/no questions because "children often fail to appreciate the underlying intentions of those who ask [such] indirect questions" (p. 237). On the other hand, in their discussion on the hierarchy of child inteNiewing questions, Poole and Lamb (1998) included the indirect question "Do you remember what you were doing when he came over?" (p. 146) in their specific but non-leading category. In this thesis, very general requests for information that were intended to elicit a narrative-type response, whether they were direct (e.g., "Tell me about that.") or indirect (e.g., "Can you tell me about that?"), were considered general questions. Questions were considered specific if they were either direct (i.e., explicit) or indirect (i.e., implicit) wh_, multiple choice, yes/no, or conditional if/then statements (e.g., "If I said your hair was purple, then would I be telling the truth or telling a lie?"). Evaluations of the use of general and specific questions in field studies has shown that child protection workers (Warren et al., 1996; Wood et al., 1996) and trained child abuse investigators (Davies, Westcott, et al., in press) tend to use few general question to elicit information from children and that the use of specific questions tends to be high. For example, Warren et al. reported that "few general, open-ended questions were asked during any portion of the inteNiews" (p. 239), and that of all questions in the abuse-related portions of the inteNiews, 10.5% were general questions. Davies, Westcott, et al. reported that in inteNiews with children who were under 8 years of age, 1.0% of the questions were general (i.e., open-ended), 59.0% were specific yet non-leading, and 40.0% were closed-ended. Thus, it appears that inteNiewers, regardless of whether or not they are trained in the use of the reportedly 30 effective child interviewing techniques, tend to rely on the use of specific questions to obtain information from children. Such reports on the use of general and specific questions, the lack of knowledge on how investigators who were trained in the use the Step-Wise Interview frame questions, and the very precise prescription in the protocol for the use of specific questions only after a disclosure of abuse has been made, provided the rationale for the second specific research question: "When interviewing children under 8 years of age, what percentage of the trained RCMP investigators' questions in the first and second halves of the Step- Wise Interviews were general questions, and what percentage in each half of the interviews were specific questions?" Specific question 3(c): Types of specific questions used. The use of specific questions can be problematic when interviewing children because children's responses to specific questions are much less accurate than their responses to general questions (Ceci & Bruck, 1993b; Goodman & Saywitz, 1994; Poole & Lamb, 1998). However, specific questions are often necessary because young children tend to provide minimal information to general questions (Goodman & Saywitz; Poole & Lamb; Poole & White, 1991 ). There are a number of factors that influence the accuracy of information that children provide in response to specific questions, and the younger the child, the " greater the risk for inaccuracy (Ceci & Bruck, 1993b; Goodman & Saywitz, 1994). Specific questions may focus on details that were never encoded by the child, or on details that the child no longer remembers. However, regardless of whether children have knowledge or memory for the specific event of interest, they may s~ill provide an answer to the question, with young children being more at risk for doing so (Poole & White, 1991 ). The context of the forensic interview may exacerbate the problem of 31 obtaining inaccurate information to specific questions because children may believe that they are required to answer the investigators' questions (Poole & Lamb, 1998). Furthermore, the types of questions that are asked to obtain specific information from children can impact on the likelihood of obtaining accurate and inaccurate responses (Peterson & Bell, 1996; Peterson & Briggs, 1997). Researchers have reported high error rates in children's responses to closed-ended yes/no questions: Peterson and Briggs reported a bias for incorrect 'no' responses, whereas Pool and Lamb reported a bias for incorrect 'yes' responses. In comparison to yes/no questions, children's responses to specific open-ended non-leading questions (e.g.," What was she wearing?") have been shown to produced lower error rates (Peterson & Briggs; Poole & White). Yuille, Hunter, et al. (1993) prescribed that when using the Step-Wise Interview, "interviewer[s] should avoid the use of multiple-choice questions as much as possible" (p. 108). Multiple-choice questions, like yes/no questions, are closed-ended and they are problematic because children may feel they should respond by choosing one of the proposed alternatives (Poole & Lamb). Raskin and Yuille (1989) recommended deleting the presumed answer in multiple-choice questions; for example, having been told by the mother that the alleged event happened in the basement, the question to the child should be framed "Did that happen in the bedroom or the kitchen?" However, there are open-ended alternatives to the foregoing example that would reduce the risk of obtaining inaccurate information (e.g., "Where did that happen?"). The literature on the use of specific questions by child protection workers shows that the interviewers tend to overuse specific questions (Warren et al., 1996), and furthermore, that specific questions are used early in the interview sessions (Wood et 32 al., 1996). In their examination of specific question types, Wood et al. found that many interviewers in their sample began interactions with the children by asking closed-ended yes/no questions such as "Do you like school?" Furthermore, the early parts of many of the interviews resembled what Wood et al. described as a quiz-show, with interviewers asking a series of specific wh_ questions about name, age, birth date, address, names and relationships of family members, teacher's name and so on. Warren et al. evaluated the use of yes/no and wh_ questions in interviews with 21 children who were 6 years of age and under (M = 3.8 year), and in 18 interviews with children who were 7 year-of-age and over (M = 8.6 years). Results for the proportions of wh_ and yes/no questions types were reported for overall interviews with each group, as well as for question type during the abuse-related portions of the interviews. In interviews with the younger children, the overall use of wh_ and yes/no questions were 36.7% and 60.3%, respectively. The overall proportions for wh_ and yes/no questions with the older children were 30.0% and 66.5%, respectively. Within the abuse-related portions of the interviews with young children, the proportions of wh_ and yes/no questions were 35.1% and 62.2%, respectively, whereas with the older children they were 26.7% and 70.2%, respectively. Taken together, the observations reported by Warren et al. and Wood et al. suggest that interviewers overuse specific questions and that the more problematic closed-ended yes/no type of specific question may be used more then the open-ended wh_ type question. The rationale for the third specific research question was based on the reported problems associated with using various question types to obtain specific information from children, the reported findings on the high use of those types of questions during interviews with children, and the lack of literature on the use of those types of questions 33 when following the Step-Wise Interview protocol. Thus, the third specific question was stated: "When interviewing children under 8 years of age, what percentage of the trained RCMC investigators' specific questions in the first and second halves of the interviews were formed as wh_, yes/no, multiple-choice, and conditional if/then probes?" This study is ecologically valid because it is based on official RCMP videotaped interviews that were held with children who were involved in cases of alleged sexual abuse. Method Participants The original population from which the sample for this thesis was to be drawn consisted of approximately 300 videotaped child sexual abuse interviews that had been conducted by the RCMP. The sample was to include only interviews with children whose cases were closed; however, the age range for those cases was not specified. The first batch of videotapes provided by the RCMP included interviews with children who were between 3 and 12 years of age. After viewing those interviews, an age restriction of 8 years of age and under was applied for the remainder of the sample collection. This was done because it was clear that the younger children presented the investigators with the most challenges during the interviews. The age restriction of under 8 years of age was chosen based on Piaget's theory on the development of thought (Dworetzky, 1987; Nairne, 1997) and research findings based on Piagetian theory. One promising application of Piagetian theory to language acquisition focuses on the emergence of metalinguistic judgements in children, particularly on the ability of children to perceive ambiguous sentences as ambiguous and to recognize when two 34 sentences are paraphrases of each other. Cairns and Cairns (1976) reported that the foregoing abilities do not seem to emerge until 6- to 8-years-old. Furthermore, psycholinguistic researchers have reported that a new phase in language development appears to being around 8-years-old. For example, Karmiloff-Smith (1979) stated, "parallel to the development of metalinguistic skills, the over 8 year old seems to attain the capacity for a more abstract level of comprehension and can cope, if need be, without the interplay of functional, syntactic, semantic and pragmatic clues used in normal discourse" (p. 323). The application of the age restriction reduced the size of the population from which samples could be drawn to 52 families . The RCMP could not locate 30 of the families associated with those 52 cases. Families involved in the remaining 22 cases were located and ask to participate; 20 families consented and 2 families declined. When the RCMP attempted to locate the videotapes and transcripts associated with the 20 consenting families, they found that the interviews associated with 5 of the families could not be provided because the videotapes had been destroyed or transcripts of the sessions had not been produced. Thus, the sample of interviews with young children was further reduced to 15 consenting families. Three families with older children had consented to participate in the study prior to the application of the age restriction. Those 3 families produced a total of 4 videotaped interviews with children: 2 were single interviews with unrelated older children; and, 2 were single interviews with older children who were siblings. The consent of 15 families whose young children had been interviewed produced a total of 17 videotaped interviews: 9 interviews with young children who were interviewed only once and who had no siblings involved; 4 interviews from 2 young children who were 35 each inteNiewed twice; and , 4 inteNiews from 2 sets of siblings (both young siblings in one case, and one younger and older sibling in the other case). In total , the researcher received 21 videotaped inteNiews and their associated transcripts. Five of those inteNiews were with older children. Of the 5 inteNiews with older children , 1 inteNiew was excluded from analysis because the child's younger sibling's inteNiew was included in the sample of young children. One older child's inteNiew was excluded because his or her older sibling's inteNiew was included in the analysis for older children. Thus, data were collected from 3 inteNiews with older children (ages ranged from 10 years 9 months to 12 years 0 months, M = 137.33 months, SO= 7.64). Sixteen of the 21 videotaped inteNiews were with young children. Exclusions of inteNiews with young children included 1 inteNiew because the child was so upset that the investigator closed the session immediately, 1 inteNiew because the child's sibling's inteNiew was included for analysis, and 2 inteNiews because they were second inteNiews and the initial inteNiews were included for analysis. Thus, 12 inteNiews with young were included for analysis (ages ranged from 3 years 1 month to 7 years 6 months, M = 60.83 months, SO= 17.70). Table 2 shows the distribution of the sample by inteNiew status (No Disclosure and Disclosure) , the children's ages and the children's and officers' genders. The inteNiews were conducted by 8 RCMP investigators from the Prince George, British Columbia, detachment. Four of the 8 investigators conducted 11 of the 15 inteNiews. Specifically, officer number 1 conducted 2 inteNiews, office number 2 conducted 5 inteNiews, officer number 5 conducted 2 inteNiews and officer number 7 conducted 2 inteNiews. The overlap in the number of inteNiews conducted by those officers 36 Table 2 Sample Distribution for Interviews by Children's Age, Gender and Interview Disclosure Status by Officer Number and Gender Officer Gender and Officer Number 2 Group Children under 8 years of age (D = 12) No Disclosure Interviews Boys Girls Disclosure Interviews Boys Girls Children over 8 years of age (n = 3) Disclosure Interviews Boys Girls 2 Male Female n=G n=2 3 4 5 6 7 8 37 produced coupled data (i.e. the data were not independent) and, therefore, they were not appropriate for analyses using inferential statistical methods. Setting of Interviews When the RCMP in Prince George receive information that a child may be at risk of being sexually abused, or that a child has disclosed that sexual abuse has occurred, they conduct an investigative interview with the child. These interviews are videotaped and written transcripts of the proceedings are usually produced. The interviews used for this project were all conducted in rooms designed specifically for the purpose of interviewing children. The rooms were comfortably furnished with a living room style couch and chair. There were no toys or other play items in the rooms that may have distracted the child's attention from the purpose of the interview. The plain clothed officers did, however, have access to a sexually inexplicit body diagram that they could use to assist the child with identification of body parts and/or body functions. The rooms were equipped with an audio recording device that sat on an end table situated between the chair and the couch. Videotape recordings were made through a one-way observation window. Procedures Observer training in the Step-Wise Interview. The two observers involved in this study independently attended a two-day training session in the use of the Step-Wise Interview protocol. The official interview-training sessions were held for RCMP investigators, child protection workers, and members of the justice system. Staff Sergeant Roberts (Prince George detachment of the RCMP) facilitated both of the twoday training sessions in which the observers participated. 38 Informed consent procedures. Informed consent to view the videotaped inteNiews was obtained from both the RCMP officers who conducted the inteNiews and the children's legal guardian(s). Initially, RCMP members who had conducted sexual abuse investigative inteNiews with children were contacted by letter to explain the purpose of the research and to obtain their consent to view archived inteNiew videotapes (Appendix B) . After the officers' consents were received, RCMP Victim SeNices representatives contacted the children's legal guardian(s) by telephone to explain the purpose of the research and to ask for their signed consent to use their child's inteNiew for data collection. The Victim SeNices representatives were provided with a script to follow while making the initial telephone contact with the parent(s) or guardian(s) (Appendix C). When the parent or guardian agreed to participate, a team of two Victim SeNice representatives visited with them. During that visit the representatives obtained signatures on the consent form (Appendix D). When the required consent documentation was in place, the researcher was given temporary possession of videotaped copies of the inteNiews and the relevant transcripts. Security and confidentiality. The RCMP loaned the inteNiew videotapes and transcripts to the researcher. While in the researcher's possession , the videos and transcripts were secured in a locked cabinet in a private laboratory room at the University of Northern British Columbia (UNBC). Viewing of the videotapes and transcripts took place on the UNBC campus in the private laboratory. All identifying information was removed from the transcripts but could not be removed from the videotapes. Development of the coding manual and establishment of interobseNer agreement. Two trained obseNers used 5 videotaped inteNiews to generate 39 definitional descriptors for the coding manual. Working together, the observers applied their definitions to videotaped sequences of behavior. Revisions and refinements of ~ definitions continued until the definitions were satisfactory to the researchers. Then, working independently, each observer applied the definitions to behaviors observed in the videotapes. When substantial disagreements in the independent application of the definitions occurred, further revisions were made to the coding manual definitions. This process was repeated until the two observers, working independently, reached and maintained 80.0% agreement in their coding of the videotaped behaviors. Percent agreement was calculated by dividing the total number of agreements by the total number of agreements plus disagreements (i.e., agreements/agreements+ disagreements). The coding manual appended to this thesis (Appendix E) contains excerpts from a larger manual that was developed by the researchers to assess various interviewer behaviors. The officer behaviors in the appended manual were of interest to the primary researcher in this study. The first draft of the coding manual was completed in early July of 1996. Refinement to that manual was an ongoing process, which lead to the completion of the entire coding manual in early December of 1997. Development of the complete coding manual was an ongoing collaborative effort between the researcher for this thesis and the second researcher; therefore, both are named as authors of that manual (Hewlett & Hardy, 1997). Observational Coding All coding was done directly from the interview videotapes. When necessary the associated transcripts were used as guides to verbal content. The coding procedure required three full passes through each of the videotaped interviews, as well as a 40 review of the transcript for each interview. The interviews with the younger age group (D. = 12) were used in all three of the passes described below because the researcher was interested in observing and evaluating both the interviewers' coverage of the key elements of the protocol and the questioning techniques used with young children. The interviews with the older age group (D. = 3) were included in Passes 1 and 2 only because the researcher was interested in evaluating any observed group differences between the officers' coverage of the key elements with the younger and older children. Pass 1. The two observers worked together for Pass 1. Each videotaped interview (N = 15) was watched and timed. While viewing each videotape, the associated police transcript of the interview was read to ensure its accuracy. The two observers discussed and agreed on any changes that were made to the police transcript. Changes involved corrections to the content of the transcripts based on what was heard in the videotaped interviews. Pass 2. During this viewing of the videotaped interviews, two observers, working independently, recorded the occurrence of the key elements of the Step-Wise Interview protocol for all 15 interviews. The Pass 2 key elements were selected for inclusion based on (a) interviewing techniques recommended in the extant child eyewitness and testimony literature, (b) information provided at the February 26 and 27, 1996, StepWise Interview training session, and (c) consultations with Staff Sergeant Roberts (personal communication, June 6, 1996). lnterobserver agreement was assessed using Cohen's kappa statistic. The observers recorded whether the key elements or events discussed below occurred or not. If the event occurred, a 'yes' response was recorded . If the criterion for the key element was not met (i.e., the event did not occur), the observers recorded a 41 'no' response . For some key elements, additional response options were provided. For example, for the key element associated with stating the date of the interview, 'yes' and 'incomplete' responses were recorded if the officer said the day and month but did not specify the year. The rationale for recording the occurrence and non-occurrence of the key elements was two-fold: (a) to determine whether the trained interviewers covered the elements prescribed in the Step-Wise Interview protocol, and (b) to determine if the coverage of the elements differed between the No Disclosure and Disclosure interviews with the younger children and between the Disclosure interviews with the younger and older children. The key elements associated with Step 1 were as follows: officer stated the time the interview began , stated the date, identified self as a police officer, identified the child by name, identified any other people in the interview room , and gave a brief description of his or her professional role. The key elements associated with Step 2 were as follows: officer attempted to elicit a free narrative about a non-abuse life event, discussed the difference between the truth and a lie, explained that it is important to only talk about things that really happened, and explained to the child that was okay to say "I don't know" if that was the truth . The key elements associated with Step 3 were as follows: the use of a general question to introduce the reason for the interview, the use of specific question to assist the child identify the reason for the interview, whether the child disclosed that abuse had occurred , and whether the child recanted after making an abuse disclosure. For Step 4 of interviews in which a disclosure of abuse was made, the observers recorded the key element associated with whether or not the officer attempted to elicit a 42 free narrative from the child about the abuse-related event(s) by using an open-ended probe. For all interviews, Step 4 key elements related to whether or not the officer used a body diagram during the interview and/or whether the officer used the same body part name or body function terms as the child. The key elements associated with Step 6 included recording whether the officer asked the child if he or she had any questions, explained what would happen next, provided the child with contact names and/or numbers should he or she want to talk again, and finally, whether the officer thanked the child for his or her participation in the interview. Pass 3. One observer watched and coded all 12 of the videotaped interviews with children under 8 years of age. The second observer independently coded 3 (25%) of the 12 interviews for Pass 3. Both observers used the associated transcript for clarification when it was necessary. Coding was done on a turn-by-turn basis and focused on the officers' interviewing segments. Cohen's kappa was used to assess interobserver agreement. For clarification of the entire Pass 3 coding process, a variable coding flow chart is provided in Appendix F. Initially, each segment was recorded as being spoken by the officer, the child, or another individual in the interview room. Each officer and child segment was then coded in the following manner to identify sections of the interview that were specifically concerned with abuse: (a) the segment was not related to abuse, (b) the segment was related to body parts or body functioning, (c) the segment was related to sexual abuse by an adult, (d) the segment was related to sexual abuse by a peer, (e) the segment was related to some form of physical abuse, or, (f) the segment was inaudible or incomplete. After all of the officer and child segments were thus 43 coded, attention was given to only the officer segments and none of the child segments were further coded. Officer segments were coded for turn type. If the officer turn type was coded as a probe, the probe was coded as either direct or indirect in form, and for probe classification which included open-ended, wh_, yes/no, multiple choice, or if/than. All officer turn types where coded for misleading content and officer turn types that occurred