LEGAL OUTCOMES AS EFFECTIVENESS RESEARCH OF FORENSIC INTERVIEWING AT THE CHILD ADVOCACY CENTER OF EAST ALABAMA (CACEA) Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. Angela B. Bradford Certificate of Approval: Alexander T. Vazsonyi Thomas A. Smith, Jr., Chair Professor Associate Professor Human Development and Human Development and Family Studies Family Studies Scott A. Ketring Stephen L. McFarland Associate Professor Acting Dean Human Development and Graduate School Family Studies LEGAL OUTCOMES AS EFFECTIVENESS RESEARCH OF FORENSIC INTERVIEWING AT THE CHILD ADVOCACY CENTER OF EAST ALABAMA (CACEA) Angela B. Bradford A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 16, 2005 iii LEGAL OUTCOMES AS EFFECTIVENESS RESEARCH OF FORENSIC INTERVIEWING AT THE CHILD ADVOCACY CENTER OF EAST ALABAMA (CACEA) Angela B. Bradford Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. Signature of Author Date of Graduation iv VITA Angela Beesley Bradford was born on February 9, 1979 to R. Mark Beesley and Doris J. Beesley in Provo, UT. She was raised in San Diego, CA and married Brian David Bradford on May 10, 2002. In 2003, she graduated from Brigham Young University with a Bachelor of Science degree in Family Science. She completed a Master of Science degree in Human Development and Family Studies, specializing in Marriage and Family Therapy at Auburn University in December of 2005. v THESIS ABSTRACT LEGAL OUTCOMES AS EFFECTIVENESS RESEARCH OF FORENSIC INTERVIEWING AT THE CHILD ADVOCACY CENTER OF EAST ALABAMA (CACEA) Angela B. Bradford Master of Science, December 16, 2005 (B.S., Brigham Young University, 2003) 48 Typed Pages Directed by Thomas A. Smith, Jr. Sexual abuse is a devastating reality with tragic effects on the victims. It affects approximately one in four children before the age of eighteen. Legislation and prosecution of sex abuse crimes are intended to deter perpetration of sexual abuse; however, obtaining the needed information from children in order to successfully pursue criminal charges is difficult. Child Advocacy Centers (CACs) have been established to aid in the forensic interviewing process. Forensic interviewing is intended to gain accurate accounts of alleged abuse from children to aid in the pursuit of criminal charges. This study tested the effectiveness of forensic interviewing at the CAC in East Alabama (CACEA) in terms of legal outcome. Seven hundred seventeen Department of Human Resources (DHR) indicated child sexual abuse cases were divided into two groups: 1) CACEA, where trained forensic interviewers were employed and 2) DHR, where traditional social workers or law vi enforcement officers were the interviewers. These two groups were compared in terms of criminal legal action outcomes. Results revealed that criminal indictments occurred twice as often with CACEA forensic interviewed cases than with DHR, traditionally interviewed cases. Additionally, convictions among charged cases occurred three times more often with CACEA interviewed cases than DHR interviewed cases. Other statistically non-significant trends were found, suggesting the need for future studies to control for confounding factors. While the results of this evaluation suggest superior results for forensic interviewing, large-scale, comparative studies can aid in further validating forensic interviewing, contributing to the goal of more effectively holding perpetrators accountable for their crimes against children. vii ACKNOWLEDGMENTS I would like to thank my God for all the challenges and blessings He has given and continues to give me for my benefit. Without Him, this thesis and my other successes would not have been possible. He has also given me my greatest blessing, my husband, Brian. I would like to thank Brian, who has always encouraged and supported me in my endeavors. Brian, your continued love, confidence in me, and humor are priceless. You will always be the top of my totem pole. I would like to thank my parents for teaching me the value of education and hard work. Finally, I would like to thank my thesis chair and committee members who have provided me with an invaluable education and experience. Tommy, Scott, and Alex, you have helped me complete this thesis, grow professionally, and most importantly, you have helped mold me into who I am today. Thank you. viii Style manual or journal used: Publication Manual of the American Psychological Association, Fifth Edition Computer software used: Microsoft Word ix TABLE OF CONTENTS LIST OF TABLES...............................................................................................................x INTRODUCTION ...............................................................................................................1 REVIEW OF LITERATURE ..............................................................................................7 METHODS ........................................................................................................................13 Overview................................................................................................................13 Participants.............................................................................................................13 Agency Procedures ................................................................................................14 Study Procedures ...................................................................................................18 Variables ................................................................................................................21 RESULTS ..........................................................................................................................24 Data Preparation.....................................................................................................24 Analyses.................................................................................................................24 Hypothesis Testing.................................................................................................25 DISCUSSION....................................................................................................................29 Limitations and Future Directions .........................................................................33 Conclusions............................................................................................................34 REFERENCES ..................................................................................................................36 x LIST OF TABLES 1. Crosstabulation of Agency and Criminal Charge .........................................................26 2. Crosstabulation of Agency and Legal Outcome ...........................................................28 1 INTRODUCTION Child sexual abuse is one of the most heinous crimes in our society. It is estimated that one in four children will be sexually abused before their eighteenth birthdays (Kellog, 2002; Spungen, Jensen, Finkelstein, & Satinsky, 1989; Woody, 2002). This tragic reality is exacerbated by the many negative effects of sexual abuse. Although some victims of childhood sexual abuse remain asymptomatic, most suffer a variety of effects for extended periods of time. Studies show that abused children exhibit more symptoms than their non-abused counterparts (Saywitz, Mannarino, Berliner, & Cohen, 2000). Common symptoms include aggression, anxiety, depression and withdrawn behavior (Kendall-Tackett, Williams, & Finkelhor, 1993). Moreover, adults who experienced childhood sexual abuse are at greater risk for developing a variety of psychiatric disorders (Saywitz et al., 2000). These are devastating facts regarding the welfare of our children and their futures as adults. Surprisingly, despite the prevalence and abounding negative effects of childhood sexual abuse, relatively few measures are taken to prevent or deter perpetration. Most preventive efforts are aimed at educating children (Shapiro, 1993) or teaching them assertiveness (Spungen et al., 1989). The idea is that if children can express their discomfort, say no, and confide in a trusted figure, victimization will be minimized. Children who participate in such programs have been found to be approximately 50% less likely to be abused than those who do not participate in such programs. It is assumed 2 that this is due to the decreased likelihood of perpetrators? approaching well-prepared children (Gibson and Leitenberg, 2000). All things considered, however, these programs cannot protect those children who have already suffered abuse. Furthermore, the programs are not structured to adequately equip young children with the skills needed to deter the most dangerous and manipulative offenders. The reality is that child sexual abuse is a continuing problem and an ever- increasing occurrence. Some states have reported as much as a 200% increase in abuse cases over an eight-year period (Martin, 1992). With the growing rate of child sexual abuse, the need to take effective action is more important than ever. Traditionally, legislation and prosecution of crimes have been the modes whereby our society has attempted to prevent criminal behavior in the home and elsewhere (Stewart, 1999). It has been asserted that criminal prosecution can prevent further criminal action through the assignment of consequences (Lindsey & Regehr, 1993). In order to decrease perpetration of sex abuse crimes, the criminal justice system has increased its attempts to prosecute offenders and has instated harsher punishments for those found guilty (Peters, Dinsmore, & Toth, 1989; The Harvard Law Review Association, 1985). This is exemplified by the ?three-strikes-you?re-out? laws and other penalties specifically intended to deter repeated sex abuse offenses (Gilstrap, 2003). Unfortunately, there is dissonance between the legal system?s efforts to deter sexual abuse offenses and reality. One study found that only 35% of cases presented to the prosecutor ended in a plea or conviction (Bradshaw & Marks, 1990). Another found that 94% of cases carried forward by the District Attorney ended in a guilty plea or conviction (Cross, Walsh, Simone, and Jones, 2003), suggesting that the greatest 3 predictor of whether sexual abuse cases in a prosecution sample lead to punishment of the alleged perpetrator is whether or not the cases are actually charged. The difference in these studies can be explained by how many cases are referred to the District Attorney and how many of those are actually charged. In a study of cases referred to the prosecutor, 65% of total interviews resulted in a disclosure of criminal sexual abuse, while only 56% of the total number were referred to the prosecutor (75.5% of which included a disclosure) (Stroud, Martens, & Barker, 2000). The same was found in a 2003 21-study meta-analysis of criminal justice decisions in sexual abuse cases. Those authors reported that the mean rate of founded child sexual abuse cases referred to the District Attorney was 56% with a variability of 40% to 80%. They also found that the District Attorneys? mean rate of charging for a sexual abuse crime was approximately 50%, with most studies being variable between 48% and 76% (Cross et al., 2003). In another study analyzing 6,096 cases, Finkelhor (1983) found that criminal charges were filed in only 24% of child sexual abuse cases nationwide. Taken