during interview steps 3, 4, and 5 were also coded for content that was considered leading. All officer segments were coded as belonging to one of the 6 steps of the interview protocol and whether they related to the topics of body parts or functioning or telling the truth or a lie. If the officer segments were not associated with specified step or topic within the protocol, they were coded as unrelated. The goal of Pass 3 was to record specific elements of the interview on a turn-byturn basis, which permitted analyses of frequencies of specific officer behaviors and of sequential patterns. The rationale for turn-by-turn coding was to determine what types of statements and probes were used by the officers in the various steps of the interview protocol and to assess the sequence in which the steps were introduced. The codes described in detail below applied to the officers' segments only. Each officer segment was identified by 'turn type' which included plain statements, tag questions, probes (requests for information) which were defined as either mixed probes (i.e., some combination of a plain statement and a probe) or simple probes (i.e. , statements, questions, or demands), repetitions (of the child's previous statement or question), and acknowledgements (e.g., okay or hm-hm). Wh_, yes/no, and multiple-choice probe segments, whether mixed or simple, were recorded as being either direct or indirect. Probe segment classifications also included conditional if/then 44 statements, requests for repetition , and other (a category for residual probes that were not classifiable as one of the foregoing forms). Leading segments were recorded to identify officer segments that were suggestive in nature and , therefore, were likely to contaminate the child's statement about abuse-related events. Thus, segments were recorded as leading only if they occurred in an interview step that was specifically concerned with abuse (i.e., Steps 3, 4 or 5). Segments were recorded as leading if they (a) contained the answer or a choice of answers, (b) named the suspected offender before the child had done so, (c) contained explicit details of the alleged offence to which the child has not previously referred in the present interview, or (d) contained the interviewer's assumptions about the alleged abuse events (e.g., an assumption that abuse did occur). Misleading segments were recorded to obtain frequencies of segments in which the officers misrepresented reality when speaking to the children. Segments in any of the interview steps were considered misleading when the officer either misrepresented his or her intentions or made a promise that he or she might not have been able to keep. Officer segments were coded as misleading if they (a) misrepresented the intentions of or for the interview, (b) were related to the officer's ability to protect the child, the child's family, or the alleged perpetrator, and (c) were misleading in ways not covered in (a) or (b) above (e.g., "Mommy won't give you any trouble if you tell me what happened"). Each officer segment was recorded as either belonging or not belonging to one of the six steps outlined in the Step-Wise Interview protocol. Observers recorded with which of the six steps of the protocol (if any) a given segment was consistent. The goals of Step 1 are to orient the child to the interview situation and to obtain case- 45 related documentation that is important to the court. The goals of Step 2 are (a) to obtain a non-abuse life event free narrative from the child which, in addition to making the child feel comfortable, is to be used to assess the child's developmental level, and (b) to model how questioning will proceed later in the interview. Incorporated in Step 2 are questions and statements that are intended to ensure that the child knows what is expected from him or her while providing testimony during the interview. The goals of Step 3 are to draw the child's attention to the reason for the interview and to obtain an initial and truthful disclosure of abuse from the child. The goal of Step 4 is to obtain a free narrative about the abuse-related events from the child once a disclosure has been made. The purposes of using a sexually inexplicit body diagram can include (a) eliciting the child's labels for body parts so that there is a clear understanding of any testimony regarding the abuse, (b) assessing the level of the child's knowledge for the names of body parts and the function of those parts, and (c) assisting the child in the abuse disclosure process. The goal of Step 5 is to clarify and/or extend information provided by the child in Steps 3 and 4. The goals of Step 6 include both ensuring that the child's questions and concerns are addressed and providing an explanation of what will (or might) happen next. Furthermore, in Step 6, thanking the child for participating and providing information on who to contact should they want to talk again shows the child that his or her testimony is valuable and appreciated. The variable for interview steps and topics included officer segments that were associated with Steps 1, 2, 3, 4, 5 and 6 of the prescribed protocol, on the topics of truth/lie, the body diagram and body parts and/or functions, and unrelated topics. Unrelated topics were either attempts by the officers to bring the children's attention 46 back to the discussions at hand (e.g., "Please sit back down in your chair.") or were associated with drifts from the then current protocol-related topic of conversation. For all of the interview steps, application of the codes required judgments about the nature of the activity in which the officers were engaged. When the officers were clearly pursuing an activity consistent with the goals of a given step, observers recorded a non-zero code even though the particular segment being coded may not have been specified in the protocol. For example, when an officer was asking specific questions about abuse events (i.e., Step 5), he or she may have used paralinguistic or verbal acknowledgement cues (e.g., "uh-huh" or "okay") to encourage the child to continue his or her answer. Because such segments were consistent with the purpose of Step 5, they were recorded as Step 5. If the officer was engaged in Step 5 questioning and the child began leaving the room, officer segments related to having the child return to his or her chair were given a step code of zero because they were not associated with a prescribed step of the protocol. Zero codes also were used when the conversation drifted from the protocol-related topic at hand. For example, if the focus of the conversation was related to naming body parts and the officer was drawn into a conversation about the child's pet, the officer segments associated with the pet topic received a step code of zero. lnterobserver agreement. Cohen's kappa values were calculated as described in Bakeman and Gottman (1997). Cohen's kappa corrects for chance agreement and, therefore, relative to calculations of percentage of agreement, it is a preferable method of evaluating interobserver agreement (Bakeman & Gottman). For Pass 2, percentages of agreement (agreements/agreements+ disagreements) and Cohen's kappas were calculated using all 15 interviews. Key 47 elements were recorded independently by both observers for Steps 1, 2, 3, 4 (when applicable), and 6 of the Step-Wise Interview. There were no key elements associated with Step 5 of the protocol. For the majority of the key elements in Pass 2, kappas were calculated by placing the frequencies of the two observers' agreements and disagreements in a 2 x 2 confusion matrix which accounted for 'yes' and 'no' responses. When the response choices were greater than two, the matrices were expanded to accommodate the additional response options (e.g., 3 x 3 or 4 x 4). The diagonal tallies in the confusion matrices reflected agreements between the two observers, whereas tallies off the diagonal reflected observers' disagreements. For Pass 3, percentages of agreement and Cohen's kappas were based on 3 of the 12 interviews that the officers conducted with children who were under 8 years of age. One observer coded all 12 interviews and the second observer randomly selected and coded 3 of the 12 interviews. The primary observer was naive to which 3 interviews were chosen by the second observer. Kappa values and percentages of agreement were calculated for the main variables of turn type, probe type, misleading segments, leading segments, and interview steps and topics. The turn type variable included 6 types of officer segments: plain statements, tag questions, mixed probes, simple probes, repetitions, and acknowledgements. The probe type variable consisted of 11 kinds of segments which included both direct and indirect wh_, multiple-choice, and yes/no probes, as well as conditional if/then probes, requests for repetition, and residual probes. The misleading segment variable consisted of statements that misrepresented the intentions for the interview or were promises that could not be kept. The leading segment variable consisted of statements or probes that were associated with the abuse-related steps of the protocol (i.e., Steps 3, 4, and 5) and which contained 48 the answer or a choice of answers, named the suspected offender before the child had done so, contained explicit details of the alleged offence to which the child had not previously referred in the interview, or contained the interviewer's assumptions about the alleged abuse. There were 9 types of segments included in the steps and topics variable: segments related to Steps 1, 2, 3, 4, 5, and 6 of the protocol; segments that were related to the topics of truth/lie, body parts and/or functions; and segments that were related to topics not specified in the protocol (e.g., an officer's attempts to have a child return to his or her chair). The percentages of agreement were calculated for the turn types of plain statements, tag questions, probes, acknowledgements, and repetitions. To obtain the percentage of agreement for probes, the mixed and simple probe types were combined to create a single category for probes (Appendix G). Percentages of agreement were calculated for the 'general', 'specific', and 'other' question type categories. To facilitate · the calculations, the original 11 probe types were collapsed in the following manner. The direct and indirect open-ended probes were combined to create the general questions category; the direct and indirect wh_, multiple-choice, and yes/no probes were combined within each of the respective probe type categories to create single categories of wh_ questions, multiple-choice questions, and yes/no questions; the categories of wh_, multiple-choice, and yes/no questions were then combined with the conditional if/then probes to create the specific questions category; the residual probes and the requests for repetition were combined to create the category labeled 'other' (Appendix G). 49 Results Results are presented firstly for the interobserver agreements obtained in Pass 2 and Pass 3 of the data collection process. Kappa values and the percentages of agreement are provided for each of the Pass 2 key elements as well as for each of the main officer segment variables for the Pass 3. Percentages of agreement are also provided for the types of statements and probes used by the officers, the steps and topics of the interview, and the general, specific, and other question categories. The second section of results reports findings that assessed the investigators' adherence to the structure of the Step-Wise Interview protocol. Coverage of the key elements in the protocol were compared for the No Disclosure and Disclosure interviews with the younger children and the Disclosure interviews with the younger and older children. The time spent in each step of the protocol, the order in which the steps were covered, and use of inappropriate interviewing techniques are described for the No Disclosure and Disclosure interviews with the young children. Results are also reported for the officers' use of various types of statements and probes, general and specific questions, and types of specific questions, for the first and second halves of the No Disclosure and Disclosure interviews with young children. For each research question, details on the sub-samples used and data preparation and data reduction (if applicable) are presented along with the findings. lnteroberserver Agreement Cohen's kappa values were initially calculated manually or by using SPSS (version 7.5). To confirm each calculated kappa value, the observers' agreements and disagreements were placed in a computer based confusion matrix (ComKappa). ComKappa computes Cohen's kappa as described in Robinson and Bakeman (1998). 50 Fleiss' (1981) benchmarks were used as guidelines for assessing the relative strength of the interobserver agreement. Kappa values below 0.40 were taken to represent poor agreement beyond chance, values between 0.40 and 0.60 were considered to be fair agreement beyond chance, values ranging between 0.60 and 0.75 were considered as good agreement beyond chance, and values over 0. 75 were taken to represent excellent agreement beyond chance (Fieiss). However, as demonstrated by Bakeman, Quera, McArthur, and Robinson (1997), "no one value of kappa can be regarded as universally acceptable" (p. 357) and, therefore, it is acknowledged that Fleiss' benchmarks do not always provide the most accurate assessment of interobserver agreement. Thus, when kappa was poor (i.e., less than or equal to 0.40), Bakeman et al.s' method for evaluating observer accuracy and the magnitude of the calculated kappa was used to assist in the interpretation of the kappa value. Pass 2 key elements. As shown in Table 3, using Fleiss' (1981) benchmarks, interobserver agreements for 21 of the 22 key elements ranged from fair (0.44) to excellent (1.00). On face, the calculated kappa of 0.36 for the key element associated with attempting to elicit a non-abuse life event free narrative suggested poor interobserver agreement; therefore, observer accuracy for that kappa value was evaluated using the methods described in Bakeman et al. (1997). Examination of the confusion matrix for the non-abuse life event variable revealed that the simple probabilities were moderately variable. Following Bakeman et al.s' guidelines, it was determined that the calculated kappa of 0.36 for the 2 x 2 confusion matrix represented observer accuracy over 80.0%. Although interobserver agreement for the life event key element was low, Bakeman et al.s' guidelines indicated that both the level of observer accuracy and the magnitude of the calculated kappa were sufficiently large to allow for 51 Table 3 Percentage of Agreement and Kappa Statistics for the Key Elements of the Step-Wise Interview (N = 15) Percent Kappa Agreement Value Officer states time interview begins 100.0% 1.00 Officer states date of interview (date/month/year) 100.0% 1.00 Officer identifies him/her self as a police officer 93.3% 0.83 Officer identifies the child 93.3% 0.83 Officer identifies other people in the room 93.3% 0.84 Officer gives brief description of his or her role 100.0% 1.00 67.0% 0.36 100.0% 1.00 that really happen 80.0% 0.57 Officer explains okay to say "I don't know" 100.0% 1.00 Interview Steps and Key Elements Step 1: Introduction Step 2: Rapport Building Officer attempts to elicit a non-abuse life event free narrative Officer discusses the difference between the truth and a lie Officer explains important to only talk about things (table continues) 52 Table 3 (cont'd) Interview Steps and Key Elements Percent Kappa Agreement Value Step 3: Introducing the Topic of Abuse Officer brings up reason for interview using a general question 100.0% 1.00 topic of abuse 80.0% 0.53 Child makes a disclosure 100.0% 1.00 Child recants after making a disclosure of abuse 100.0% 1.00 80.0% 0.69 100.0% 1.00 80.0% 0.44 Officer asks if child has any questions 93.3% 0.86 Officer explains what will happen next 87.0% 0.73 Officer provides child with contact information 100.0% 1.00 Officer thanks child for his or her participation 100.0% 1.00 Officer asks more specific question to introduce the Step 4: Abuse Free Narrative Officer attempts to elicit a free narrative about abuse event Officer uses body diagram to discuss body/function terms Officer uses same private body part terminology/gestures as the child Step 6: Closure 53 meaningful yet cautious interpretation of results for that variable. Pass 3 turn-by-turn officer segments. As shown in Table 4, using Fleiss' (1981) benchmarks, the calculated kappa values for the main officer segment variables of turn type and probe type were excellent and the kappa for interview steps and topics was good. The percentages of agreement for the various categories within the main officer segment variables ranged from a low of 76.1% for open-ended probe types to a high of 100.0% for the topic of truth and lie. Thus, the results for interobserver agreement and percentages of agreement for the foregoing officer segment variables and their associated subcategories were sufficient for meaningful interpretations of the results associated with those variables to be made. However, the frequency of occurrences for both the misleading and leading segment types were low and the calculated kappa values and percentages of agreement for those variables reflect agreements on the non-occurrence of those officer segments. Because this was the case, the researcher could not make a definitive determination as to whether misleading and leading segments rarely occurred or whether real occurrences were undetected because of faulty measurement techniques. As a result, the assumption was made that the low frequency of coding the occurrences of misleading and leading segments by both of the observers could be used as a reasonable measure for limited interpretation of results associated with those variables. General Research Question #1: Adherence to the Step-Wise Interview Protocol The results presented in this section assess the officers' adherence to the structure of the Step-Wise Interview by evaluating the coverage of the key elements in the protocol during interviews with the younger and older children, the percentage of time spent in the various steps of the interviews with younger children, and the order in which the 54 Table 4 Percentage of Agreement and Cohen's Kappa for Turn-by-Turn Officer Segment Types. Steps and Topics. and General and Specific Questions (n = 3) Officer Segment Variables Turn Type Percent Kappa Agreement Value 84.4% 0.80 Plain statement 93.0% Tag questions 97.3% Probes (mixed and simple) 95.7% Repetitions 98.2% Acknowledgements • Probe Type 93.7% 82.1% Open-ended (direct and indirect) 76.1% Wh_ (direct and indirect) 90.9% Multiple-choice (direct and indirect) 99.1% Yes/no (direct and indirect) 93.6% If/then 99.4% Repetition Requests 99.7% Residual probes 96.4% 0.75 (table continues) 55 Table 4 (cont'd) Percent Kappa Agreement Value Misleading segments (very low frequency) 99.9% 1.00 Leading segments (low frequency) 99.0% 0.50 Interview Steps and Topics 78.2% 0.70 Step 1 Introduction 98.5% Step 2 Rapport building 97.9% Step 3 Introducing the topic of abuse 92.8% Step 4 Abuse-related free narrative 96.6% Step 5 Specific questioning 94.0% Step 6 Closure 94.7% Topic of truth/lie 100.0% Topic of body (parts-functions/diagram) 99 .9% Topics unrelated to the protocol 82.6% Officer Segment Variables 56 steps of the protocol were introduced during inteNiews with the younger children. 1(a): Coverage of Key Elements in lnteNiews with Younger and Older Children The results for the coverage of the Pass 2 key elements are descriptive in nature. A ratio of 2:1 was used as the criterion for identifying substantial differences in the coverage of the key elements between the two age groups and between the No Disclosure and Disclosure inteNiews. The 2:1 ratio was used because the data were not appropriate for use with inferential statistics, no other precedent for detecting differences of this nature was found in the literature, and a ratio of less than 2:1 was deemed too low to demonstrate that differences were substantial. Table 5 shows the percentage of officers who covered the key elements associated with the various steps of the Step-Wise lnteNiew protocol when inteNiewing children from both the younger and older age groups. For the younger children, the results are presented separately for the No Disclosure and Disclosure inteNiews. All of the older children made a disclosure of sexual abuse during their inteNiews. Step 1. The goals of Step 1 are to orient the child to the inteNiew situation and to document case-related information that is important to the court. No substantial differences (hereafter simply referred to as differences), based on the 2:1 ratio, were obseNed in the coverage of the Step 1 key elements between the No Disclosure and Disclosure inteNiews with the younger children or between the Disclosure inteNiews with the younger and older children. In Step 1, the legal documentation of the inteNiews (time, date, and identification of the child and others in the inteNiew room) were covered fairly well. Most of the officers, regardless of inteNiew status or age group, identified themselves as "police officers" and described their professional 57 Table 5 Percentage of Officers who Covered the Step-Wise Interview Key Elements with Children Under 8 Years of age and Children Over 8 Years of age by Interview Status Interview Status and Child's Age Younger Children Key Element No Disclosure n=5 Older Children Disclosure Disclosure n=7 n=3 Step 1: Introduction States time interview begins 80.0% 57.1% 66.7% States date of interview 80.0% 71.4% 100.0% Identifies child 80.0% 100.0% 100.0% Identifies others (when applicable) 100.0% 100.0% 100.0% Identifies self as police officer 100.0% 85.7% 66.7% Describes his or her role 80.0% 100.0% 66.7% (table continues) 58 Table 5 (con't) Interview Status Younger Children Key Element No Disclosure D.=5 Older Children Disclosure D._=? Disclosure D.=3 Step 2: Rapport Building Attempts to elicit a life event free narrative 40.0% 14.3% 0.0% Discusses truth and lie 40.0% 85.7% 66.7% 40.0% 71.4% 66.7% 20.0% 85.7% 33.3% Discusses only talking about things that really happened Explains it is okay to say "I don't know" if child doesn't know Step 3: Introducing the Topic of Abuse Uses general question 60.0% 57.1% 66.7% Uses specific questions 100.0% 71.4% 33.3% (table continues) 59 Table 5 (can't) Interview Status Younger Children Key Element No Disclosure D.=5 Disclosure n=7 Older Children Disclosure D.=3 Step 4: Abuse-Related Free Narrative Attempts to elicit a free narrative about the abuse event(s) N/a 71.4% 100.0% 40.0% 50.0%a 66.7% Explains what will happen next 20.0% 66.7% 33.3% Provides contact information 20.0% 16.7% 0.0% Thanks child 100.0% 50.0% 33.3% Step 6: Closure Asks child if he or she has questions (table continues) 60 Table 5 (can't) Interview Status Younger Children Key Element Older Children Disclosure n=7 Disclosure !}=3 60.0% 71.4% 66.7% 100.0% 100.0% 66.7% No Disclosure !}=5 Body Parts and Functions Body diagram is used Officer uses same body part names as child (when applicable) a One Disclosure interview ended abruptly because the child was very upset and, therefore, the officer conducting that interview was not provided with the opportunity to cover the Step 6 key elements. Accordingly, the results for coverage of the Step 6 key elements are based on a sample of n =6 for the Disclosure interviews. 