together, these statistics indicate that 24% to 28% of alleged perpetrators are actually charged, with approximately 22.5% to 26% of total cases ending in a guilty plea or conviction. It is reported that these statistics are largely due to the fact that successful prosecution of child sexual abuse is difficult to attain (The Harvard Law Review Association, 1985). Reliance on young children to disclose abuse restricts the extent to which we can know of its prevalence. Children may not know in whom to confide, fear punishment, be intimidated by the perpetrator, not fully comprehend the events, not recognize the events as wrong, and/or be poor reporters of fact and time. Sexual abuse is such an unreported crime that it is assumed to be the most under-reported type of child 4 abuse (Hodson & Skeen, 1987). Even when children have disclosed their sexual abuse and subsequent reports have been made, convictions have remained astoundingly low due to the very nature of interviewing young children (The Harvard Law Review, 1985). Several problems have been noted, including child suggestibility, synthesis of several abuse experiences, recollection inconsistencies, and misunderstanding between child and interviewer (Kuehnle, 1996). Nevertheless, children?s own accounts or statements are often the only evidence available. Furthermore, there has been some evidence that children?s statements are essential to the process. In looking at prosecution outcomes in cases, De Jong and Rose (1991) found no significant difference in the rate of felony convictions between cases with and without physical evidence of child sexual abuse. This underscores the importance of the child?s ability to give an accurate account of his/her abuse. Forensic interviews can be used to obtain such statements. To make forensic interviewing more efficient and safe against bias, Child Advocacy Centers (CACs) have been established. The first CAC was formed in Huntsville, AL when Madison County District Attorney Bud Cramer heard the story of a thirteen-year old who had to retell her account of being sexually abused fifteen separate times. Horrified, he recognized the need for an agency designed to protect children from re-victimization by the criminal justice system through repeated interviewing. This agency could coordinate the needs of various agencies involved into one set of interviews (Blodgett, 1987). The CAC strives to provide a safe, neutral environment, in which a child suspected of being abused can comfortably be interviewed by a trained forensic interviewer. An effective interview aids the criminal justice and child protective systems 5 in their respective efforts to gain information needed to make the best decisions regarding the child?s welfare and prosecution of the offender (Carnes, Wilson, & Nelson-Gardell, 1999; Kellog, 2002). CAC interviewers are trained to gain information from children without being suggestive or leading. Also, when deemed necessary, the interviewing model at the CAC allows children more than one interview in which they may disclose their experiences (Carnes, Nelson-Gardell, Wilson, & Orgassa, 2001). This evaluation is intended to provide familiarity and comfort for the child and can minimize the dangers associated with synthesizing several events into one account. CAC services help reliably determine whether a child has in fact suffered sexual abuse or what the nature of the abuse was, which aids in the process of prosecuting the alleged perpetrator (Carnes et al., 1999). Prior studies have shown that 71% to 73% of the cases in which credible disclosures were obtained in extended evaluations at a CAC led to successful legal action (Carnes et al., 1999; Carnes et al., 2001). This is a promising statistic; however, it is not known whether the results of these studies are meaningful to CAC interview effectiveness because a comparison group was not utilized. Furthermore, most children do not continue on from the interview to the evaluation process. Forensic interviews are a much more common occurrence (J. Payne, personal communication, April 18, 2005). Still, whether forensic interviewing is effective in obtaining the information necessary to successfully prosecute alleged perpetrators has not yet been established. An extensive review of the literature and direct inquiry to the NCAC yielded no studies and certainly no comparative studies focused on whether or not forensic interviewing as outlined by the American Professional Society on the Abuse of Children (APSAC) and 6 used by the National Child Advocacy Center (NCAC) is effective in terms of legal outcome (J. Payne, personal communication, February 9, 2005). As such, the purpose of this study is to explore the effectiveness of the interviewing model used by CACs. The proposed research will be conducted examining legal outcomes in Lee County, Alabama before and after forensic interviewing was established at the Child Advocacy Center of East Alabama (CACEA). 