61 role by saying something like "my job is to talk to children about things that are happening in their lives and I talk to lots of nice children who are just like you." Overall, the officers tended to cover each of the Step 1 key elements. Step 2. The goals of Step 2 include obtaining a life event free narrative that is unrelated to abuse and modeling how the questioning will proceed throughout the interview. Obtaining the life event free narrative is intended to serve two purposes: (a) to help the child relax and feel comfortable, and (b) to provide information for assessing the child's cognitive developmental level. Incorporated in Step 2 are questions and statements that are intended to set the parameters for testimony expectations and to clarify that the child understands those expectations. Differences, based on the 2:1 ratio, were found between the officers' attempts to elicit a non-abuse life event free narrative from the young children in the No Disclosure and Disclosure interviews and between the interviews with the younger and older children. For the younger group, attempts to elicit a non-abuse life event occurred more often in the No Disclosure interviews than in the Disclosure interviews; however in the No Disclosure interviews the key element was covered by only 40.0% of the officers. No attempts were made to elicit a non-abuse free narrative in the interviews with the older children and that result differed from both the No Disclosure and Disclosure interviews with the young children. It is important to note that none of the officers attempted to elicit a non-abuse life event free narrative from the older children during Step 2. It appeared that this key element was not necessary because the older children were prepared to discuss the abuse-related events early in the interview sessions. Differences, based on the 2:1 ratio, were observed between the No Disclosure and Disclosure interviews with the younger children in the coverage of the parameter 62 setting elements of discussing truth/lie and explaining that it is was okay to say "I don't know". For the younger group, the forgoing key elements were covered more often in the Disclosure interviews than in the No Disclosure interviews. There was also a 2:1 ratio difference between the coverage of "I don't know" between the Disclosure interviews with the younger and older children as that key element was covered more often with the younger children. Thus, results showed that less than half of the officers covered the parameter setting key elements with the young children in the No Disclosure interviews and that the coverage of those elements tended to be substantially higher with the young children in the Disclosure interviews. Coverage of the parameter setting key elements of truth/lie and only talking about things that really happened were different than the coverage of "I don't know" in the interviews with the older children, as both of the former key elements were covered more often than the latter. In general, with the exception of attempting to elicit a non-abuse life event narrative, a greater percentage of officers covered the Step 2 key elements in the Disclosure interviews than in the No Disclosure interviews. The results showed that the Step 2 elements were skipped by more that half of the officers in the No Disclosure interviews. The key elements related to discussing the difference between the truth and a lie and telling the child it is okay to say "I don't know" were covered more often in the Disclosure interviews than in the No Disclosure interviews. It is possible that covering those topics is an important factor in making the child feel comfortable and, thus, more likely to disclose abuse if it did occur. Step 3. The goals of Step 3 are to draw the child's attention to the reason for the interview and to obtain an initial and truthful disclosure of abuse from the child if abuse 63 had in fact occurred. The interview protocol prescribes that officers first use general questions to introduce the topic of abuse and, if necessary, specific questions can later be used to focus the child on the reason for the interview. However, the purpose of Pass 2 coding was to assess the coverage of the key elements but not the sequence in which they were introduced. Thus, the Pass 2 codes for the use of the general and specific questions do not reflect the order in which the questions were introduced but rather whether the two question types were used or not. Comparisons between the interviews with the younger and older children showed that the use of specific questions was different between the groups, with specific questions being more prevalent in both the No Disclosure and Disclosure interviews with the young children than in the Disclosure interviews with the older children. No differences were found in use of general questions between either type of interview with the younger children or between interviews with the younger and older children. However, within the older group the use of general questions was more prevalent than the use of specific questions. As outlined in the effective child interviewing literature and in the Step-Wise Interview protocol, investigators are urged to use general, open-ended questions to obtain as much information as possible from the child in a narrative form . The results showed that it is not uncommon for officers to ask specific questions while introducing the topic of abuse, especially when interviewing the younger children. In the No Disclosure interviews, all of the officers used specific questions which may reflect their ongoing attempts to obtain a disclosure of abuse from the child. Step 4. The goal of Step 4 is to obtain a free narrative account of the abuse event(s) from the child and, therefore, was only applicable to the Disclosure interviews. 64 When abuse was disclosed, the majority of the officers attempted to obtain a free narrative about the abuse-related event(s) from both the younger and older children. However, those attempts seldom produced a narrative response from the young children, whereas the older children were able to provide such narratives. Step 6. The goals of Step 6 include ensuring that the child's questions and concerns are addressed and providing an explanation of what will (or might) happen next. The goals also include demonstrating to the child that his or her testimony is valued and appreciated. This is achieved by thanking the child for participating and providing information on who to contact should he or she want to talk again. Regardless of the age group or interview status, coverage of the Step 6 key elements was generally poor. The results showed that in the Disclosure interviews with the younger children the officers explained what would happen next more often than in the No Disclosure interviews with the younger children and the Disclosure interviews with the older children. However, in the No Disclosure interviews the officers thanked the children for their participation more often than in the Disclosure interviews with either the younger or older children. While the officers often skipped providing contact information with younger children, they omitted it entirely with older children. It is important that interviewers cover the Step 6 key elements because they provide the child with important information about the investigative procedure and they give the child an opportunity to ask questions. With complete coverage of the key elements the child is likely to leave the interview with a sense of reassurance and closure. Additionally, but importantly, if sexual abuse had occurred but the child did not make a disclosure during the interview, the child may be more likely to disclose at a later date if he or she knew who to tell and felt confident that their statement was valuable and 65 appreciated. Closing the interview without ensuring that the child feels comfortable, and not ensuring that the tone is supportive, is counter to the guidelines for effective child interviewing. Body parts and functions. The rationale for using a sexually inexplicit body diagram and/or discussing body parts and functions involves (a) eliciting the child's labels for body parts so that there is a clear understanding of any testimony regarding the abuse, (b) assessing the level of the child's knowledge for the names of body parts and the function of those parts, and (c) assisting the child in the abuse disclosure process. Officers are to use the same terminology for body parts and functions as the children to ensure there is no miscommunication or confusion. There were no differences observed with regards to the use of the diagram between the No Disclosure and Disclosure interviews with the younger children, or between the younger and older disclosure groups. In most of the interviews, regardless of the interview status and age, the officers used the same terminology for body parts or functions as the child. This is an important finding because, had the officers used different terms, they may have confused the children and possibly contaminated the interview process. 1(b): Percentage of Time Spent in the Various Steps and on Various Topics As recently noted by Poole and Lamb (1998), little research has been conducted on the structure of the Step-Wise Interview protocol. Furthermore, literature on the length of time interviewers spend in investigative interview sessions, or the length of time spent in the various stages or steps of those sessions is sparse. Literature on the optimal length of time to spend in interview sessions or in the various stages or steps of the interviews is lacking; however, Davies, Westcott, et al. (in press) suggested that 66 extended rapport building (e.g., Step 2 greater than 8 min) can have a powerful influence on the course of the interview, especially with children under 12 year-of-age. The results in this section describe the percentages of interview time that the officers spent in Steps 1 through 6 of the Step-Wise interview protocol for the No Disclosure and Disclosure interviews with the young children. The results also describe the percentages of interview time that was spent discussing body terms and on topics that were not related to the specific steps prescribed in the protocol. The numbers of officer segments in any given step or on the various topics were used as a measure of time spent. To ensure the validity of using the number of officer segments for this analysis, a correlation between the total number of officer segments and the length of time of the interviews was obtained, r (1 0) = .96, Q < .01. The difference between the mean length of time for the No Disclosure (21.2 min, SD = 16.2; range approximately 11 to 50 min) and Disclosure (23.2 min, SD = 7.5; range approximately 13 to 34 min) interviews was not statistically significant,! (1 0) = .29, Q = .78, and, therefore, it was appropriate to make comparisons between the percentages of time spent in the various steps and on topics in the two types of interviews. Data preparation. Only data from the interviews with children under 8 years of age were used for this analysis. To calculate the mean percentage of time spent in each step of the interview protocol, the number of officer segments from each interview was calculated. The number of segments associated with each interview step (i.e., 1 through 6) or topic category (i.e., topics related to body parts and/or functions and topics unrelated to the protocol) were then calculated for each interview. Those step and topic related totals were divided by the total number of officer segments on a per 67 interview basis to obtain the mean percentage of time spent in the steps or on the topics for each interview. The percentages so calculated served as the dependent variables. The mean percentages and standard deviations for the time spent in each of the interview steps, on topics related to body parts and functions, and the time spent in discussions that were not considered to be associated with specific interview related topics are provided in Appendix H. The data were not appropriate for use with inferential statistics therefore, as in comparisons for the coverage of key elements, a ratio of 2:1 was used as the criterion for identifying substantial differences in the percentage of time spent in the various steps and on topics both within and between the two types of interviews. Time spent. As depicted in Figure 1, using the 2:1 ratio, differences between the two types of interviews were found in the percentage of time spent introducing the topic of abuse (Step 3), asking specific questions (Step 5), and in closing the interview sessions (Step 6). In the No Disclosure interviews, the officers spent 21.6% (approximately 4.58 min) of the interview time in Step 3, whereas the officers in the Disclosure interviews spent approximately 10.0% of the interview time in that step. Those results indicated that when a disclosure of abuse was not forthcoming the officers spent more time attempting to obtain one. The results also showed that when a disclosure was not made the interviewers moved on and continued to ask specific questions that were in some way related to the suspected abuse event(s). This is evident in that interviewers spent 6.5% of the interview time in Step 5. Although the officers in the No Disclosure interviews asked specific abuse-related questions, the time spent in Step 5 was substantially less for those interviews than for the Disclosure interviews. Thus, once a disclosure of abuse was made by the children, the officers 68 50.0% D No Disclosure 40.0% • Disclosure Q) E i= 30.0% 0 Q) Ol eel "EQ) ...u 20.0% Q) 0.. 10.0% 0.0% Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Body Other Interview Steps and Topics Figure 1. Figure 1 depicts the percentage of time the officers spent in the various steps of the interview (Stepl: Introduction, Step 2: Rapport Building, Step 3: Introducing the Topic of Abuse, Step 4: Abuse-Related Free Narrative, Step 5: Specific Questioning, and Step 6: Closure) in the No Disclosure (!! =5) and Disclosure (!! =7) interviews. The figure also depicts the percentage of time spent on the topic of body parts and functioning (Body) and in discussions that were not considered to be associated with specific interview related topics (Other, e.g., asking the child to return to his or her chair). 69 spent a good percentage of the interview asking specific abuse-related questions in an attempt to clarify and/or extend the information the children provided during Step 4. Those results may be indicative of the problems the investigators encountered with regards to obtaining comprehensive free narrative accounts about the abuse-related events from the young children. Officers in the No Disclosure interviews spent substantially more time closing the interviews (Step 6) than did officers in the Disclosure interviews; however, the least amount of time was spent in that step of both types of interviews. There were no 2:1 ratio differences between the two types of interviews with regards to the percentages of time spent on the topic of body parts and functioning or topics unrelated to the protocol. However, for both types of interviews, the time spent on unrelated topics was greater than the time spent in all other steps with the exception of Step 3 in the No Disclosure interviews and Step 5 in the Disclosure interviews. In summary, approximately 2 minutes of the interview sessions were spent in Steps 1 and 2, and on the body topic in the No Disclosure interviews, and in Steps 1, 2, 3 and on the body topic in the Disclosure interviews. In the No Disclosure interviews, approximately 5 minutes were spent introducing the topic of abuse and just over 1 minute was spent asking specific abuse-related questions, whereas in the Disclosure interviews, about 2 and 6 minutes, respectively, were spent in those two steps. Less than 1 minute of the interview time was spent closing the sessions in both types of interviews and about 8 minutes was spent on topics unrelated to the Step-Wise Interview protocol. 1(c): Order in which the Six Steps of the Protocol were Introduced As noted, research on the structure of the Step-Wise Interview protocol is lacking (Poole & Lamb, 1998); therefore, the results in this section answer the question of 70 whether the officers in this study followed the recommended order for introducing the various steps of the Step-Wise Interview protocol when they interviewed the younger children. Data preparation. To determine the order in which the steps of the protocol were introduced some of the original officer segment codes were re-coded. Officer segments, originally coded as 2.1 to identify them as interview parameter setting segments during the rapport building step, were re-coded to 2.0 which was consistent with rapport building for Step 2. Segments associated with the use of the body diagram and/or discussions about body parts or functions were not used in the step order analysis. Step order coding began with the first officer segment for each interview and continued to the last officer segment of that interview. Once an interview step order code had been assigned, it was not assigned or recorded a second time. For example, if an officer's segments were consistent with Step 1, the step order code of 1 was assigned, when the officer's segments were consistent with Step 2, the step order code of 2 was assigned, if the officer's segments then switched back to those consistent with Step 1, those segments continued to be assigned the step order code of 2. The step order code was not changed until there were five consecutive officer segments that were consistent with an alternative interview step. For example, if the officer's ongoing segments were constant with Step 1 of the interview protocol (and, therefore, were assigned the step order code of 1) and the officer said "thank you for coming here today" (first segment not consistent with Step 1 but consistent with Step 6) "here is my card in case you want to phone me sometime after today's interview" (second segment not consistent with Step 1 but consistent with Step 6) the step order 71 code remained 1. Once five consecutive officer segments consistent with an alternative interview step occurred, the step order code change began with the first of the five consecutive segments of the new step. Officer segments that had an original StepWise Interview code zero because they were not associated with a prescribed step or topic of the protocol received a step order code consistent with the step in which they occurred. Order of step introduction. Step order analysis was conducted on the interviews with the children who were under 8 years of age, subdivided by interview disclosure status. The officers' use of the six interview steps, and the order in which the steps were introduced during the No Disclosure and Disclosure interviews, are described in Tables 6 and 7, respectively. The tables show (a) the number of interviews in which each step was used, (b) the number of interviews in which the step was used in the prescribed Step-Wise Interview protocol order, and (c) if the step was used out of order, whether it was introduced early or late. Of the 5 No Disclosure interviews, 3 were conducted in the correct Step-Wise Interview order. All of the No Disclosure interviews began with the introduction step, but in one interview the officer introduced the topic of abuse before performing the key elements associated with rapport building. Specific questions, a step that is not prescribed in the protocol when a disclosure has not been made, were asked in 3 of the No Disclosure interviews, and in one of those interviews, the officer closed the interview and then moved back to asking specific questions. The most consistent problem in the No Disclosure interviews occurred with Step 5. Technically, Step 5 of the protocol is to be used, if necessary, to clarify and extend the information obtained from the child 72 Table 6 Order in which the Steps of the Step-Wise Interview were Introduced in No Disclosure Interviews with Younger Children (n = 5) Number of interviews in which Step was: Used in Steps Used Correct Order Used Early Used Late 1: Introduction 5 5 0 0 2: Rapport Building 5 4 0 1 3: Introduce Topic 5 4 1 0 4: Free Narrative N/A N/A N/A N/A 5: Specific Questions 3 2 0 1 6: Closure 5 4 1 0 73 Table 7 Order in which the Steps of the Step-Wise Interview were Introduced in Disclosure Interviews with Young Children (n = 7) Number of interviews in which Step was: Used in Steps Used Correct Order Used Early Used Late 1: Introduction 7 7 0 0 2: Rapport Building 7 7 0 0 3: Introduce Topic 6a 6 0 0 4: Free Narrative 5b 3 0 2 5: Specific Questions 7 3 4 0 6: Closure 5c 5 0 0 a One spontaneous disclosure of abuse occurred. b No attempt was made to elicit an abuse-related free narrative in two interviews. cOne interview ended abruptly. 74 during his or her free narrative about the abuse-related events. However, 60.0% of the officers asked specific questions that were in some way related to the topic of abuse even though the child had not made a disclosure that the abuse had occurred. Of the 7 Disclosure interviews, 5 were conducted in the order prescribed by the Step-Wise Interview protocol. All of the Disclosure interviews began with Steps 1 and 2 in the correct order. In one of the interviews, the child made a spontaneous disclosure of abuse, so the officer did not have to introduce the topic. In the 6 remaining interviews, the topic of abuse was introduced in the correct order. In 2 of the 7 Disclosure interviews, the officers did not attempt to elicit an abuse-related free narrative; however, when there was such an attempt, it was introduced in the correct order in 3 interviews but after the introduction of Step 5 in the remaining 2 interviews. The specific questioning step occurred in all 7 Disclosure interviews and was introduced in the correct order in 3 of those interviews; however, specific questions were asked before attempts were made to elicit an abuse-related narrative in 4 interviews. When the Disclosure interviews did not end abruptly, all of the officers closed the interviews in the correct order. In the Disclosure interviews, the most problematic areas appeared in the steps associated with eliciting an abuse-related free narrative and asking specific questions. When a disclosure of abuse was made, 28.6% of the officers skipped the abuse free narrative step, but they did ask specific questions related to the abuse event(s). When there was an attempt by the officer to obtain a free narrative about the abuse events, 40.0% of the interviewers asked abuse-related specific questions before attempting to obtain the free narrative account of the events from the child. 75 Summary of Findings on the Adherence to the Structure of the Step-Wise Interview Protocol The evaluation on the coverage of the key elements of the Step-Wise Interview showed that the officers did well in covering the various elements associated with Step 1 of the protocol. In Step 2, however, the key elements were covered by less than half of the officers in the No Disclosure interviews, with higher coverage in the Disclosure interviews. Few officers attempted to elicit a non-abuse life event free narrative from the younger children and no such attempts were made with the older children. The officers used both general and specific questions when introducing the topic of abuse but the use of specific questions was more prevalent with the younger children than with the older children . When the children made a disclosure of abuse, the majority of officers attempted to elicit a free narrative about the abuse event(s). The coverage of the key elements associated with closing the interview was generally poor. Most of the officers used the body diagram during the interviews and when there were discussions related to body parts and functioning, all of the officers' use the same terminology as the younger children and the majority of the officers did so with the older children. The percentage of interview time spent in Steps 1 and 2, and on the topic of body parts and functioning, was similar for both the No Disclosure and Disclosure interviews with young children. The officers in the No Disclosure interviews spent more time introducing the topic of abuse than did the officers in the Disclosure interviews. There was interview time spent asking specific abuse-related questions in both the No Disclosure and Disclosure interviews with the young children but, as expected, more time was devoted to that step of the interview in the Disclosure interviews than in the No Disclosure interviews. In both the No Disclosure and Disclosure interviews, the 76 percentage of time spent on topics unrelated to the protocol was higher than the percentage of time spent in any of the prescribed steps, and the least amount of time was spent closing the interviews. In general, most of the interview steps were introduced in the correct Step-Wise Interview order. Specifically, 3 of the 5 No Disclosure interviews, and 5 of the 7 Disclosure interviews, were conducted in the order prescribed in the protocol. However, some important steps were skipped by some officers, some officers introduced steps that were not prescribed in the protocol, and some steps were introduced out of order. The foregoing findings suggest that in some areas, the adherence to Step-Wise Interview protocol by the officers in this study was good. However, according to the literature on effective child interviewing techniques, many of the areas in which adherence to the protocol was low are areas in which adherence seems most crucial. General Research Question #2: Use of Inappropriate Interviewing Techniques The results presented in this section address the extent to which the RCMP investigators used inappropriate interviewing techniques when interviewing children under 8 years of age. Analysis was conducted on the frequency of occurrences for officer segments related to misleading the children by either misrepresenting the intention of the interview or making promises that they could not keep. Analysis was also conducted on the frequency of occurrences for officer segments related to leading the children (i.e., segments which contained the answer or a choice of answers, named the suspected offender before the child had done so, contained explicit details of the alleged offence to which the child had not previously referred, or contained the interviewer's assumptions about the alleged abuse events). 