7 REVIEW OF LITERATURE In the mid-1980s, appellate court judges began questioning interviewing techniques and children?s suggestibility (Myers, Diedrich, Lee, Fincher, & Stern, 1999). Children?s statements were often no longer being accepted as credible. What was once enough to obtain a conviction was perceived as insufficient. There is a large body of literature supporting the notion of children?s high suggestibility (e.g., Martin, 1992; Roberts & Powell, 2001). Questions aimed at gaining specific information tend to be close-ended, potentially giving children ideas about experiencing fictitious events (Wood & Garven, 2000). If an interviewer asks a child ?Did Uncle Bob touch your private parts with his mouth?? the child may imagine the event described and gain a false memory. Even without gaining a false memory, there is an increased risk of the child concocting elaborate, fictitious stories and making false allegations based on what was asked or said (Bradford, 1994; Sauer & Okerblom, 1993). Should the prosecution rely on these mistaken accounts to make its case, the child?s credibility can be destroyed as the defense draws attention to inconsistencies in the child?s story. The defense can also challenge the interview process, stating that the questions were suggestive or leading. Thus, doubt enters the jury box, and a conviction becomes more difficult to attain. Children?s statements are also difficult to use successfully in court due to the complexity of integrating several events into one account (Prior, Glaser, & Lynch, 1997). For example, if a police officer interviews a child suspected of having been abused, that 8 child is implicitly expected to give information from several abusive experiences in one interview. When children experience an event (such as sexual abuse) several times, they establish a ?general event representation,? including the common characteristics across events into one cognitive script (Roberts & Powell, 2001, p. 1646). These are used to trigger recall of past events, but they also generalize the various events, making specifics of individual incidents difficult to remember (Roberts & Powell, 2001). This can result in contradictions within the story because time, place, or specifics change across events. If more than one interview is conducted, the difficulty arises when the child is required to summarize everything into one testimony on the courtroom stand. Again, the defense can find inconsistencies in the child?s account, and a conviction by jury becomes less probable. All of this is compounded by the fact that children generally have a hard time remembering specific details long after the fact and often need coaching to be able to accurately recall events (Geddie, Beer, Bartosik, & Wuensch, 2001; Roberts & Powell, 2001). In order to gain more credible child statements and/or disclosures, various guidelines of forensic interviewing have been established (Kuehnle, 1996). The National Child Advocacy Center (NCAC) uses a model of interviewing consistent with the practice guidelines of the American Professional Society on the Abuse of Children (APSAC) (J. Payne, personal communication, February 9, 2005). The NCAC trains forensic interviewers for the various CACs across the country according to these guidelines. APSAC guidelines for investigative interviewing are intended to ?maximize the opportunity for the child to describe what may have happened? while decreasing the likelihood of obtaining incomplete or inaccurate information (APSAC, 2002). Forensic 9 interviews are undertaken to determine if a child has been abused and if so, by whom and in what way. Forensic interviewing is designed to reduce the chances of child suggestibility and the other problems associated with relying on child narratives. The forensic interview is conducted in most cases by a single interviewer. In the rare situations when there are two interviewers, a lead interviewer is designated. Children are interviewed independently of any siblings, parents, or other children. Video recording and/or a one-way mirror can be used. Generally, the forensic interviewer knows the specifics of the allegations before conducting the interview (APSAC, 2002). Interviewers are trained to approach the situation with an open mind and to be warm and supportive toward the child. They are not supposed to make any negative statements about the alleged perpetrator because children are more suggestible when a person has been portrayed as bad in some way. The interviewer can use different methods to help the child give more details about the abuse experience. Anatomical dolls, drawings, anatomical drawings, puppets, flash cards, and a number of other media can be used (APSAC, 2002). APSAC trains its interviewers to build rapport with a child before and in the beginning stages of the interview. They are supposed to follow the pace of the child, and not show any frustration if the child seems unwilling to talk. Ultimately, the forensic interviewer should be non-threatening and supportive. When there are questions about the terminology a child is using, the interviewer clarifies (APSAC, 2002). Before beginning the interview, the forensic interviewer usually sets ground rules (e.g. ?No lying or guessing,? ?You know more about this than I do,? ?If I ask a question more than once, 10 it doesn?t mean you got it wrong the first time.?). This is to reduce the likelihood of the child guessing or misinterpreting the interviewer?s reiterated questions as doubt. In order to elicit as much information as possible, the forensic interviewer first invites the child to talk. This can be done in a very general way (e.g., ?Tell me all about??). This technique allows the child to give as much detail as he/she wants without prompting on the part of the interviewer. The interviewer then moves on to more open-ended questions based on the child?s narrative to acquire greater detail (e.g., ?What did he do when he took off your pants??). Then, if necessary, the interviewer moves to closed-ended questions. The interviewer is trained to avoid using suggestive questions during this and all phases of the interview (e.g. He/she may ask ?Was it day time or night time?? rather than ?Did he touch your privates at night??). Leading and suggestive questions and comments are specifically avoided (APSAC, 2002). The interviewer attempts to get as much detailed information about the event as possible. This includes