77 2(a): Misleading segments. The frequency of misleading segments in this sample was very low. They occurred only 6 times out of a total of 3600 officer segments. Four of those 6 statements will be described to illustrate the types of officer statements that were considered misleading. In one No Disclosure interview, there was a misleading segment related to misrepresenting the intentions for the interview that occurred in Step 1. The remainder of the misleading segments occurred in one of the Disclosure interviews and each occurrence took place in Step 3. Each of those 5 segments were related to telling the child that neither he or she, nor some other individual, would be in trouble if the child disclosed the abuse. The examples given are not verbatim but do, in general, reflect the officer's segments and/or the interactions between the officer and the child. References to names are replaced with 'X' (followed by a number if more than one person is referred to), references to mother or father are replaced with 'Parent' and gender identity is replaced with 'him or her' or 'he or she'. Examples of misleading officer segments. The following example occurred very early during Step 1 of a No Disclosure interview. This officer segment was recorded as misrepresenting the intentions for the interview because he or she implied that the chat would be about nothing more than playing. Officer segment 1: Remember how we were just talking about playing and going for a bike ride, that's alii wanted to have a chat about. The example below occurred in sequence during Step 3 of a Disclosure interview. The officer segments numbered 2, 3 and 4 were recorded as misleading because they implied that the child would not be in trouble if he or she disclosed the abuse events. Officer segment 1: You just have to tell me the truth that's all. 78 Officer segment 2: You are not in trouble. Officer segment 3: You are not going to get into any trouble. Officer segment 4: Your Parent will not give you trouble. 2(b): Leading segments. Officer segments were recorded as leading if they were likely to contaminate the child's statements about the abuse-related events and, therefore, only segments associated with Steps 3, 4, and 5 of the protocol were coded as leading. The frequency of leading statements and/or leading questions was low. Of the 1281 officer segments in Steps 3, 4 and 5, leading segments occurred only 21 times. There was a single occurrence of a leading question in Step 3 of a No Disclosure interview. Eighteen of the 21 leading segments occurred in Step 3 of one Disclosure interview. The remaining 2 leading segments occurred in two other Disclosure interviews and each was associated with Step 5. Examples of leading officer segments. The following officer-child interactions provide an example for the types of officer segments that were recorded as leading. This sequence occurred during Step 3 and the officer segments numbered 2 and 4 were considered leading because the child had not previously mentioned the names of X1 or X2 in officer segment 2, and had not mentioned telling his or her parent in officer segment 4. Officer segment 1: Can you tell me about what happened yesterday? Child: I forget. Officer segment 2: Okay, what happened yesterday with you and ... X1 or X2? Child: I don't remember. Officer segment 3: Hm. Officer segment 4: But you told your Parent after that happened, eh? 79 The next example of a leading officer segment occurred during Step 5. Officer segment number 2 was recorded as leading because the showing of the hand contained the interviewer's assumptions about the alleged abuse events. Officer segment 1: When X touched you down there, what part of his or her body did he or she use to touch you? Child: I dunno. Officer segment 2: What do you call this? (the officer shows the child a hand) The results suggested that in this sample of interviews, the frequencies for both misleading statements and leading questions were very low. The minimal use of such statements and questions is a positive finding because it indicates that the officers avoided the use of the inappropriate investigative techniques that have been reported as so problematic in the child interviewing literature. General Research Question #3: Types of Statements and Questions Used The literature on the types of questions child interviewers should use to obtain information from children is substantive. However, little research has been conducted on the use of the Step-Wise Interview protocol and, therefore, little is known about the questioning styles and types of questions that trained investigators use when following the protocol. The results presented in this section describe the percentages of statements and probes, general and specific questions, and the types of specific questions that were used by the officers in the first and second halves of the interviews Disclosure and No Disclosure interviews with young children. Data preparation. The only officer segment turn type categories that were collapsed for this analysis was the probe type; mixed and simple probes were combined to create a single category for probe turn type. As described in the method section of 80 this thesis, and shown in the appended flow chart (Appendix H), the probe type codes were collapsed into three question type categories: General questions which was comprised of the open-ended probes that required a multiple-word response; Specific questions which consisted of the wh_, multiple-choice, yes/no and if/than probes; and, Other which incorporated both the requests for repetition and the residual probes. The interviews were divided using a method similar to that in Underwager and Wakefield (1990). The total number of officer turns in each interview was tallied and divided in half to split the interviews. The frequencies for each segment type was tallied and divided by the number of interviews to obtain the mean percentages for the different types of officer segments (i.e., plain statements, tag question, probes, repetitions, and acknowledgements) that were used during each half of the interviews. Those mean percentages served as the dependent variables. The same method was used to obtain the mean percentages for the general, specific and other question categories as well as for the various specific question types. The data were not appropriate for use with inferential statistics so a ratio of 2:1 was used as the criterion for identifying substantial differences between the use of: 3(a) plain statements, tag questions, probes, repetitions, and acknowledgements; 3(b) general, specific and other questions; and, 3(c) specific wh_, yes/no, multiple-choice, and if/then questions. 3(a): Use of statements and probes. Officer turn type segments were analyzed to obtain baseline measures on the use of the various types of statements and the use of probes. The percentages of all officer turn type segments (plain, tag, probes, repetition, and acknowledgement) for the first and second halves of the interviews are presented in Table 8 (SDs are provided in Appendix 1). 81 Table 8 Percentage of Officer Turn Types for the First and Second Halves of the interviews with Young Children by Interview Status Officer Turn Types AcknowleInterview Half Plain Tag Probe Repetition dgement No Disclosure Interviews (n = 5) First Half 28.0% 5.0% 48.5% 10.6% 7.9% Second Half 27.8% 5.1% 43.2% 12.8% 11 .1% Disclosure Interviews (n = 7) First Half 19.1% 9.0% 49.3% 9.4% 13.2% Second Half 24.1% 4.3% 46.3% 8.6% 16.7% 82 The results showed , using the 2:1 ratio, that there was no difference between the use of plain statements and probes in either the first or second halves of the No Disclosure interviews. However, both plain statements and probes were used more often than tag questions, repetitions of the child's previous statement, and acknowledgements in the No Disclosure interviews. In the first half of the Disclosure interviews, there was a difference between the use of plain statements and probes, with probes being used more often . As in the No Disclosure interviews, plain statements and probes were used more often than tag questions, repetitions, and acknowledgements in both the first and second halves of the Disclosure interviews. However, in the Disclosure interviews, tag questions were used more often in the first half of the interviews than in the second half. Overall, the results showed that, regardless of interview status or interview half, the officers were probing for information almost as often as they were making some kind of statement. 3(b}: Use of general. specific and other questions. The rationale for evaluating the use of general, specific, and other types of questions during the first and second halves of the No Disclosure and Disclosure interviews was twofold: (a) Little is known about the use of general questions relative to the use of specific questions by officers trained in the Step-Wise Interview protocol, and (b) the protocol is very precise in its prescription for the use of specific questions only after a disclosure of abuse has been made. Table 9 shows the percentages of general, specific, and other questions types that were used during the first and second halves of the No Disclosure and Disclosure interviews (SDs are provided in Appendix J). 83 Table 9 Percentage of Officer General. Specific and Other Question Types For First and Second Halves of Interviews with Young Children by Interview Status Officer Question Types Interview Half General Questions Specific Questions Othe(i No Disclosure Interviews (n =5) First Half 1.6% 91.0% 7.4% Second Half 6.0% 88.1% 5.9% Disclosure Interviews (n = 7) First Half 3.2% 90.7% 6.1% Second Half 7.3% 87.8% 4.9% a Includes probes what were recorded as requests for repetition and unclassified (residual) questions. 84 The analysis using the 2:1 ratio to detect differences showed that in the No Disclosure interviews, specific questions were used more often than general question in both the first and second halves of the sessions. The same result for the higher use of specific questions, relative to general questions, was found for both the first and second halves of the Disclosure interviews. Comparisons between the use of general questions and 'other' questions (i.e., requests for repetition of the child's previous statement and residual probes) showed that in the first half of the No Disclosure interviews, general questions were used less often than those in the 'other' question category. In the Disclosure interviews, the use of general questions was lower in the first half than in the second half of those interviews. Overall, the results showed that regardless of interview status or interview half, between 87.8% and 91.0% of the questions asked were specific in nature. 3(c): Types of specific question used. Researchers have reported that the use of wh_ questions produces lower error rates in the responses obtained from children than yes/no and multiple-choice question types (Peterson & Briggs, 1997; Poole & White, 1991 ). Additionally, Yuille, Hunter, et al. (1993) prescribed that interviewers should avoid the use of multiple-choice questions as much as possible when using the StepWise Interview protocol. Analysis on the use of different types of specific questions by the trained officers was conducted and the results are shown in Table 10 (SDs are provided in Appendix K). A 2:1 ratio was used to determine differences in the officers' use of wh_, yes/no, multiple-choice, and if/then specific questions for both types of interviews. In the No Disclosure interviews, there were no differences found between the use of wh_ and yes/no questions in either the first or second halves of the interviews. However, wh_ 85 Table 10 Percentages of Officer Specific Question Types For First and Second Halves of Interview with Children Under 8 Years of Age by Interview Status Officer Specific Question Types MultipleInterview Half Wh_ Yes/No Choice lf!fhen No Disclosure Interviews (!1 = 5) First Half 50.2% 45.2% 4.1% 0.5% Second Half 49.4% 49.8% 0.0% 0.8% Disclosure Interviews (n = 7) First Half 52.0% 41.6% 3.8% 2.6% Second Half 49.7% 45.1% 2.4% 2.8% 86 and yes/no questions were both used more than multiple-choice and if/then questions in both halves of the No Disclosure interviews. Comparisons between the use of multiplechoice and if/then questions in the No Disclosure interviews showed that multiplechoice questions were used in the first half of the interviews but not in the second half. Additionally, multiple-choice questions were used more often in the first half of the No Disclosure interviews than if/then questions in either the first or second halves of those interviews. The pattern of wh_ and yes/no questions used in the Disclosure interviews was the same as in the No Disclosure interviews: No differences were found in the use of the two types of questions and both of those types of specific questions were asked more than multiple-choice and if/then questions. However, the use of multiple-choice and if/then questions differed in the Disclosure interviews from that seen in the No Disclosure interviews. No differences were found in the use of either type of question within or between the Disclosure interview halves. Comparisons showed that the use of if/then questions was higher in both the first and second halves of the Disclosure interviews than in the first and second halves of the No Disclosure interviews. Overall, the results showed that regardless of interview status or interview half, the officers were equally likely to ask wh_ questions as they were to ask yes/no questions and the use of multiple-choice and if/then questions was relatively low. Summary of Findings on the Use of Various Types of Statements and Questions in the First and Second Halves of the Step-Wise Interviews To provide an overall picture of the types of segments used by the officers, the results were initially reported for the percentages of each segment type (i.e., plain statements, tag questions, probes, repetitions and acknowledgements). The results 87 showed that during the first and second halves of both the No Disclosure and Disclosure interviews, the officers were probing for information just about as often as they were making some kind of statement. To further refine the analysis on the use of probes, the probe segments were categorized as either general, specific, or other type of question. The pattern that emerged for both the No Disclosure and Disclosure interviews was similar in that very few general questions were asked during the first or second halves of the interviews and the high use of specific questions, in both types of interviews, remained the same throughout the interview process. Taken together, these findings demonstrate that regardless of the interview status or interview half, almost one half of the officer segments were questions and that the use of specific questions far outweighed the use of general questions. The majority of the specific questions used by the officers, regardless of interview status or interview half, were of the wh_ or yes/no form; however, there were no differences found between the use of those two question types. This showed that the officers where just as likely to ask a less problematic wh_ question as they were to ask a more problematic yes/no question. The results showed that the officers set the tone by using specific questions early in the interview sessions and that few general questions were used in attempts to obtain free narrative responses from the children. The implications of the use of specific questions, and the impact they have on the investigative process are discussed in detail in the following section . Discussion Lamb and his colleagues (1 995) stated (cited in Poole and Lamb, 1998): The demonstrable fact that investigative interviews with young children can be rendered worthless by inept practice should not blind us to the substantial 88 literature demonstrating that reliable information can be elicited from young children who are competently interviewed. . .. This emphasis reflects our firm belief that the informativeness of interviews with child victims is strongly influenced by the skill and expertise of the interviewer and that the interviewer characteristics, unlike the characteristics and abilities of the child, can be (and must be) improved. (p. 34) Such comments from leading researchers in the field of child eyewitness testimony and child interviewing techniques led to the development of this research project. The purpose of the study was to determine whether investigators who were trained in the use of Step-Wise Interview, a reportedly effective child interviewing protocol, adhered to the structure and recommended questioning techniques prescribed by the protocol when they interviewed children in actual cases of alleged sexual abuse. The objectives of the Step-Wise Interview are to (a) minimize the trauma of the investigation for the child, (b) minimize the contaminating effects of the interview on the child's memory for the event(s), and (c) maximize the amount of information obtained from the child about the alleged events(s) (Yuille Hunter, et al., 1993; Yuille Marxsen, et al., 1993). To control for the potentially negative effects of the forensic context, the investigative interviews were conducted in casual, ~ i and non-intimidating environments. For example, the interview rooms were setup similar to a home living room with a couch, an armchair, and a coffee table. The interviewers were dressed in plain clothes and were trained to maintain a non-authoritative demeanor described by Poole and Lamb (1998) as "calm, unhurried, and accepting, with pauses to permit spontaneous additions by the child and opportunities for the interviewer to develop thoughtful questions" (p. 97). The casual physical environment, the relaxed demeanor. 89 of the interviewer, and coverage of the key elements associated with setting the parameters for testimony expectations are in place to mitigate the child witness's tendencies to acquiesce to social pressures and provide inaccurate testimony. Adherence to the structure of the Step-Wise Interview protocol was assessed by examining (a) coverage of the key elements associated with the various steps of the protocol and the use of terminology for body parts and functioning, (b) the amount of time spent in each step of the protocol, on the topics of body parts and functions and on topics unrelated to the prescribed protocol, and (c) the order in which each of the interview steps were introduced during the interviews. Adherence to recommended child questioning techniques was assessed by evaluating (a) the use of inappropriate misleading and leading questions and/or statements, and (b) the use of statements and probes, general and specific questions, and specific question types in the first and second halves of the interviews. The trained officers in this study covered most but not all of the key elements prescribed in the Step-Wise Interview protocol. Coverage of the key elements varied between the younger and older age groups and between interview status (No Disclosure and Disclosure) within the younger age group. In interviews with both the younger and older children, the officers did well in covering the various key elements associated with Step 1 of the protocol. However, the Step 2 key elements were covered by less than half of the officers in the No Disclosure interviews with the young children, with better coverage of most Step 2 key elements in the Disclosure interviews with the younger and older children. The Step 2 key elements related to discussing the difference between the truth and a lie and telling the child it is okay to say "I don't know", were covered more often in the Disclosure interviews than in the No Disclosure. 90 inteNiews. It is possible that covering those topics is an important factor developing a comfortable and trusting environment for the child making it more likely that he or she will disclose abuse if it did occur. Few officers attempted to elicit a non-abuse life event free narrative from the young children in the No Disclosure inteNiews and substantially fewer attempts to do so were obseNed in the Disclosure inteNiews with that age group. None of the officers attempted to elicit a non-abuse free narrative from the older children. The officers used both general and specific questions when introducing the topic of abuse but the use of specific questions was more prevalent with the younger children than with the older children. When the children made a disclosure of abuse, the majority of officers attempted to elicit a free narrative about the abuse event(s) from the younger children and all of the officers made the attempt with the older children. The key elements associated with closing the inteNiews were poorly covered although all of the officers who conducted the No Disclosure inteNiews with the younger children thanked them at the end of the session. Most of the officers used the body diagram during the inteNiews and when there were discussions related to body parts and functioning, all of the officers used the same terminology as the younger children and the majority of the officers did so with the older children. Some of the findings on coverage of key elements from this study were similar to those reported by Warren et al. (1996). Specifically, in both studies it was found that the investigators tended not to conduct the warm-up or practice sessions by asking general questions about non-abuse-related life events, and they often did not explain that it was okay for the child to answer with "I don't know" type responses. Additionally, in this study, it was found that the majority of officers who conducted the No Disclosure inteNiews skipped the parameter setting key 91 elements associated with discussions about truth/lie and only talking about things that really happened. The implications associated with missing key elements of the interview protocol are discussed in the next section of this thesis. In this study, the percentage of time spent in some of the interview steps varied, depending on the interview status, within the interviews with young children. In the Disclosure interviews, approximately the same amount of time (1 0.0% or approximately 2 min) was spent in each of the first three steps of the protocol. In the No Disclosure interviews, the time spent in the first two steps of the interviews was similar to that of the Disclosure interviews; however, in the No Disclosure interviews, approximately twice as much time was spent introducing the topic of abuse. The findings also showed that in the No Disclosure interviews, after introducing the topic of abuse, some officers moved on and asked specific abuse-related questions and they spent about the same amount of time in Step 5 as in Steps 1 and 2. In the Disclosure interviews, after the children made a disclosure of abuse, the officers spent about the same amount of time asking specific abuse-related questions as the officers in the No Disclosure interviews spent introducing the topic of abuse. The foregoing pattern indicates that, on average, the children who disclosed abuse did so within approximately 2 minutes once the topic of abuse was introduced, and then the officers moved on to Steps 4 (about 1 min) and 5 (about 6 min) of the protocol. When disclosure of abuse was not forthcoming the officers spent about 5 minutes in Step 3 and just under 2 minutes in Step 5. The time spent on discussions related to body part names and functions did not differ between the two types of interviews in this study, and the time spent on that topic was approximately the same as the time spent in Steps 1 and 2. Regardless of interview status, much of the interview time was spent on topics unrelated to the protocol (38.4% 92 in the No Disclosure and 31 .7% in the Disclosure interviews) and the time spent closing the interviews was less than 1 minute. Wood et al. (1996) reported that when interviewers had difficulty establishing rapport, they typically moved forward in the interview process and spent a great deal of time attempting to get the children to talk about any abuse events that may have occurred. As shown in this study, however, the difference between the time spent in the rapport building step did not differ between the No Disclosure and Disclosure interviews. What did differ was the time spent in Steps 3 and 5 with the officers in the No Disclosure interviews spending more time attempting to obtain a disclosure in Step 3 and less time asking questions about the alleged abuse events in Step 5. Goodman and Saywitz (1994) cautioned that the reliability of eyewitness reports from children could be compromised if hesitant preschoolers are pressured into answering questions by authority figures. The results from this study indicated that when there was no disclosure of abuse, the officers continued to ask questions related to the topic of abuse and the suspected abuse events. However, there is little known about the optimal amount of time that should be spent in any given step of the Step-Wise Interview protocol, or more generally, in the various interview stages that are recommended in the child interviewing literature. It is most likely that the optimal time would vary across children and investigators. Nevertheless, it is evident in this study that some No Disclosure interview time was spent in Step 5 which was a deviation from the prescription for use of the Step-Wise Interview protocol. Underwager and Wakefield (1990) reported that the 9 videotaped child sexual abuse interviews they reviewed lasted, on average, 36 minutes and they ranged between 15 and .50 minutes. Davies, Westcott, et al. (in press) reported the majority of 93 interviews they reviewed lasted, on average, between 40 and 49 minutes, and that the minimum and maximum amount of time for the entire sample was 20 and 90 minutes, respectively. In this study the No Disclosure interviews lasted, on average, about 21 minutes and ranged from approximately 11 to 50 minutes and the Disclosure interviews lasted, on average, about 23 minutes and ranged from approximately 13 to 34 minutes. Thus, the time the officers spent conducting the Step-Wise Interviews was similar to the duration of other reported interviews. In this study, regardless of interview status, the time spent in the rapport building step was about 2 minutes, or about 10.0% of the interview time. Warren et al., (1996) found that when the interviewers in their sample attempted to build rapport, the average proportion of interview time spent in that stage was 15.0% (based on total words for interviewer and child). Davies, Westcott, et al. reported that less than 10 minutes was spent building rapport in the majority of interviews they reviewed and pointed out that further studies on rapport are required to provide guidance on how it should best be handled. Overall, the majority of the interviews in this study were conducted in the correct Step-Wise order; 3 of the 5 No Disclosure and 5 of the 7 Disclosure interviews were conducted correctly. Of note was the introduction of Step 5 (specific abuse-related questioning) in 3 of the No Disclosure interviews and the skipping of Step 4 (eliciting an abuse-related free narrative) in 2 of the Disclosure interviews. The remaining prescribed interview steps were introduced in both the No Disclosure and Disclosure interviews but some were not used in the correct order; however such deviations were few. The implications of introducing non-prescribed steps (i.e., Step 5- specific abuserelated questioning in the No Disclosure interviews) and skipping prescribed steps are discussed in the next section of this thesis. 