who, what, where, and when questions. Contextual information (such as what was happening with each person?s clothing) and information about any items that may have been used in or present in the abuse is included. Also, any information about a possible witness is important to obtain. At the end of the interview, the interviewer tries to change the conversation to a more neutral topic so as to end with a neutral or positive feeling. The child can be given the opportunity to ask any questions or say anything he/she wants. The forensic interviewer can thank the child for participating and may give information about what may happen next (without promising anything). The interviewer can also help the child 11 establish a brief safety plan, identifying an adult he/she would talk to if other issues arise (APSAC, 2002). Before the CACEA began conducting forensic interviews by an NCAC-trained interviewer in 1998, the Department of Human Resources (DHR) in Lee County was responsible for them. At that time, a social worker or law enforcement officer conducted the interviews. It is unknown by current DHR workers how interviews were conducted by social workers before the CACEA got an interviewer (D. Sheppard, personal communication, April 14, 2005). There was no set protocol, procedure, or set of guidelines for the law enforcement officers conducting interviews (W. Ramsey, personal communication, April 14, 2005). They often tried to have a female officer conduct the interviews, but when that wasn?t possible, any investigator could interview the alleged child victim. Each interview was highly individualized to the child and detective conducting the interview. Each detective?s training and experience varied; however, they tried not to conduct multiple interviews unless the physical evidence did not corroborate the child?s statement (A. Elkins, personal communication, April 15, 2005). Currently, children are referred to the CACEA when a report of suspected abuse has been made to the DHR or a law enforcement agency. This project evaluated the effectiveness of the forensic interviewing at the Child Advocacy Center of East Alabama (CACEA) in comparison to the traditional abuse interviews employed by the DHR and law enforcement personnel prior to the employment of a trained forensic interviewer by the CACEA. The specific questions to be answered in the current study are: ? Is there an association between agency responsible for interviewing and whether the case was indicted and charged? If so, what is the association? 12 ? Is there an association (among cases that were indicted and charged) between the agency responsible for interviewing and court disposition? If so, what is the association? These questions are answered by the tested hypotheses of the study, which are: ? CACEA forensic interviewing will be statistically more likely to be associated with obtaining an indictment and charging cases than forensic interviewing by the DHR/law-enforcement group. ? CACEA forensic interviewing will be statistically more likely to be associated with successful legal action (pleas and trial convictions) than DHR/law- enforcement, whereas DHR/law-enforcement forensic interviewing will be statistically more likely to be associated with dropped cases and acquittals than the CACEA. 13 METHODS Overview To determine whether the CACEA is fulfilling its purpose of providing more efficient forensic interviewing, the legal outcomes of sexual abuse cases interviewed by the APSAC-trained interviewer at the CACEA were compared to those of cases interviewed by DHR and law enforcement prior to the establishment of forensic interviewing at the CACEA. The independent variable (interview-type) has two levels, CACEA and DHR. There are two dependent variables. The primary dependent variable (charge) has two levels, charged and not charged. The secondary dependent variable (disposition) has four levels: guilty plea, conviction, acquittal, and nolle pross (or dropped). A more detailed operationalization of the variables will be presented later in the Methods section. Participants The participants of this study were the indicated sexual abuse victims for the years studied. Indicated sexual abuse victims are those for which DHR has made the determination that the allegations of abuse can be substantiated and the abuse did occur. It is from this pool that cases are chosen to pursue in criminal court; however, both those whose cases were pursued in criminal court and those whose cases were not were included as participants. Two groups were compared. Group One consists of child 14 sexual abuse cases from the fiscal years 1993 through 1997. These are the last full five years before the CACEA began providing forensic interviews. Group Two consists of indicated cases reported from the calendar years 1999 through 2003. This is the first full five years after the CACEA employed a trained forensic interviewer who was responsible for the interviewing. The distinction in fiscal and calendar years is due to the total number of cases available from the Alabama DHR (ALDHR). More discussion on this can be found in Study Procedures. There was not any direct contact with the participants. Therefore, an expedited IRB was filed and approved for the method of data collection. The participants? identities remained anonymous to the researcher. Agency Procedures DHR-indicated sexual abuse cases. When DHR received a report of suspected or alleged sexual abuse, the case was assigned to a social worker from the Child Abuse and Neglect (CAN) Unit. Each child included in the initial allegation was assigned his/her own case number. For example, if the allegation included three children, the assigned social worker was given