94 In this study, it appeared that the officers rarely made misleading comments that could potentially undermine the child's confidence in the interviewer or the investigative process, and the use of leading statements or questions was minimal. The low frequency of leading questions reported for this study parallel Davies, Westcott, et al. (in press) findings and the results from both studies are positive. They indicate that the forensic investigators avoided the use of some inappropriate interviewing techniques that could jeopardize the validity of the children's eyewitness testimony. These findings are important because the use of leading questioning can contaminate the child's statement and , therefore, contribute to legal decisions that may either allow a guilty person to avoid penalties for their actions or put an innocent person on trial for a crime they did not commit. The outcome of the investigative interview can also impact decisions made about to child protection issues. For example, if the alleged perpetrator was an immediate family member, the Ministry for Children and Families may decide to apprehend the child and place him or her in a safe living environment. As pointed out by numerous researchers (e.g. , Poole & Lamb 1998; Yuille, Hunter, et al., 1993), the credibility of the child's testimony becomes a very important issue because most often there are no third party witnesses or physical evidence to support the allegations of the sexual abuse. The results on the use of general and specific questions in this study were similar to the findings reported by Warren et al. (1996) , Wood et al. (1996) and Davies, Westcott, et al. (in press). Specifically, like Wood et al., this study found that the investigators established a pattern of asking specific and close-ended questions in the early stages of the interview session and, like Warren et al., Wood et al. and Davies, Westcott, et al., the interviewers seldom used general, open-ended questions. 95 Specifically, Warren et al. reported that "few general, open-ended questions were asked during any portion of the interviews" (p. 239), and that of all questions in the abuserelated portions of the interviews, 10.5% were general questions. In this study, the use of general questions in the interviews with the young children ranged from a low of 1.6% in the first half of the No Disclosure interviews to a high of 7.3% in the second half of the Disclosure interviews. Davies, Westcott, et al. reported that in interviews with children who were under 8 years of age, 1.0% of the questions were general (i.e., open-ended), 59.0% were specific yet non-leading, and 40.0% were closed-ended. In this study, about 91.0% of the questions asked in the first halves of the No Disclosure and Disclosure interviews with the young children were specific questions and about 88.0% of the questions in the second halves of those interviews were specific in nature. Observations on the types of specific questions used by interviewers, as reported by Warren et al. and Wood et al., suggested that the more problematic closed-ended yes/no type of specific question may be used more often than the less problematic open-ended wh_ type question. This was not the case in this study as the results showed that regardless of interview status or interview half, the officers were just as likely to ask the young children wh_ questions as yes/no questions. Additionally, in this study, the use of multiple-choice and if/then questions was low relative to the use of wh_ and yes/no questions. Nevertheless, it does appear that interviewers, regardless of whether or not they are trained in the use of the reportedly effective child interviewing techniques, tend to rely on the use of specific questions to obtain information from young children and the use of specific questions begins early in the interviewing process. The use of specific questions can be problematic and the implications of doing so are discussed in the next section of this thesis. 96 The results on the trained RCMP investigators' adherence to the structure of Step-Wise Interview protocol, and the questioning techniques used in the Step-Wise Interviews, were good in many respects. Many of the key elements of the protocol were covered by most of the officers, the duration of the interviews and the time spent in the various steps of the protocol were typical of reports in the literature, and the steps of the interview protocol were introduced in the correct order in the majority of the interviews. The officers' use of general questions and their reliance on the use of specific questions parallel reports in the literature and their use of wh_ questions relative to yes/no questions appeared to be more positive than results reported elsewhere. However, the deviations from the prescribed protocol that were observed do have implications and those are discussed below. Implications During the initial steps of the interview protocol, investigators are to set the ground rules or parameters for testimony expectations and to clarify that the child understands those expectations. Because children tend to provide responses to questions asked by adults (Poole & Lamb, 1998), not covering the key elements associated with parameter setting could lead to children spontaneously answering questions that they did not understand or those for which they actually did not possess the knowledge required to respond correctly. Thus, children could inadvertently confirm or disconfirm abuse-related information that was critical to the investigation. Additionally, when an investigator begins the interview by asking specific questions, he or she may unwittingly be setting an example for the types of questions and responses that are expected in later steps the interview (Wood et al., 1996). The child's expectation to respond with confirmatory and/or short answer responses would be 97 further reinforced when the investigator continued to use specific questions while attempting to build rapport and when introducing the topic of abuse. In everyday contexts, adult speech to children is often simplified, and based on such experience, children likely enter interviews with the expectation of being asked a series of specific questions (Poole & Lamb, 1998). As research confirms, forensic interviewers often behave in accordance with those expectations (Davies, Westcott, et al., in press; Poole & Lamb; Warren et al., 1996; Wood et al., 1996). In addition to children's tendency to simply provide some kind of answer to an adult's question, research has shown that children's responses to specific questions are much less accurate than their responses to general questions (Poole & Lamb, 1998). Specific questions may relate to memories that are weakened by insufficient encoding (Brainerd & Ornstein, 1991) or to script memories (Nelson, 1986) which increase the risk of response error. Especially problematic is the use of specific questions that are framed as forced-choice and yes-no probes for information (Poole & Lamb). Overall, the use of specific questions generally impairs performance, although older children, relative to younger children, are better able to resist the implicit social pressure to respond. Children, whether older or younger, tend to provide more accurate answers to specific questions when target events are extremely salient or memorable and when the questions imply something that violates expectations about what might reasonably have happened (Poole & Lamb). When the free narrative prompt for information about a non-abuse life event is skipped, children are not provided with the free narrative warm-up or practice session that is expected during the abuse-related narrative step of the interview. Additionally, not obtaining narrative responses from the child could inhibit the investigator from 98 accurately assessing the child's developmental level and linguistic competency. Attempting but failing to obtain a free narrative about a non-abuse life event would provide the investigator with important information about the child's ability to provide an abuse-related narrative later in the interview session. It must be remembered that "researchers consistently have shown that children provide fewer details in response to open-ended questions than in response to a series of specific questions" (Poole & Lamb, 1998, p. 52) . Therefore, it is important that investigators follow-up on narrative accounts by asking a series of carefully planned specific questions while being aware that the children's responses to those questions may be much less accurate than their responses to the open-ended questions. If the investigator skips key elements associated with closing the interview the child may be left feeling confused about the investigative process. The child may be left with unanswered questions and, furthermore, with a sense that his or her statement was not important or appreciated. Not closing the interview in the informative and supportive way recommended in the literature may play a role in inhibiting the child from wanting to provide additional information about disclosed abuse-related events, or from wanting to make a disclosure at a later date. Recommendations The recommendations put forth are based on the observations from this study and on findings reported by researchers such as Davies, Marshall, et al. (1998), Davies, Westcott, et al. (in press), Poole and Lamb (1998), Warren et al. (1996), and Wood et al. (1996) . However, the recommendations must be taken cautiously because the sample used for this study was quite small and the researcher was unable to control for 99 or evaluate the impact of factors such as gender, the children's' emotional, cognitive, or motivational states, or the investigators' adherence to the prescribed protocol. Since "information presented in lecture format rarely promotes significant behavior change" (Poole & Lamb , p. 240) , child interviewing training sessions should provide opportunities for practice and critical feedback on performance. Interviewing skills develop gradually and , therefore, the techniques must be practiced repeatedly over time. "Systematic feedback is the key to successful training" (Poole & Lamb, p. 241). Training and practicum experience. It is not uncommon for child interviewing training programs to include a role-playing module in which the participants practice the investigative protocol. However, the participants often interview one another and the training time devoted to role-playing may be relatively limited. To optimize training, the participants should be provided with the opportunity to interview children (Davies, Marshall, et al. , 1998; Poole & Lamb, 1998). Interviewing adult cohorts will not provide trainees with the challenges that they will inevitably encounter when interviewing children and , therefore, having trainees interview children about a staged event would be most beneficial (Davies, Marshall, et al.) . Additionally, sufficient time must be allocated to interviewing practice so that trainees have the opportunity to both learn the process and experience some of the difficulties that arise when interviewing children (Wood et al., 1996). Davies, Marshall, et al. reported that as much as 30% to 40% of the total training time was suggested as a reasonable figure for interviewing practice. Justification for these recommendations is found in the results reported in this thesis. For example, although it is not possible to determine why the majority of the officers did not make attempts to have the children provide narrative accounts of non-abuse life 100 events, attempts to do so were extremely low with the younger children and nonexistent with the older children. Training and practice with children of various ages would provide interviewers with (a) experience in framing questions in ways that promote narrative responses, and (b) a broad understanding of the quality and quantity of the narrative responses children can provide to such prompts for information. Constructive criticism and ongoing feedback. It is recommended that both trainees and active child abuse investigators receive feedback on their interviewing techniques (Davies, Marshall, et al., 1998; Poole & Lamb, 1998; Wood et al., 1996). The practice interview sessions during training should be videotaped and then reviewed by the instructor with the trainee. Being able to view the interview while receiving performance feedback would provide the trainee with immediate and pertinent information about his or her interviewing strengths and weaknesses. Early identification of potential problem areas would assist the trainee in learning how to overcome them and would also reduce the likelihood of the problems becoming 'ingrained bad habits'. It is also recommended that trained investigators be provided with ongoing opportunities to review and obtain feedback on their videotaped interviews with children involved in actual abuse cases (Davies, Marshall, et al.; Poole & Lamb; Wood et al.). This process would provide the investigators with the opportunity to discuss their interviewing techniques and to obtain current information with regards to developments in the field of child interviewing. Davies, Marshall, et al. reported that the officers they interviewed recognized the need for formal feedback and that some police agencies routinely sampled videotaped child interviews to assess interviewer standards. A number of results obtained in this study provide justification for the recommendation that interviews receive constructive and ongoing feedback on their 101 performance. For instance, some officers in the No Disclosure interviews asked specific abuse-related questions, which according to the Step-Wise Interview protocol, were not prescribed. Reviewing the videotapes of the interviews and discussing the nature of the abuse-related questions with an instructor would assist the officers in recognizing the types of questions which lead them to enter Step 5 of the protocol. Coverage of the Step 6 key elements of the Step-Wise Interview protocol was poor. Reviewing the videotaped interviews would bring this matter to the interviewers' attention and feedback from an instructor would reinforce the importance of covering the key elements associated with closing investigative interviews with children. Refresher courses. Investigators who deal with child protection and who interview children need to be kept up-to-date on developments in those areas (Poole & Lamb, 1998; Warren et al., 1996). It is recommended that refresher courses and/or information seminars be provided for trained personnel. The sessions could address topics such as operational issues, forensic or special interest research findings, and changes to the judicial process (Davies, Marshall, et al., 1998). Davies, Marshall, et al. viewed ongoing training sessions as essential for two important reasons: (a) the dissemination of valuable information, and (b) providing officers with the opportunity to share the experiences they face in the line of duty. A review of findings from field studies on child interviewing practices clearly shows that investigators rely on specific questions to obtain information from young children and that the use of specific questions begins early in the interview sessions. In this study, the results showed that the officers followed the same trend in the use of specific questions as interviewers in other studies. The high use of specific questions reported in this and other studies may be indicative of the problems the investigators 102 encountered with regards to obtaining comprehensive free narrative accounts about the abuse-related events from the young children. Thus, the results from this thesis support the recommendation for refresher courses. Having the officers attend a refresher course would provide them with the opportunity to learn more about current findings on the types of response errors that are associated with different types of specific questions. It would also give the officers an opportunity to discuss among themselves, and with researchers and instructors, why the use of specific questions seems so prevalent in interviews with children. Limitations of the Current Study Cohen's kappa was used as a measure of interobserver agreement and Fleiss' (1981) benchmarks were used to assess the relative strength of agreement. When interobserver agreement (according to Fleiss' benchmarks) was poor, additional analysis was conducted to determine whether the observer accuracy was acceptable. The interobserver agreements were all satisfactory except for the key element associated with the officers' attempts to elicit a non-abuse life event free narrative from the children. Further evaluation of the low kappa for the life event variable showed that observer accuracy was acceptably high and, therefore, meaningful interpretations of the results for that key element could be made. Limitations to the interpretation of results for misleading and leading officer segments are a result of the low frequencies of their occurrence in this study. Thus, with the exception of misleading and leading variables, any limitations to the generalizability of the findings from this study are not a result of unsatisfactory interobserver agreement. The small sample sizes for both the number of children interviewed and the number of investigative officers who conducted the interviews limits the generalizability 103 of the result obtained in this study. The fact that the data were drawn from interviews conducted by 8 officers also limits the results because some of the officers interviewed more than one child. Nevertheless, results similar to the ones reported in this thesis have been found by researchers whose sample sizes were substantially larger (e.g. , Davies, Westcott, et al., in press, N = 36 and Warren et al. , 1996, N = 42). Based on the supportive evidence provided by those studies, the findings of this study were not considered atypical. Suggestions for Future Research Although promising structured child interview protocols have been developed, to date, rigorous testing of the protocols with children from the population of interest is sparse (Poole & Lamb, 1998). Researchers have not yet demonstrated that any one type of interviewing protocol is reliable and valid for use with children of various ages. Researchers are reporting that when interviewing children, rather than using general questions to obtain information, investigators typically use specific questions throughout the session (Davies, Westcott, et al., in press; Warren et al., 1996; Wood et al., 1996). Future research should focus on determining whether the level of cognitive development has a moderating influence on the value and utility of general and specific questions. Perhaps carefully planned and appropriately worded specific questions are required when interviewing young children and children who have cognitive deficits. Additionally, researchers have demonstrated that even with training, investigators often do not follow the recommended guidelines and protocols developed for interviewing children (Geiselman et al., 1993; Davies, Westcott, et al). Future research should be conducted to determine how much training and experience is required for investigators 104 to learn and adhere to prescribed and recommended interviewing techniques and processes. Conclusions The results obtained in this study support the conclusions reported by researchers who are investigating child interviewing and the adherence to structured interview protocols (Davies, Marshall, et al., 1998; Davies, Westcott, et al., in press; Geiselman et al., 1993; Warren et al., 1996; Wood et al., 1996). Specifically, in this study, adherence to the structure of the Step-Wise Interview protocol was low in some very key areas. Evaluations on the use of inappropriate questioning techniques were positive in that the use of misleading and leading statements and questions was very low. However, examination of the types of questions used showed that few general questions were asked and that the use of specific questions was consistently high. Findings from this study, and similar field studies (Davies, Westcott, et al.; Underwager & Wakefield (1990); Warren et al.; Wood et al.), may imply the need for a more practical approach to conducting investigative interviews with children. It is quite possible that somewhat different techniques are required to elicit accurate information from children depending on their age, developmental level, linguistic ability and emotional state. Although the effective child interviewing literature espouses the benefits associated with using general questions, the use of well planned specific questions may be necessary to assist younger children with focusing their attention on the matter at hand (Poole & Lamb, 1998). Appropriately phrased specific questions contain cues that trigger retrieval of details that young children may not otherwise remember. However, how to balance and optimize the use of general and specific questions with children of different ages will not be known until researchers establish which types of interviewing methods 105 lead to the accurate disclosure of actual abuse. As Poole and Lamb pointed out, "we cannot currently specify which [interviewing] techniques are in the best interests of particular children" (p. 71 ). Therefore, at this point in time, no single child interviewing method can be considered as the gold standard. 