three case numbers. The social worker was responsible for investigating the allegation. In cases of a disclosure, he/she attempted to ascertain whether the disclosure was credible. This was done by conducting interviews with family members, school officials, daycare providers, and other acquaintances of the child to determine if the child had a pattern of telling the truth or lying. The social worker also spoke with household members in whom the child may have confided about the alleged abuse or who may have been an eyewitness of the alleged abuse. Any medical evidence 15 was also considered. With supervisory consultation, the social worker then determined whether there was enough evidence to substantiate the allegation (i.e., whether it was indicated). Each case/child received a determination. The method of determining whether an allegation was indicated was consistent across both Groups? time spans (S. Godfrey, personal communication, October 20, 2005). DHR/law-enforcement interviews. As previously stated, there was not a standard interview format during the time span DHR and law enforcement were responsible for the interviewing. However, both DHR and law enforcement officers used techniques that were generally consistent across agencies. The interviewer asked the child to tell him/her what happened. If the child did not know, the interviewer either told him/her or asked if anyone had ever touched him/her in a way he/she did not like. The interviewer took time to build some rapport by asking about who the child lived with, what a typical day at home was like, and other historical questions. Some time was taken to establish the difference between truth and untruth (S. Godfrey, personal communication, October 20, 2005). Police officers stayed with the child for the duration of the interview, intentionally not breaking contact with the child (V. Jackson, personal communication, October 19, 2005). The interviewer asked general questions and tried to clarify what the child said. Generally, interviews lasted an hour, and both DHR and law enforcement tried not to conduct more than one interview with the child (S. Godfrey, personal communication, October 20, 2005; V. Jackson, personal communication, October 19, 2005). CACEA interviews. The CACEA interviews were conducted by a former police officer who was a retired 30-year veteran. The interviewer was trained in forensic 16 interviewing by the National Children?s Advocacy Center in Huntsville, AL and received certification by The American Professional Society on the Abuse of Children (APSAC). CACEA interviews were conducted when DHR or law enforcement agencies received a report of suspected abuse. The agency that received the report assigned a social worker or investigating officer. That person then called the CACEA to schedule an interview for the alleged victim with the forensic interviewer. He/she also provided the interviewer with a referral form with information about the alleged victim, his/her family, and the allegation. The interview was done with the child while the DHR and/or law enforcement officer observed and took notes behind a one-way mirror. The interviews generally lasted 30-40 minutes, although they may have varied from 10 minutes to an hour. The CACEA forensic interviewer began the interview by taking some time to build rapport. She asked the child with whom he/she lived and what school he/she attended. She asked who his/her favorite teacher was, what his/her favorite subject was, and if there were any teachers or subjects he/she did not like. She then asked the child if he/she knew why he/she was at the center. If the child knew, she confirmed that what he/she said was what they would be talking about. If he/she did not know, she told him/her that it was her job to talk with boys and girls who may be having problems at home or when things may have happened to them. She then explained the rules of the forensic interview as follow: - You know more than I about what happened. - Always tell the truth. - No guessing. 17 - If you don?t know or don?t remember, say so. - If I repeat a question, it doesn?t mean the first answer was wrong. - If a question is too hard, we can come back later. - You can correct me if I get something wrong. - You can tell me if you don?t agree with me. The interviewer asked the child if he/she understood the rules. For children who were younger and may have had some difficulty differentiating truth from lies or reality from pretend, the interviewer used examples to clarify. For example, she pointed to a large, gray teddy bear in the room and asked if it was truth or a lie if she were to say that the bear is red. Another thing she may have done was make a comment about the current weather conditions, such as saying it was snowing when it was not, and ask if that was real or pretend. Once the interviewer established that the child knew the difference between truth and lies and real and pretend, she gained confirmation from the child that he/she would follow the rules. She then proceeded with the interview. Often, the children made a partial disclosure by that point. This gave the forensic interviewer an entry point. She asked the child to tell her more about what happened and fleshed out as many details as she could with open-ended questions. Once she obtained a reliable disclosure, she may have asked more closed-ended questions to obtain specifics that had not yet been addressed. If the child had not made a disclosure, the interviewer asked open-ended questions about the people in his/her family. She also may have asked open-ended questions about the alleged perpetrator. If the child did not disclose anything, the interviewer asked the child if he/she knew about the private parts of his/her body. She explained that those parts are the parts covered by the swimming suit. She 18 asked the child if he/she had ever gotten touches on those parts that may have hurt him/her or felt yucky. The interview was then structured based on whether a disclosure was made. Regardless of whether a disclosure had been made, the interviewer left the room toward the end to consult with the DHR and/or law-enforcement officer observing. She did so to insure that all questions those involved may have had were asked. Then she re-entered the room, asked whatever questions she may have been requested to ask, and completed the interview by thanking the child for talking to her. Multi-disciplinary team. Following interviews of both groups, the cases were discussed during the Lee County multi-disciplinary team. The multi-disciplinary team was functional for the duration of both time periods. The team consisted of one of the Assistant District Attorneys, who did not remain constant within or across time periods, the Victim?s Service Officer of the DAs office, who remained constant throughout all ten years included in the study, law enforcement officers, DHR social workers, and during the CACEA time period, the forensic interviewer. The team discussed the allegations and known facts of the case. Using this information and consultation with the team, the ADA decided whether the case should be continued for further investigation, closed, or prosecuted. No standardized procedure was used for making this decision during either of the time periods studied. Study Procedures The DHR interviews were conducted by law enforcement officers and caseworkers in the Child Abuse and Neglect (CAN) Unit at DHR. There was no NCAC- trained forensic interviewer present. The five-year time spans were chosen to assist in 19 obtaining an adequate sample size. Each group has a span of five years included in the study to control for any variance due to time span. The year in which the interviewer began at the CACEA was not included in the study to control for variance due to transition from DHR interviews to CACEA interviews. The total number of indicated child sexual abuse cases for each year included in the study was obtained from ALDHR. ALDHR compiles yearly reports on their abuse cases and provided the number requested as they do not contain any confidential or identifying information about any child victim. During the years included in Group One, ALDHR compiled reports according to the fiscal year. During the years included in Group Two, it compiled reports according to the calendar year. Information on the cases that were charged in criminal court was obtained from the Lee County Justice Center. The circuit clerk obtained reports from the court administrator and provided them for the purpose of this study. The reports included information on the court case number, defendant name, charge, disposition, and court action date for the fiscal and calendar years that correspond to those for which information from ALDHR was obtained. The offense charges included were Rape in the first and second degree, Sodomy in the first and second degree, and Sexual Abuse in the first and second degree. Not all cases carrying these charges involve a minor child. To eliminate the cases involving adult victims, the reports were cross-referenced with CACEA records and those of the Victim Service Officer working in the Lee County District Attorney?s office. The charges chosen for inclusion in the study were done so because it is expected that most sexual abuse reports made to the DHR will involve an allegation of one of these (rather than, for example, transmitting obscene materials to a minor via the internet). The number of 20 cases that were not charged each year is the total number of indicated sexual abuse cases less the number of charged cases per year. No further information was needed on these cases as their years indicated which agency was responsible for the interview (independent variable) and it is understood that they were not charged (primary dependent variable). The report obtained from the Justice Center included dispositions on probation cases with an original charge of child sexual abuse. These cases were not included in the study because they had already ended in a conviction at an earlier date. However, if the original court case was disposed during one of the time periods studied, it was included in its corresponding group. Also excluded from the study were criminal court cases in which the defendant was being charged with child sexual abuse, but the alleged victim was an adult when he/she reported the abuse. This is because these cases would not have been included among the indicated DHR cases since the victim was an adult when the abuse was reported. In cases where there were multiple charges for each child victim, all charges were considered one case since the number of indicated cases from DHR signifies the number of child victims, not the number of possible charges. When each charge had a different disposition but one of the dispositions was a conviction or guilty plea, the case was considered convicted or plead. For example, a defendant may be charged with first degree Rape and first degree Sexual Abuse for abusing one child. The defendant may plead guilty to the charge of Sexual Abuse in return for having the Rape charge dropped. This case would be considered a guilty plea even though one of the charges would be dropped. If the defendant is acquitted of the Rape charge but convicted of the Sexual Abuse charge, it is still considered a conviction. This is because the focus 21 is on whether or not the alleged perpetrator must