106 References Anderson, J. R. (1990). Cognitive psychology and its implications (3rd ed.). New York: W. H. Freeman. Baddeley, A. (1990). Human memory: Theory and practice. Boston: Allyn and Bacon. Bakeman, R., & Gottman, J. M. (1997). Observing interaction: An introduction to sequential analysis (2nd ed.). New York: Cambridge University Press. Bakeman, R., Quera, V., McArthur, D., & Robinson, B. F. (1997). Detecting sequential patterns and determining their reliability with fallible observers. Psychological Methods. 2, 357-370. Brainerd, C., & Ornstein, P. A. (1991 ). Children's memory for witnessed events: The developmental backdrop. In J. Doris (Ed.), The suggestibility of children's recollections: Implications for eyewitness testimony (pp. 10-20). Washington, DC: American Psychological Association. Bruck, M., Ceci, S. J., Francoeur, E., & Barr, R. (1995). I hardly cried when I got my shot! Influencing children's reports about a visit to their pediatrician. Child Development, 66, 193-208. Cairns, H. S., & Cairns, C. E. (Eds.). (1976). Psycholingustics: A cognitive view of language. New York: Holt, Rinehart and Winston. Ceci, S. J. (1991 ). Some overarching issues in the children's suggestibility debate. In J. Doris (Ed.), The suggestibility of children's recollections: Implication for eyewitness testimony (pp. 1-9). Washington, DC: American Psychological Association. 107 Ceci, S. J., & Bruck, M. (1993a) . Child witnesses: Translating research into policy. Social Policy Report: Society for Research in Child Development, 7, 1-30. Ceci , S. J. , & Bruck, M. (1993b). Suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113, 403-439. Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific analysis of children's testimony. Washington, DC: American Psychological Association. Ceci, S. J., Leichtman, M.D., & Putnick, M.E. (Eds.). (1992). Cognitive and social factors in early deception. Hillsdale, NJ: Erlbaum . • Ceci , S. J., Ross, D. R., & Toglia, M. P. (1987) . Suggestibility in children's memory: Psycholegal implications. Journal of Experimental Psychology, 116, 38-49. Davies, G., Marshall, E., & Robertson, N. (1998). Child abuse: Training investigating officers (Police Research Series Paper No. 94). London: The Home Office. Davies, G. , Westcott, H., & Horan, N. (in press). The impact of questioning style on content of investigative interviews with suspected child sexual abuse victims. Psychology, Crime and the Law. Dent, H. R. (1991 ). Experimental studies of interviewing child witnesses. In J. Doris (Ed.), The suggestibility of children's recollections: Implications for eyewitness testimony (pp. 138-146). Washington, DC: American Psychological Association. Doris, J. (Ed.). (1991 ). The suggestibility of children's recollections: Implication for eyewitness testimony. Washington, DC: American Psychological Association. Dworetzky, J.P. (1987). Introduction to child development (3rd ed.). St. Paul, MN: West Publishing. 108 Faller, K. C. (1988) . Child sexual abuse: An interdisciplinary manual for diagnosis. case management, and treatment. New York: Columbia University Press. Fivush, R., Hamond, N. R., Harsch, N., Singer, N., & Wolf, A. (1991 ). Content and consistency in young children's autobiographical recall. Discourse Processes. 14. 373-388. Fivush, R., & Hudson, J. (Eds.). (1990). Knowledge and remembering in young children. New York: Cambridge University. Fleiss, J. L. (1981 ). Statistical methods for rates and percentages. New York: Wiley. Geiselman, R. E., & Fisher, R. P. (1988). The cognitive interview: An innovative technique for questioning witnesses of crime. Journal of Police and Criminal Psychology, 2, 2-5. Geiselman, R. E., Fisher, R. P., Firstenberg, 1., Hutton, L.A., Sullivan, S. J., Avetissian, I. V., & Prosk, A. L. (1984). Enhancement of eyewitness memory: An empirical evaluation of the cognitive interview. Journal of Police Science and Administration, 12, 74-80. Geiselman, R. E., Saywitz, K. J., & Bornstein, G. K. (1993). Effect of cognitive questioning techniques on children's recall performance. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims. child witnesses: Understanding and improving testimony (pp 71-93). New York: Guilford. Goodman, G. S., & Bottoms, B. L. (Eds.). (1993). Child victims. child witnesses: Understanding and improving testimony. New York: Guilford. 109 Goodman, G. S., & Clarke-Stewart, A. (1991 ). Suggestibility in children's testimony: Implications for sexual abuse investigations. In J. Doris (Ed.), The suggestibility of children's recollections: Implication for eyewitness testimony (pp.92105). Washington, DC: American Psychological Association. Goodman, G. S., Rudy, L., Bottoms, B. L., & Aman, C. (1990). Children's concerns and memory: Issues of ecological validity in the study of children's eyewitness testimony. In R. Fivush & J. Hudson (Eds.) , Knowledge and remembering in young children (pp. 249-284). New York: Cambridge University Press. Goodman, G. S., & Saywitz, K. J. (1994). Memories of abuse: Interviewing children when sexual victimization is suspected. Child and Adolescent Psychiatric Clinics of North America. 3, 645-661. Hewlett, M. G., & Hardy, C. (1997). Coding procedures for the Step-Wise Interview. Unpublished coding manual, University of Northern British Columbia, Prince George, British Columbia, Canada. Karmiloff-Smith, A. (1979). Language development after five. In P. Fletcher & M. Garman (Eds.), Language Acquisition: Studies in first language development (pp. 307-323). New York: Cambridge University Press. King, M.A., & Yuille, J. C. (1987). Suggestibility and the child witness. InS. J. Ceci, D. R. Ross, & M. P. Toglia (Eds.), Children's eyewitness memory (pp. 24-35). New York: Springer-Verlag. Lamb, M. E., Hershkowitz, 1., Sternberg, K. J., Esplin, P. W., Hovav, M., Manor, T., & Yudilevitch, L. (1996). Effects of investigative utterance types on Israeli children's responses. International Journal of Behavioral Development, 19, 627-637. 110 Lamb, M. E., Sternberg, K. J. , & Esplin, P. W. (1994). Factors influencing the reliability and validity of statements made by young victims of sexual maltreatment. Journal of Applied Developmental Psychology, 15, 255-280. Leippe, M. R., Manion, A. P., & Romanczyk, A. (1993). Discernibility or discrimination?: Understanding jurors' reactions to accurate and inaccurate child and adult eyewitnesses. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving testimony (pp 169-201 ). New York: Gilford Press. McGough, L. S., & Warren, A. R. (1994). The all-important investigative interview. Juvenile and Family Court Journal. 45, 13-29. Nairne, J. S. ( 1997). The adaptive mind. Pacific Grove, CA: Cole Publishing Company. Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Erlbaum Perry, N. W., & Wrightsman, L. S. (1991). The child witness: Legal issues and dilemmas. Newbury Park, CA: Sage. Peterson, C., & Bell, M. (1996). Children's memory for traumatic injury. Child Development, 67, 3045-3070. Peterson, C., & Briggs, M. (1997). Interviewing children about trauma: Problems with "specific" questions. Journal of Traumatic Stress. 10. 279-290. Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. 111 Poole, D. A., & White, L. T. (1991 ). Effects of question repetition on the eyewiness testimony of children and adults. Developmental Psychology. 27, 975-986. Powell , M. B., & Thomson, D. M. (1994). Children's eyewitness-memory research: Implications for practice. Families in Society: The Journal of Contemporary Human Services. 41. 204-215. Raskin, D. C., & Yuille, J. C. (1989). Problems in evaluating interviews of children in sexual abuse cases. InS. J. Ceci, D. F. Ross, & M. P. Toglia (Eds.). Perspectives on children's testimony (pp. 184-207). New York: Springer-Verlag. Robinson, B. M., & Bakeman, R. (1998). ComKappa: A Windows '95 program for calculating kappa and related statistics. Behavior Research Methods. Instruments. and Computers. 30, 731-732. Saywitz, K. J. (1995) . Improving children's testimony: The question, the answer, and the environment. In M. S. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 113-139). Newbury Park, CA: Sage. Saywitz, K. J. , & Snyder, L. (1993). Improving children's testimony with preparation. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims. child witnesses: Understanding and improving testimony (pp. 117-146). New York: Guilford. Steller, M. (1991 ). Commentary: Rehabilitation of the child witness. In J. Doris (Ed.), The suggestibility of children's recollections: Implication for eyewitness testimony (pp. 106-109). Washington, DC: American Psychological Association. Sternberg, K. J. , Lamb, M. E., Hershkowitz, 1., Esplin , P. W., Redlich, A., & Sunshine, N. (1996). The relation between investigative utterance types and the 112 informativeness of child witnesses. Journal of Applied Developmental Psychology, 17, 439-451. Underwager, R., & Wakefield, H. (1990). The real world of child interrogations. Springfield, IL: Charles C. Thomas. Walker, A. G. (1993). Questioning young children in court: A linguistic case study. Law and Human Behavior. 17. 59-81. Warren, A. R., Woodall, C. E., Hunt J. S., & Perry, N. W. (1996). "It sounds good in theory but. .. " Do investigative interviewers follow guidelines based on memory research? Child Maltreatment. 1. 231-245. Wood, J. M., McClure, K. A., & Birch, R. A. (1996). Suggestions of improving interviews in child protection agencies. Child Maltreatment. 1, 223-229. Yuille, J. C. (1988). The systematic assessment of children's testimony. Canadian Psychology, 29, 247-262. Yuille, J. C., Hunter, R., Joffe, R., & Zaparniuk, J. (1993). Interviewing children in sexual abuse cases. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims. child witnesses: Understanding and improving testimony (pp. 95-115). New York: Guilford. Yuille, J. C., Marxsen, D., & Menard, K. (1993). Penultimate draft of the report to the Ministry of Social Services of British Columbia of the Systematic Interviewing project. Ministry of Social Services Research and Valuation Branch, Victoria, British Columbia, Canada. Zaragoza, M. S. (1991 ). Preschool children's susceptibility to memory impairment. In J. Doris, (Ed.) The suggestibility of children's recollections: Implication 113 for eyewitness testimony (pp. 28-39). Washington, DC: American Psychological Association. Zaragoza, M. S., Graham, J. R., Hall, G C. N., Hirschman, R., & Ben-Porath, Y. S. (Eds.). (1995). Memory and testimony in the child witness. Newbury Park, CA: Sage. Appendix A 114 THE 6 STEP INTERVIEW GOALS INTRODUCTION RAPPORT BUII.DING INTRODUCE THE TOPIC - explain role - - relax the child - assess developmental level - discuss the "I don't know" answer - elicit information EXAMPLES "My name is and I'm a social worker/police officer. Our job is to talk with children and their families." "Tell us about your last birthday." "We're only going to talk about things that really happened." "If you don't know you can tell us you don't know." "Do you know why we're talking today?" -focus the child's "Has anything happened to attention you that you'd like to tell me about?" - provide a context for the child to narrate "Tell us about who lives with you/looks after you." information "Tell us about the people you like/don'tlike." "Tell me about it." "Uh-huh." "Where were you?" "What happened next?" FREE NARRATIVE - the child to provide information UNINTERRUPTED SPECIFIC QUESTIONS - clarify/extend the information provided - follow up on inconsistencies "Whose house was it?" "What colour was the car?" "Can you help me understand this?" "Who knows about this?" -answer child's questions - thank and reassure the child - explain what will happen next "Do you have any questions for us?" "Thanks for talking with us - we may need to talk with you again." "We're going to talk to now." protection Issue MUST be resolved CLOSURE Appendix B 115 THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA 3333 University Way Prince George, B.C. Canada V2N 4Z9 Tel: (604) 960-5555 Fax: (604) 960-5794 August, 1996 Dear Officer, We are conducting an evaluation of the interviewing techniques RCMP officers use with children during investigations of alleged child sexual abuse. To carry out this project it is necessary that we view a number of archived videotaped interviews with child witnesses. Records indicate that videotapes of interviews you conducted with child witnesses are archived under the care and control of RCMP Staff Sergeant Wayne Roberts. We are asking your permission to use one or more of these videotapes in our study. If you give us your permission to view one or more videotapes of your interviews, we will then contact legal guardians of the children on the videotapes and ask their permission as well. Both you and the legal guardian must give permission to use a given videotape before it can be included in the study. The main goal of this study is to evaluate current interviewing practices. The final report will highlight current areas of strength and identify areas in which training might be improved. In The final report, only group data will be reported and on individual officer or child will be identified. There is a place on the consent form for you to indicate whether you would like to receive a copy of the final report. Please complete the attached consent form, seal it in the enclosed envelope, and send it to Staff Sergeant Wayne Roberts. He will collect the completed forms and forward them to me. If You have any questions or concerns about this project, please contact me at the number given below. Thank you for your interest in our study. Sincerely, Cindy Hardy, M.A. Psychology 960-5814 116 RCMP OFFICER'S CONSENT FORM (ONE) Name of Officer: - - - - - - - - - - - - - - - - - (please print) I understand that Professor Cindy Hardy is doing a study to evaluate interviewing techniques used with children during investigations of child sexual abuse. If I agree to participate in this study, videotapes of one or more of the interviews I have conducted with child witnesses will be viewed by the research team. I understand that confidentiality will be protected, all reports will be based on group information, and on individual will be identified in reports. -----------------------------------------------Please check one of the following : _ I AGREE to participate. One or more of the interviews I have conducted with child witnesses can be used in the study. _ I PROVISIONALLY AGREE to participate. One or more of the interviews I have conducted with child witnesses can be used in the study but I want to be told which interviews will be used and may withhold consent for use of particular interviews. (If you choose this option we will contact you again once we have obtained consent from legal guardians. At that time, you will be asked to complete a consent form like the one shown on the next page.) _ I DO NOT AGREE to participate in the study. None of the interviews I have conducted with child witnesses can be used in the study. Signature: - - - - - - - - - - - - - - - - Date: _ _ _ _ _ _ _ _ _ __ -----------------------------------------------___ Check here if you would like to receive a copy of the final report from this project. ================================================ If you agree or provisionally agree to participate, please complete the following questions. Have you attended an inter-agency joint training session on interviewing children? YES NO If yes, who conducted the workshop? (check one) _ _ Staff Sergeant Wayne Roberts _ _ Dr. John Yuille, University of British Columbia _ _ Other (please s p e c i f y - - - - - - - - - - - - -.) Unknown If yes, what was the approximate date of the training session? - - - - - - - - - - - - - 117 RCMP OFFICER'S CONSENT FORM (TWO) Name of Officer: - - - - - - - - - - - - - - - - - Dear Officer, When we contacted you earlier, you provisionally agreed to participate in our evaluation of interviewing techniques used with children during investigations of alleged child sexual abuse. You withheld full consent pending our notification of which videotaped interview(s) we would use in the study. The legal guardian of the child named below has given us permission to use the following interview. We are now asking you for your permission to use the interview. ================================================ Clinic Tape N u m b e r : - - - - - - - - - - - - - - - - Date of Interview: Please check one of the following: _ _ I AGREE to participate. You may use the interview identified above. _ _ I DO NOT AGREE to participate. You may not use the interview identified above. Signature: - - - - - - - - - - - - - - - Date: _ _ _ _ _ _ _ _ __ ================================================ Appendix C 118 TELEPHONE SCRIPT Hello, my name is _ _ _ _ _ _ _ _ _ and I am a Victim Services volunteer. Are you the parent (legal guardian of ______ ? (name of child) (If YES, continue as outlined below. If NO, ask for parent and/or check whether you have the correct number). I am calling because your help is required. The Prince George RCMP are involved in an evaluation of the interviewing procedures they use with children. This evaluation project is being conducted by Professor Cindy Hardy of the Psychology Program at UNBC. In order to carry out this project, the researchers must view videotapes of interviews that were conducted by RCMP officers. I am calling to ask for you permission to use the videotaped interview of your child that is already on file in our office here in Prince George. Before you can decide whether you want to give your permission, there are a few things you need to know about this evaluation project. 1. This project does not require further contact with your child. 2. Confidentiality will be fully protected in this evaluation project. Each videotape will be assigned an identification number so that your child's name will not be used. You can be assured that no records will be produced using your child's name or any other identifying information. 3. The researchers are not interested in the details you child reported during the Interview. They are only interested in the procedures the police officer used during the interview. 4. The goal of the project is to find ways to improve the interviews RCMP officers do with children. Your assistance with this project may benefit other children who must be interviewed by police. Do you think you might be willing to give permission for researchers to view the videotape of the interview with your child? If parent says NO: Record that permission was denied and thank them for their time. If parent says YES and lives in Prince George, make arrangements to meet parent (see below). If parent says YES and lives outside Prince George, explain that a local RCMP officer or Victim Services worker will contact them (see below). 119 For families in Prince George We need written permission from you to use the videotape of your child's interview in the study. Can we find a time to meet, either at your home or elsewhere, so I can deliver a letter qescribing the study and get your written consent to use the videotape? (Set time and date, check address, and give tern a number where they can reach you in case they have to change the meeting time. Ask them whether they would like you to call the night before the scheduled meeting to confirm. Thank them for their time) For families outside Prince George We need written permission from you to use the videotape of your child's interview in the study. I will arrange for a local RCMP officer or Victim Service worker to contact you. He/She will meet with you to deliver a letter describing the study and get your written consent to use the videotape. (Ensure that they are comfortable with being contacted by a local RCMP representative. · Thank them for their time.) Appendix D 120 THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA 3333 University Way Prince George, B. C. Canada V2N 4Z9 Tel: (604) 960-5555 Fax: (604) 960-5794 August, 1996 Dear Parent, , from RCMP Victim Services, is forwarding this letter to you on my behalf. In collaboration with the Prince George RCMP, I am conducting a research study to evaluate the interviewing procedures RCMP officers use with children. In order to carry out this project, I must view videotapes of interviews that were conducted by RCMP officers. I am writing to ask for your permission to use the videotaped interview of your child that is already on file with the RCMP in Prince George. Before you can decide whether you want to give your permission, there are a few things you need to know about this evaluation project. 1. This project does not require further contact with your child. 2. Confidentiality will be fully protected in this evaluation project. Each videotape will be assigned an identification number so that your child's name will not be used. You can be assured that no records will be produced using your child's name or any other identifying information. 3. I am not interested in the details you child reported during the interview. I am only interested in the procedures the police officer used during the interview. 4. The goal of the project is to find ways to improve the interviews RCMP officers do with children. Your assistance with this project may benefit other children who must be interviewed by police. When you spoke with < name of Victim services worker> recently, you indicated that you might be willing to give me permission to use the videotape of your child being interviewed by an RCMP officer. Before the videotape can be used in the study, I need your written permission. Please complete the attached consent form. If you have questions or concerns about the study, please contact me at the number given below. Thank you for your interest in the study. Sincerely, Cindy Hardy, M.A. Psychology 960-5814 121 PARENT CONSENT FORM Name of Parent/Legal Guardian: (please print) Clinic Tape N u m b e r : - - - - - - - - - - - - - - - - - - - Date of Interview: -----------------------------------------------I understand that confidentiality will be fully protected in this study and that there will be no contact with my child. Please check on of the following: _ I AGREE to participate. The existing videotaped interview with my child can be used in the study. _ I DO NOT AGREE to participate in the study. The existing videotaped interview with my child can not be used in the study. Signature:---------------- Date: _ _ _ _ _ _ _ _ __ ================================================ If you received this letter by mail, please send this page by return mail as soon as possible. Thank you for your assistance. Appendix E 122 Coding Procedures for the Step-Wise Interview Maureen G. Hewlett and Cindy Hardy University of Northern British Columbia December 10, 1997 Overview Coding requires three full viewings of each videotaped police interview as well as a review of the transcript for each interview. The sequence to be followed is: Pass 1: Watch the videotape, time the interview, and read the transcript of the interview to be sure it is accurate. Coders will work together for this pass only to ensure they agree on the accuracy of the transcript and to make any changes to the transcript that may be required. Pass 2: Watch the videotape a second time and rate the interview on global processes, which are referred to as key elements, that are to occur in of each of the six phases of the Step-Wise interview. Coders will work independently to assess interobserver agreement. Pass 3: Watch the videotape a third time and use the transcript for clarification. Coding in this pass will be done on a tum-by-tum basis and will focus primarily on the officers' interviewing techniques. Coders will work independently to assess interobserver agreement. The procedures and relevant definitions for each of the steps outlined above are detailed below. 123 Pass 1: Comparing Videotape with Transcript For Pass 1 the coders will work together. They will record the length (in minutes) of the interview and will correct any inconsistencies between the verbal and written records of the interview. If the written transcript does not match the dialogue on the videotape, the coders will replay the videotape until they are confident of what is being said and then correct the written transcript so it reflects the verbal content on the videotape. For this coding system, the word "segment" is defined as "a communication bounded by a conventional punctuation mark such as a period, question mark, or exclamation mark". All segment analyses will be conducted from the written transcripts once the transcripts have been checked for accuracy and corrections, if any, have been made. The transcripts are provided by the RCMP and are generally very accurate so few revisions are anticipated. In some cases, words like "Okay" may be typed as stand alone segments in the transcript when they should be included in a longer tagtype segment. In such situations, coders will make small editorial changes to the transcript during Pass 1. Segment numbers will be assigned to each segment uttered by the officer and the child. If each line of the written transcript has been numbered sequentially the coder may have to assign a decimal point subcode for lengthy communications. For instance, if there is more than one segment on a single line, each segment would be identified by the transcript line number (e.g. , 1) and a subcode (e.g., 1.1): 1 (1) The brown fox jumped over the fence. (1.1) He jumped high. If a single communication is comprised of a number of segments that require more than a single transcript line the coder should assign subcodes in the follow way. 1 (1) The brown fox jumped over the fence. (1.1) He jumped high 2 but landed very hard. (2) The farmer came and the fox ran 3 away. (3) The farmer jumped. (3.1) He got his gun and fired at 124 4 the fox. (4) I thought the farmer had killed the fox but I guess not. Any prominent disrupting child behaviors should be noted on the written transcript. These notations should be made when the child is agitated, distracted, crying, or fidgeting, and when the child's behavior elicits requests for attention from the officer (e.g., touching, jumping, moving from seat). 125 Pass 2: Global Key Element Ratings In the global key element rating process, the coder records whether the events listed below occurred or not. Although the events should follow the sequence indicated, coding whether or not they do is not important at this point in the coding system. Unless otherwise indicated in the coding instructions, coders will score the key elements that do occur with a yes (Y) response. If the criteria for the key element was not met (i.e., the event did not occur) the coders will score a no (N) response. Phase 1: Introduction Phase Officer states the time interview begins. Score a yes (Y) if the officer states time (e.g., "2:00PM", "1400 hours"). Disregard the time shown on the videotape; the officer must state the time to score a yes. Officer states the date of the interview. This may or may not include the name of day of the week but must include date, month, and year. Score a yes (Y) if the complete date is given and incomplete (INC) if a partial date is given (e.g., "August 1994"; "24th of August"). Disregard the date shown on the videotape; the officer must state the date to score a yes. Officer identifies him/her self as a police officer. To score a yes (Y), this must be a verbal statement. If the officer shows his/her badge, coders will circle (B) on the coding sheet. Officer identifies the child. Score a yes (Y) when the officer either says the child's full name (first and last) or has the child say his/her full name. If only part of the child's name is given by either party, record an incomplete (INC). Officer identifies any other people in the interview room. If person(s) other than the officer and child are present at the beginning of the interview, score a yes (Y) if the officer identifies the person(s) by name or by his/her relationship to the child (e.g., "Ms. Smith from the Ministry for Children and Families"; "your mom"). If people enter the room during the interview, identification may be done in a less formal manner (e. g., "Okay, here is mom"; "Ms. 126 Smith has come in to see you"), and will be scored as yes (Y). If the officer asks people to identify themselves and the people do so, score a yes (Y). If there is no one other than the officer and the child in the interview room throughout the interview, score not applicable (NA). Officer gives a brief description of his/her role. Score a yes (Y) response when the officer describes his/her role as a police officer. The description may be very general (e.g., "My job is to talk to kids."; "I wear plain clothes because .. . "). Note. If the officer and child have previously met, the officer may begin the interview by reminding the child of their previous meeting. In such situations, the officer may not handle the introduction phase of the interview the way he/she would if meeting the child for the first time (e.g., officer may say "Like I told you before, I'm a police officer and part of my job is to talk to kids"). However, the officer still needs to meet the criteria outlined above to receive yes (Y) scores for the elements listed above. Phase 2: Rapport Building Phase The child is asked to describe in detail an event unrelated to the abuse. To score a yes (Y), the officer must attempt to elicit a free narrative from the child about some life event unrelated to abuse (e.g., the child's last birthday party, a trip to the museum). Do not score a yes (Y) if there is no attempt to elicit a free narrative about an event. The officer discusses the difference between telling the truth and telling a lie. Score a yes (Y) if the officer engages the child in a discussion of the difference between truths and lies. The officer may ask the child to explain the difference or may ask the child to differentiate truths and lies by describing scenarios in which someone is telling truths and/or lies. Officer explains that it is important to only talk about things that really happened. Score a yes (Y) if the officer reinforces the importance of telling only the truth during the interview. The officer's statement may be rather general (e.g., "All we are going to talk about here today are things that really happened"). 127 Officer explains that it is okay to say "I don't know" if that is the truth. Score a yes (Y) if the officer somehow expresses to the child that if he/she cannot answer a question that it is fine to say so. For example the officer may say "If I ask you something and you don't know, it is okay, you can tell me that you don't know" or "If I ask you something and you don't remember, then you can tell me that you don ' t remember". Phase 3: Introducing the Topic Phase Officer uses general questions to introduce the reason for the interview. Score a yes (Y) if the officer asks one or more general questions to help the child identify the purpose of the interview (e.g., "Do you know why we are here today?"). Officer asks specific questions to introduce the reason for the interview. Score a yes (Y) if the officer asks one or more specific questions to help the child identify the purpose of the interview (e.g., "Has anyone done anything to you?"). Also score a yes (Y) if the officer states the purpose of the interview either directly (e.g., "Your mom told me there was some stuff going on that you wanted to talk to me about") or indirectly (e.g., "I just want to know, has anything happened to you that you would like to tell me about?"). Child makes a disclosure. Score a yes (Y) if the child discloses that he/she has had sexual contact with a peer or an adult, or has experienced some type of physical abuse. Indicate the type of abuse (adult sex, peer sex, or physical abuse) on the coding sheet. If the nature of sexual contact is unclear because the age of the perpetrator is never specified, record adult sexual abuse. The child's disclosure must clearly describe an abusive situation (see definition of disclosure given in Pass 3) to score yes (Y). Child recants. If, after making a disclosure, the child denies that previously reported abuse occurred, score yes (Y). Score no (N) if the child never retracts his/her disclosure or if he/she retracts then makes another disclosure. If the child never makes a disclosure, score not applicable (NA). 128 Phase 4: Free Narrative Phase Officer attempts to elicit a free narrative about abuse events. If the officer uses an openended type probe to elicit a narrative description about the abuse incident(s), score yes (Y), regardless of whether the officer' s attempt to elicit a free narrative is explicitly structured (e.g. "Tell me everything you can remember and start from the beginning") or implicitly structured (e.g., "Can you tell me what happened?"). If the officer uses an open-ended probe but immediately follows it with a close-ended probe leaving the child no opportunity to respond to the initial open-ended probe, score no (N). If the officer never uses an open-ended probe to elicit a narrative, score no (N). If the child does not make a disclosure, score not applicable (NA). Officer uses body diagram to discuss body/function terms. If the officer uses a pre-made or hand-drawn body diagram to discuss body part names and/or functions, score yes (Y), even when use of the drawing does not successfully elicit body part names and/or functions from the child. Officer uses the same body part or body function terms/gestures as the child. If the officer consistently uses the same (or very similar) verbal terms or body gestures as the child when referring to genitalia or breasts, score yes (Y). If the officer never uses the same verbal body terms or gestures as the child, score no (N). If the officer sometimes uses the same terms and gestures as the child but sometimes uses different words or gestures, or introduces new words or gestures, score inconsistent (INC). Examples If the child refers to his/her genitals as "wee-wee" and the officer consistently uses "wee-wee", score yes (Y). If the child points to his/her body and/or the body diagram to indicate a specific place and the officer reminds the child that he/she pointed there (e.g. "You pointed down there, remember?"), score yes (Y). 129 If the child says "privates" and officer says "private parts", score yes (Y). If the child refers to his "penis" and the officers uses "genitals", score no (N). If the child has only pointed and the officer introduces new language by saying "You pointed to your breast", score no (N). If the child refers to "boobies" and the officer sometimes uses the word "boobies" and sometimes uses the word "breasts", score as inconsistent (INC). Phase 5: Specific Questioning Phase The material given here is for information only; no codes will be assigned in Pass 2 for Phase 5 because all officers use specific questions with young children. In Phase 5, the officer is supposed to ask specific questions to clarify and extend information provided by the child during the free narrative phase. Questions that cover specific details about abuse-related information, such as who, what, where, when, and why, are typical Phase 5 questions. If the officer has begun specific questioning and is told of abuse occasion(s) not described previously, the officer is supposed to ask for a free narrative regarding that occasion. When this occurs, coders are to revise and/or extend their coding of items (a) through (e) for Phase 4, Free Narrative Phase, following coding rules given in that section. Specific sections of Phase 4 coding that will likely need to be revised include whether and in what order items (a) to (e) were covered, and whether the officer helps the child label each abuse incident. Phase 6: Closure Phase Officer asks the child if he/she has any questions. If the officer attempts to ensure that the child's questions and concerns are addressed, score yes (Y). Examples "Is there anything you would like to ask me before we leave here today?" "Is there anything else you would like to talk to me about today?" 130 Officer explains what will happen next. If the officer explains (generally or specifically) what will happen next, score yes (Y). Examples "Next we will talk to your mom and dad and then we may want to talk to you again, okay?" "Mom will take you home now and I will call to make another appointment with you." Officer provides child with appropriate contact names/numbers. If the officer explains how the child may contact the officer, score yes (Y). Examples "Call me if you remember anything else." "Tell your mom if you remember something else and she will know what to do." Officer thanks child for his/her participation. If the officer thanks or otherwise acknowledges his/her appreciation of the child's participation in the interview, score yes (Y). Examples "Thanks for coming to talk to me today." "I appreciate you telling me these things." 131 Pass 3: Turn by Turn Coding of Officer's Segments The goal of Pass 3 is to record specific elements of the interview on a turn-by-tum basis, which will permit analyses of frequencies of specific behaviors and of sequential patterns. Unless otherwise specified, the codes detailed below apply to the officer's segments only. Data File Preparation for Pass 3 Coding Preparation of an Excel spreadsheet file is necessary before Pass 3 coding can begin. In addition to listing all the variable names across the top row of the spreadsheet, coders will prepare the file by entering Line Number and Speaker in the first two columns of the spreadsheet. Line Number For all segments in the interview, the line numbering that was recorded on the transcripts in Pass 1 will be used to associate specific segments with Pass 3 codes. Before beginning Pass 3 coding of a given interview, coders will create an Excel spreadsheet file in which the line number for each segment of the interview is entered, in sequence, in the first column of the spreadsheet. Pass 3 codes associated with each segment will be entered in the spreadsheet in subsequent columns. Speaker For all segments in the interview, coders will record the identity of the speaker of each segment by entering one of the following values in the second column of the Excel spreadsheet. Speaker (1 digit code) 0 =Officer 1 =Child 3 =Other 132 Pass 3 Variables to be Coded Segment Referring to Abuse Rationale. This variable will be used to identify sections of the interview that specifically concern abuse. It will be coded for both officer and child segments. General definitions. Segments may refer to one of three types of abuse. The three types are (a) sexual contact with an adult, (b) sexual contact with a peer, and (c) physical abuse. Begin coding segments as referring to abuse when the child provides a disclosure statement which clearly describes an abusive situation or when the child replies affirmatively when specifically asked whether abuse occurred. Ambiguous segments such as "He had a bad idea" or "It was a bad thing she wanted to do" or "He touched me" are not sufficient to begin coding segments as referring to abuse. After disclosure, continue coding segments as referring to abuse unless the topic changes to something unrelated to the initial disclosure. If and when the discussion returns to topics related to the initial disclosure, resume coding segments as referring to abuse. Discussions about body parts and body functions will be assigned a score of 2 for this variable, unless the discussion is explicitly linked to the child's specific abuse experiences. Specific definitions. The values for this variable and examples for each value are given below. Assign scores to both officer and child segments. Note that the value of "1" is not used. O=No Assign a score of 0 whenever a non-zero score can not be assigned. In most cases where the score is 0, the segment is unrelated to abuse or the segment is an attempt by the officer to elicit an initial disclosure. Examples Officer "Do you know why we are here today?" "Thanks for coming to speak to me today." Child "I had a lot of friends at my birthday party." 133 "I live at ... " 2 =Body Assign a score of 2 if the segment is part of a discussion about body part names and/or body part functions, provided the discussion is not explicitly linked to the child's experience of abuse. Examples OF: "What is this part called?"/ CH: "Wee wee." 3 =Adult sex Assign a score of 3 when the child gives an explicit disclosure of sexual contact with an adult and continue scoring 3 as long as that sexual contact is the topic of discussion. Examples Child: "He kissed my privates." "She wanted to touch my privates." OF: "Did he touch your private parts?" I CH: "Yes." OF: "Show me on the picture where he touched you." I CH: "Here on the weewee." 4 =Peer sex Assign a score of 4 when the child gives an explicit disclosure of sexual contact with someone younger than 16 years of age. Continue scoring 4 as long as that sexual contact is the topic of discussion. Examples Child "My friend Billy touched my tits." 5 = Physical abuse Assign a score of 5 when the child gives an explicit disclosure of physical abuse. Continue scoring 5 as long as that physical abuse is the topic of discussion. 134 Examples "She hit me across the face with the belt." 33 =Inaudible or incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify specific situations. 1. If, after the child has made an initial disclosure, the officer begins asking the child about a second alleged abuser, segments related to that activity will be assigned a score of 0 until an explicit disclosure has been made. 2. Differentiation of types of sexual contact (i.e., adult versus peer) should be based on the coder's knowledge of the remainder of the interview, even though the initial disclosure may not contain information about the age of the alleged abuser. 135 Tum Type Rationale and general definitions. This variable will be used to record the type of segment spoken by the officer. Each officer segment will be assigned a score for this variable. Specific definitions. The values for this variable and examples for each value are given below. 1 =Plain statement A plain statement is a statement of fact or opinion and is not a request for information, a repetition of what the child said, nor an acknowledgement of what the child said. Examples "Time now is 1100 hours." "My name is ... " 2 =Tag question Like plain statements but with brief questions appended to them, tag questions are not requests for information but rather are attempts to elicit agreement and/or cooperation. The question ending is often "okay?", "right?", or "isn't it?". Examples "You have a seat in the chair, okay?" "It's cold out today, isn't it?" "Your name is ... , right?" 3 =Mixed probe This type of segment contains a probe (see definition below) along with some other form of communication (e.g., plain statement or acknowledgement). The probe portion of the segment will usually be at the end of the segment, but may occasionally come at the beginning of the segment. If a probe is of the If-then type, do not 136 code it as a mixed segment unless it contains something besides the If-then probe. Examples "Today is Monday, and the last time you saw him was when?" "Okay, tell me about the last time you were with him." "That's good, now what if I said I had green hair, what's that?" But not "If I said my hair was green, would that be a truth or a lie?" 4 =Simple probe Probes are intended to elicit information and can take the form of statements, demands, or questions. Examples "Tell me more about it." "Can you tell me more about that?" "Pardon me?" "If I said my hair was green, would that be a truth or a lie?" "You must tell me what happened." 5 =Repetition Score repetition when the officer repeats exactly what the child said, allowing for minor changes which do not alter the content of the statement, such as the use of appropriate pronouns. Repetitions are often punctuated with question marks in the transcripts but are not intended to elicit information. To be scored as a repetition, the officer' s segment must follow directly after the child's segment and must contain some content directly expressed by the child. Examples 137 Child: "I went to the park to play." I Officer: "You went to the park to play?" Child: "No." I Officer: "No, okay." Child: "It's a back." I Officer: "Yeah, it's a back." 6 =Acknowledgement Brief segments or non-meaningful paralinguistic utterances are to be coded as acknowledgments when that is the apparent intent of the segment. When the officer is responding to a probe from the child or seeking information from the child, do not code his/her segment as an acknowledgement. Examples "Okay." "Yeah, okay." "Is that right?" "Hm-hm." But not "Is that right?" (if context suggests officer is seeking information) Child: "Can I color?" I Officer: "Okay." 33 = Inaudible or Incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify difficult coding decisions. 1. When coders can not determine, from context and intonation, whether a segment is a repetition or a probe, code it as a repetition. 2. When coders can not determine, from context and intonation, whether a segment is an acknowledgement or a probe, code it as an acknowledgement. 138 3. When coders can not determine whether a segment is a repetition or an acknowledgement, code it as an acknowledgement. Probe Type Rationale. Probe Type is to be scored whenever Tum Type = 3 "mixed segment" or 4 "probe" , and will be used to classify questions and probes according to their linguistic structure. These codes draw on work by Peterson & Biggs (1995) and Walker (1993). General definitions. Direct probes have relatively simple syntax and their meaning is unambiguous. In contrast, indirect probes have more complex syntax and the meaning of indirect probes and their answers are often ambiguous. At a concrete level, indirect probes can be answered with either a "yes" or a "no" response, but at a less concrete level, they contain requests for further information. For example, in the indirect probe "Do you remember if he was there?", it is not clear whether the questioner is asking about the interviewee's memory or the presence of the person referred to as "he". Furthermore, a "Yes" response could mean two things ("yes, he was there" or "yes, I remember") and a "No" response could mean two things ("no, he wasn't there" or "no, I don't remember"). Phrasings such as "Do you remember X?", "Can you tell me X?" or "Do you know X?" ALMOST ALWAYS signify an implicit structure (where X is any proposition). Open-ended probes are very general requests for information intended to elicit a narrative-type response. The content of the expected answer is left open for the interviewee to interpret as he/she sees fit. Wh- probes are intended to elicit specific information, such as details about who, what, where, when, why, how, how many, or how much, and do not have potential answers embedded in the probe. Questions regarding the names and functions of body parts will usually be coded as Wh- questions. Multiple choice probes offer a choice of responses embedded in the question. Yes/No probes ask the interviewee to indicate agreement or disagreement with a proposition. If, then probes contain conditional statements, as in "if X, then 139 Y", where X is some condition and Y is the probe requiring an answer, and may be phrased in a variety of ways. Officers frequently use If-then structures when attempting to assess the child's knowledge of truths and lies. Requests for repetition are probes used to request the interviewee to repeat his/her immediately preceding response. Requests for clarification should not be coded as requests for repetition and will be codable in other categories (e.g., open-ended). Other probe types will be scored whenever a probe is not classifiable as one of the above types and includes commands issued by the officer to the child in a probe form. This is a "residual" category and should be used sparingly. Specific definitions. The values for this variable and examples for each value are given below. 0 = Not applicable. Assign a score of 0 when Tum Type is not scored as 3 or 4 (i.e., when the officer's tum is not a mixed segment or a probe). 1 =Direct open-ended Probes that have an explicit structure and are intended to elicit a narrative-type response. Examples "Tell me about that." "What do you mean?" "Describe what happened yesterday." "Your birthday was a long time?" (officer is asking for clarification of child's previous statement) 2 = Direct wh- Probes that have an explicit structure and are requests for specific details but do not contain potential answers. Examples "Who was there?" "What was he wearing?" 140 "Where did that happen?" "When was that?" or "What time was it?" "Why did you do that?" or "How come you did that?" "How many people were there?" "What's this part called?" "What do you use your legs for?" 3 = Direct multiple choice Probes that have an explicit structure and have a choice of answers embedded in the question. Examples "Was his hair brown or black?" "Were you wearing pants or a skirt?" "Were there four or six people there?" 4 =Direct yes/no Probes that have an explicit structure and ask for agreement or disagreement with a proposition. Examples "Was he there?" "Do you forget?" "Did you like that?" "Are you sure?" "No, no one?" 5 =Indirect open-ended Probes that have an implicit structure and are intended to elicit a narrative-type response. Examples "Can you tell me more about that?" "Can you explain what you mean?'' 141 "Can you tell me what happened yesterday?" 6 = Indirect wh- Probes that have an implicit structure and request specific details but do not contain potential answers. Examples "Do you remember who was there?" "Can you tell me what he was wearing?" "Do you know where that happened?" "Do you remember what time it was?" "Do you know why he did that?" "Can you tell me how many people were there?" "Do you have other names you call it?" 7 =Indirect multiple choice Probes that have an implicit structure and have a choice of answers embedded in the question. Examples "Do you know if his hair is brown or black?" "Do you remember if there were three or six people there?" "Can you tell me if you were wearing pants or a skirt?" 8 = Indirect yes/no Probes that have an implicit structure and ask for agreement or disagreement with a proposition. Examples "Do you know if he was there?" "Can you tell me whether you forget?" "Do you remember if you liked that?" "Do you think you can help me?" 142 9 = "If-then" Probes that contain conditional statements, as in "if X, then Y", where X is some condition and Y is the probe requiring an answer Examples "If I said your hair was purple, then would I be telling the truth or telling a lie?" "If I said your hair was purple, what's that?" "What about when I take a drink from the cup, what do I touch it with?" 10 =Request for repetition Probes used to request the interviewee to repeat his/her immediately preceding response. Examples "Pardon me?" or "Excuse me?" 11 =Other Any questions not classifiable as one of the above types. Includes implied commands. Examples "Do you want to have a seat here?" "Will you come in?" 33 = Inaudible or incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify difficult coding decisions. 1. If coders can not decide whether a probe has an indirect implicit structure or is a direct explicit Yes/No probe, it should be coded as a direct Yes/No question. 2. When a probe has the If, then structure, coders must decide whether the condition X is essential to understand the probe properly or whether the condition X is being used for purposes other than listener comprehension (e.g., to pressure the interviewee). If the 143 condition X is essential to proper understanding, then code the question as an If, then type, even though it may also fit into one of the other probe types (e.g., multiple choice). If the condition X is NOT essential to proper understanding, then code the question as some other category (as appropriate). Examples "If you don 't remember, then why did you tell your mom that you did?" "If I said to you, can you tell me what's telling a lie, what would you say?" The conditions are not necessary for understanding these questions, so code as direct wh- questions. 3. DO NOT code a segment as an If, then probe if you can not decide whether a condition is essential for understanding the question, . Use another appropriate category. Examples "If it's the truth, do we get in trouble for telling the truth?" "If he came in here right now, would you recognize him?" It is not clear whether the condition is essential for understanding the question, so code as a direct yes/no question. 4. On occasion, multiple choice questions are phrased as a series of separate segments. In such situations, use a decimal to indicate that the question has multiple segments. The following sequence illustrates this coding rule. OF: direct wh- "What day is it?" CH: "I don't know." OF: "Is it Monday?" first part of multi-segment MC question, code as 3.1 OF: "Tuesday?" second part, code as 3.2 OF: "Wednesday?" third part, code as 3.3 OF: "Thursday?" fourth part, coded as 3.4 144 In this example, if the child had replied after the officer's question "Is it Monday?", that question would have been coded as an direct yes/no, and "Tuesday?" would have been coded as 3.1, "Wednesday?" as 3.2, and "Thursday?" as 3.3. 5. When coding indirect questions, coders will usually have to decide what the implied question is before they can identify the type of implied question. One method to help decide this is to ask yourself what information you would provide if you were asked the question. Examples: "Can you tell me about your last birthday party?" would be interpreted by most adults as meaning "provide an account of my last birthday party", and it is an indirect open-ended probe. "Do you remember what time it was?" would be interpreted by most adults as meaning "what time was it?", and it is an indirect wh- probe. 6. When a series of probes contains some grammatically incomplete probes which are incomprehensible by themselves but make sense in context, code the grammatically incomplete probes the same way the probe occurring immediately before the grammatically incomplete probe was coded. Apply this rule only in those situations where the officer's turns are sequential with no reply from the child occurring between the parts of the officer's conversational tum. Example "Can you tell me anything else about him? About what he looked like?" code both segments as indirect wh- probes because that's what the first probe is and the second probe is grammatically incomplete. 145 Leading Segment Rationale. This variable will be used to identity officer segments that are suggestive in nature and are, therefore, likely to contaminate the child's statement about abuse-related events. To code a segment as leading or nonleading, the segment must occur in interview phases specifically concerned with abuse, that is, it must be occur in Phase 3, 4.1, or 5 (see definitions of Phase, below). General definitions. In Phases 3, 4.1, and 5, segments are defined as leading when they (a) contain the answer or a choice of answers, (b) name the suspected offender before the child has done so, (c) contain explicit details of the alleged offense to which the child has not previously referred in the present interview, or (d) contain the interviewer's assumptions about the alleged abuse events (e.g., an assumption that abuse did occur). The context of the ongoing discussion must be considered when deciding whether a segment is leading. If the child has already referred to a fact or opinion and the officer later repeats that fact or opinion when asking another question, the repetition is not to be considered leading. Similarly, if the officer asks about some detail or concept that is difficult for a young child to understand and simplifies the child's task by askng a multiple choice probe, that probe would not necessarily be considered leading, particularly when the choices provided in the probe exhaust all possible answers. Finally, segments which reflect an assumption that the child told someone about the alleged abuse are not to coded as leading, because that assumption is often warranted (i.e., the child's prior report to a trusted person prompted the interview). Specific definitions. The values for this variable and examples for each value are given below. 0 =Not leading When Phase= 3, 4.1, or 5, a score ofO means the segment is not or not applicable 146 leading. When Phase is not equal to 3, 4, or 5, a score of 0 means not applicable. Examples (when Phase= 3, 4, or 5) "No one is going to be in trouble." "Did you tell your mommy anything else?" "Did it happen before or after lunch?" (not considered leading because options simplify a difficult concept) "Were you wearing pants or a skirt?" (not considered leading if child previously said she was wearing clothes; options are exhaustive) 1 =Leading Assign a score of 1 when a segment is suggestive as defined above. Examples (taken from the Step-Wise Interview training manual; some of these may not be considered leading given the appropriate context) "Did he tell you not to tell anyone?" "Were you scared, angry or sad?" "Was it your Dad who touched you?" "We've been told you are having a problem with your Uncle." 33 = Inaudible or incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify difficult coding decisions. 1. On occasion, officers attempt to encourage children to make disclosures by making statements about the anticipated consequences of disclosure. These statements may or may not be leading. Base coding of such segments on the degree to which the officer's 147 assumptions are directly expressed in the segment. Coders may also decide to code such segments as Misleading or Coercive (see definitions below). Example (as part of discussion about whether someone touched the child) "No one is going to get in trouble." Code as Leading= 1; Misleading= 3. 148 Misleading Segment Rationale. This variable will be used to obtain frequencies of segments in which the officer misrepresents reality when speaking to the child. General definitions. Misleading segments are segments in which the officer either misrepresents his/her intentions or makes a promise that he/she may not be able to keep. Three categories of misleading segments will be coded. Specific definitions. The values for this variable and examples for each value are given below. 0 = Not misleading Assign a score of 0 to any segment which does not fit one of the categories listed below. 1 = Intentions Assign a score of 1 when the officer misrepresents the intentions of or for the interview. Example "Remember how we were chatting outside about playing in your yard, that's all I wanted to have a chat about." 2 = Promises Assign a score of 2 when the officer makes a promise that he/she may not be able to keep. Such promises may relate to the officer's ability to protect the child, the child's family, or the alleged perpetrator. Examples "No one will hurt you like that again." 3 =Other Assign a score of 3 when the officer makes a misleading statement that is neither a misrepresentation of the interview process nor a promise. Examples 149 "Mommy won't give you any trouble." "No one will be in trouble." 33 = Inaudible or incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify difficult coding decisions. 1. When coders are uncertain whether a given segment is misleading, code it as misleading (i.e., code liberally rather than stringently). 150 Interview Phase Rationale. This variable identifies phases of the Step-Wise Interview protocol. It will be used to determine (a) whether the officer covered key elements of the Step-Wise Interview protocol, (b) the order in which the key elements were covered, and (c) the proportion of the interview spent in each phase of the interview. General definitions. Coders are to record which of the six phases of the Step-Wise Interview protocol (if any) a given segment is consistent with. The goals of Phase 1 are to orient the child to the interview situation and to provide procedural information important to the court. The goals of Phase 2 are (a) to obtain a free narrative from the child which, in addition to making the child feel comfortable, is to be used to assess the child's developmental level, and (b) to model how questioning will proceed later in the interview. The goal of Phase 2.1 is to ensure that the child knows what is expected from him/her during the interview. The goals of Phase 3 are to draw the child's attention to the reason for the interview and to obtain an initial disclosure of abuse from the child. The goal of Phase 4.1 is to obtain a free narrative of abuserelated events from the child. The goal of Phase 4.2 is to elicit the child's labels for body parts. The child's knowledge of body part functions may also be assessed. The goal of Phase 5 is to clarify and/or extend information provided by the child in Phases 3 and 4.1. The goals of Phase 6 are to thank the child for participating in the interview and to ensure that the child's questions and concerns are addressed. For all interview phases, application of the codes requires judgments about the nature of the activity the officer is engaged in. When an officer is clearly pursuing an activity consistent with the goals of a given phase, assign a non-zero code even though the particular segment being coded may not be specified in the protocol. For example, when the officer is asking specific questions about abuse events, he/she may use paralinguistic or verbal acknowledgement cues (e.g., "uh-huh" or "okay") to encourage the child to continue his/her answer to a given question 151 or may restate what the child says to ensure he/she understands it. Because such segments are consistent with the purpose of Phase 5, they would be coded as Phase= 5. If, however, after asking some specific questions about the abuse, the officer asks the child to explain the difference between truths and lies, segments associated with that activity would be coded as Phase= 2.1, provided the segments were not directly linked to the child's experience of abuse (in which case the segments would be coded as Phase = 5). Specific definitions. The values for this variable and examples for each value are given below. 0 =No phase Assign a score of 0 when the officer is engaged in an activity not explicitly specified in the Stepwise Interview protocol. Examples When officer temporarily leaves room mid-way through interview, all officer segments related to that activity are Phase= 0. When, early in the interview and before introducing the topic of abuse, the officer and child are talking about some nonabuse material that is not specified as part of Phase 1 or Phase 2, all segments related to that activity are Phase= 0. 1 =Introduction phase Assign a score of 1 when the officer's segment is related to one of (a) date of interview and/or time the interview begins, (b) identification of self as a police officer and/or descriptions of his/her role, (c) identification of the child and other people present for the interview, or (d) information-gathering segments about the child's birthdate, age, address, name of school, and/or grade. 152 See decision rules for specific instructions. 2 = Rapport phase, Assign a score of 2 when segments are related to the child free narrative providing a free narrative about a non-abuse event. See decision rules for specific instructions. 2.1 = Rapport phase, Assign a score of 2.1 when the officer's segment is related to one of setting parameters (a) determining whether the child knows the difference between truths and lies, (b) ensuring the child knows he/she is to tell the truth (may be phrased as "talking about things that really happened"), or (c) telling the child that it is acceptable to say "I don't know" or "I don't understand" or "I don't remember" if that is the truth. 3 = Introduce topic Assign a score of 3 when the officer's segment is related to one of (a) helping the child identify the reason for the interview, or (b) obtaining an initial disclosure of abuse from the child. These activities may be very general or specific. If a disclosure is not forthcoming from the child, this phase may be quite long. Examples "Do you know why you are here today?" "Has anyone done anything to you?" "Who has seen or touched your genitals and/or whose genitals have you seen or touched?" "Are you afraid someone will get in trouble if you tell?" 4.1 =Free narrative phase, Assign a score of 4.1 when the officer's segment is related to 153 free narrative getting the child to provide a free narrative about abuse events. The officer may ask the child to tell him/her about the most recent time, the first time, unusual times, and times remembered well. Probes must be open-ended to count as Phase = 4.1. Examples "Tell me all you can remember." "Tell me about the first time it happened." "Tell me what usually happened." "Tell me about the time you remember most." 4.2 =Free narrative phase, Assign a score of 4.2 when the officer's segment is related to one of body diagram (a) finding out child's names for body parts, or (b) assessing child's knowledge of body part functions. If the body diagram is used to elicit information about abuse-related experiences, code the activity as Phase= 4.1 (if questions are of narrative type) or Phase= 5 (if questions are of specific type). 5 = Specific questioning Assign a score of 5 when the officer's segment is related to phase clarifying or extending information provided by the child in Phases 3 and 4.1 . See decision rules for specific instructions. Examples "Whose house was it?" "What color was the car?'' "Who knows about this?" "Was he touching you on top of or underneath your clothes?" 6 = Closure phase Assign a score of 6 when the segment is related to one of 154 (a) asking the child if he/she has any questions or concerns, (b) giving the child information about what will happen next, (c) providing the child with contact names/numbers, or (d) thanking the child for participating in the interview. 33 = Inaudible or incomplete Assign a score of 33 when a segment is inaudible or incomplete. Decision rules. The following decision rules should be used to clarify difficult coding decisions. 1. When Phase 1 elements other than statements of date and time of interview come up late in the interview, code them as Phase = 1 unless they are embedded in a segment which also contains a key element of another phase, in which case the segment should be coded as being from that other phase. 2. When the officer states the time after the interview is well under way, code that statement as Phase = 0 unless it is embedded in a segment which also contains a key element of another phase, in which case the segment should be coded as being from that other phase. 3. Segments related to whether the child likes school and information-gathering about who the child lives with are not coded as Phase 1 because they are not specified in the Stepwise Interview protocol. 4. In cases where the officer and child have met prior to the current interview, segments related to refreshing the child' s memory of the previous meeting should be coded as Phase 1 only when they serve the purposes of Phase 1. That is, to be coded as Phase 1, such segments must provide information important in a court setting or help the child (a) identify the officer as a member of the RCMP, (b) identify other people present for the interview or at the previous meeting, or (c) identify the officer's role. 5. Start coding Phase= 2 when the officer asks the child to provide a free narrative about a non-abuse event such as the child's last birthday party. Note that because one of the 155 goals of the phase is to obtain a free narrative, the topic of discussion must be an event, not the child's likes, dislikes, or attitudes. Continue coding Phase= 2 as long as any specific non-abuse event remains the main topic of conversation. If the officer engages the child in a discussion about non-abuse events but never asks for a free narrative, do not code the officer's segments as Phase= 2. Such discussions will usually be coded as Phase= 0. 6. If the officer does not attempt to elicit a free narrative and immediately after disclosure begins asking specific questions about abuse-related material, code specific questions about the abuse as Phase= 5. This will record that no free narrative about abuse events was elicited from the child during the interview. 7. Technically, the Step-Wise Interview protocol specifies that a free narrative about abuse events be provided by the child, without interruption from the officer, and only then should the officer ask specific questions to clarify information. When the child provides new information (i.e., describes a previously undescribed abuse event) in response to specific questions, the officer is supposed to elicit a narrative about the new information, then continue asking specific questions. Consequently, it is possible that Phase 4.1 and Phase 5 will alternate during the child's statement. 8. If the officer begins seeking information about abuse by asking specific questions, without requesting a free narrative, code Phase= 5, even though the child may not yet have made an explicit disclosure. 156 Checking for Illegal Code Combinations The final step in coding of Pass 3 is to check for illegal code combinations. Illegal code combinations reflect errors in coding which should not be allowed to affect interobserver agreement. The steps for checking for illegal code combinations are outlined below. 1. Transfer data file from Excel to SPSS by copying and pasting scores into the SPSS file called "Kappa Master", which has pre-defined variable names and specifications. 2. In SPSS, select DATA-> SELECT CASES. 3. Select if (turn_s_x = 0) and ((turnty_x = 3) or (turnty_x = 4)) and (answer= 0). Then request frequencies for line number, to get a list of segments where a probe was. not scored for answerability. 4. Select if (turn_s_x = 0) and ((turnty_x = 3) or (turnty_x = 4)) and (probety_x = 0). Then request frequencies for line number, to get a list of segments where a probe was not scored for probe type. 5. Correct any coding errors identified in this process. Corrections will be made to either or both of the SPSS file and the Excel file, as required. 6. The SPSS final data file is to be stored in "f:\rcmpstud data" and on floppy disk. The file (if needed) for Kappa calculations is to be stored in "f:\rcmpstud kappa" and on floppy disk. Appendix F 157 Pass 3 Variable Coding Flow Chart TURN SEQUENCE Officer or Child Segment SEGMENT REFERENCE TYPE Not About Abuse About Body Parts/ Function About Adult Sex Abuse About Peer Sex Abuse About Physical Abuse or Inaudible or Incomplete CODE ONLY OFACER SEGMENlS TURN TYPES MISLEADING Intention of the ~ ~ ~ ~ ~ i l ~~ ... .. Repetition Acknowledgment __ , PROBE TYPE Open-ended or Wh_, Yes/No, Multiple-Choice, If/Then Form either Direct or Indirect and Request for Repetition Residual Probes i~ ~ ~s ~~ TURN TYPES LEADING (steps 3, 4, 5) STOP CODING CHILD SEGMENlS STEPS OF INTERVIEW 1. Introduction 2. Rapport Building 3. Introducing Topic of Abuse 4. Free Narrative Abuse 5. Specific Questions 6. Closure TOPICS Body Parts/Functions Truth/Lie Segments Unrelated to Protocol Steps and Topics Appendix G 158 Pass 3 Officer Segment Turn Type and Probe Type Variables and Collapsed Categories Flow Chart Turn Type as coded Plain Statement Tag Question Mixed Probe Simple Probe Repetition Acmowledgment Mixed and Simple Probes =Probes ~ Plain Statement, Tag Question, Repetition and Acmowfedgment =Not Probes Probe Type first reduction as analyzed Probe Type as coded Direct Open-ended Indirect Open-ended DirectWh Indirect Wh Direct Ye&'No Indirect Ye&'No Direct Multiple-Choice Indirect Multiple-Choice Conditional lffThen Request for Repetition Residual Probes Turn Type as analyzed r-. Direct & Indirect Open-ended =Open-ended Direct & Indirect Wh =Wh_ Direct & lndi rect Yes'No =Yes/No Direct & Indirect Multiple-Choice =Multiple-Choice Conditional If/Then =ConditionaiiVThen Request for Repetition & Residual Probes =Other Probe Type second reduction as analyzed Open-ended =General Question r-.. Wh Ye&'No Multiple-Choice Conditional lffThen =Specific Question Other 159 Appendix H Mean Percentages and Standard Deviations for the Time Spent in Each of the Interview Steps. on Body Topics and Topics Unrelated to the Protocol in the No Disclosure and Disclosure Interviews with Young Children No Disclosure Disclosure (.!1 =5) (.!1 =7) Steps and T epics M M Step 1: Introduction 9.6% 5.2% 6.8% 3.6% Step 2: Rapport Building 9.9% 5.7% 9.4% 4.7% Step 3: Introducing the Topic of Abuse 21.6% 6.1% 9.4% 13.8% Step 4: Abuse-Related Free Narrative N/A N/A 4.9% 8.0% Step 5: Specific Questions 6.5% 8.7% 26.1% 23.2% Step 6: Closure 3.5% 1.8% 1.3% 1.2% Topic of Body Parts or Functions 10.5% 10.6% 10.3% 9.3% Topics Unrelated to the Protocol 38.4% 9.2% 31.7% 19.0% Appendix I 160 Standard Deviations for Officer Turn Types for the First and Second Halves of the interviews with Young Children by Interview Status Officer Turn Types Acknowle- Interview Half Plain Tag Probe Repetition dgement so so so so so No Disclosure Interviews (D = 5) First Half 10.6% 3.8% 4.8% 5.8% 3.1% Second Half 11 .7% 4.2% 8.5% 6.4% 8.1% Disclosure Interviews (D = 7) First Half 8.5% 4.6% 6.8% 5.2% 9.7% Second Half 13.8% 3.2% 8.4% 7.6% 14.5% Appendix J 161 Standard Deviations for Officer General. Specific and Other Question Types For First and Second Halves of Interviews with Young Children by Interview Status Officer Question Types Interview Half General Questions Specific Questions Othe(i so so so No Disclosure Interviews (D. = 5) First Half 2.6% 4.8% 3.8% Second Half 7.5% 11.0% 3.9% Disclosure Interviews (D. = 7) First Half 6.9% 6.5% 3.0% Second Half 8.5% 8.0% 3.1% a Includes probes what were recorded as requests for repetition and unclassified (residual) questions. 162 Appendix K Standard Deviations for Officer Specific Question Types For First and Second Halves of Interview with Children Under 8 Years-of-Age by Interview Status Officer Specific Question Types MutipleInterview Half Wh_ Yes/No Choice If/Then so so so so No Disclosure Interviews (n = 5) First Half 5.5% 6.0% 6.0% 1.1% Second Half 12.4% 12.81% 0.0% 1.7% Disclosure Interviews (n = 7) First Half 10.2% 11.9 2.9% 2.4% Second Half 15.0% 12.0% 2.4% 7.0%