account for his crime. A guilty plea or conviction on a charge assumes that he does. Using all indicated child sexual abuse cases, the two groups were compared in terms of the number of criminal court cases each had for its five-year time span. Using charged case, the CACEA-interviewed group and DHR/law enforcement-interviewed group were compared in relation to legal outcomes (or dispositions) obtained in the alleged perpetrators? criminal cases. Variables The independent variable (interview type) has two levels that were compared, CACEA and DHR. CACEA refers to the cases of children who were interviewed by the NCAC trained forensic interviewer employed by the CACEA. DHR refers to the cases of children who were interviewed either by a DHR worker or law enforcement officer. There are two dependent variables. The primary (criminal charge) refers to whether the case was indicted by the grand jury and charged. The following two outcomes are possible when a case is presented to the grand jury: 1.) An indictment or true bill is when the grand jury rules that there is enough evidence to continue with the case and go to trial if necessary. 2.) A no-bill is when the grand jury rules that there is not enough evidence to proceed with the charges. Cases that were indicted were charged and comprise the first level of this variable, whereas cases that were no-billed were not charged and comprise the second level of the variable. In addition to cases that were no- billed, the second level includes cases that were not pursued by the District Attorney or presented to the grand jury. It cannot be determined how many of the uncharged cases 22 were not pursued and how many were no-billed because grand jury proceedings are considered secret, and the researcher did not have access to how many cases were charged and no-billed (N. Abbett, personal communication, April 15, 2005). Therefore, the second level of this variable comprises all cases that were not charged with one of the criminal charges included in the study, whether no-billed or not pursued. The secondary dependent variable (disposition) refers to the legal results or the court disposition of the charged abuse cases. The following six outcomes are possible when a case is charged: 1.) A plea is when the alleged perpetrator chooses to enter into a bargain with the District Attorney to avoid trial. He/she pleads guilty, usually in return for a lesser sentence. It is considered a conviction of the crime. Similarly, 2.) a waived to grand jury disposition means that a defendant wishes to waive his/her right to a grand jury proceeding and enter a guilty plea. These two dispositions are grouped together as one level of the secondary dependent variable since they are both considered by the court to be a guilty plea. 3.) A conviction is when the jury finds the alleged perpetrator guilty of the crime. 4.) An acquittal is when the case goes to trial, but a jury finds the alleged perpetrator not guilty. 5.) A nolle pros means that the charge was not processed, or dropped by the DA. This may happen for a number of reasons, including a plea to a lesser charge, death of the defendant, or obtaining new facts to the case that change the DAs desire to pursue the case. 6.) A withdrawn and filed disposition means that the charge was withdrawn from the docket and filed in the clerk?s office to be processed at a later date. One example of a reason for this may be that the defendant is carrying a capital charge in another state and that charge needs to be processed first (V. Teague, personal communication, July 27, 2005). In the sample for this study, certain cases were 23 also incomplete. This is because the case is still in the middle of the legal process and no court disposition has been made. These incomplete cases, along with the cases that were withdrawn and filed were not included in the study. Therefore, the second dependent variable (disposition) has four levels: guilty pleas (including those waived to grand jury), convictions, acquittals, and nolle prosses. Guilty pleas and trial convictions are considered successful legal action due to the idea that perpetrators should account for and be penalized for their actions. However, sentencing was not considered in this study because the emphasis is on whether or not CACEA forensic interviewing increases the likelihood of positive court action (indictment, plea bargain or conviction), not which specific legal penalties follow. This information came directly from the court?s records. 24 RESULTS Data Preparation The original data set consisted of 721 cases. Four cases were excluded because they were indicted by the grand jury but as of this writing are incomplete (have not received a disposition in the court). The data set then consisted of a total of 717 cases. A total sample of cases was used for the first analysis, testing the first hypothesis. The CACEA group included 229 cases, while the DHR group consisted of 488 cases. Only the cases that were charged comprise the sample for the second analysis, testing the second hypothesis. The secondary dependent variable, criminal charge, was controlled, resulting in N = 105 cases. In this analysis, the CACEA group included 53 cases, while the DHR group included 52 cases. Analyses In order to address the hypotheses of this study, Pearson chi-square tests were used. The chi-square is a nonparametric test of association. The assumptions for the chi- square are that the sample was randomly drawn from the population, the data was reported in raw frequencies, the measured variables are independent, and the categories on independent and dependent variables are mutually exclusive and exhaustive (Connor- Linton, 2003). The level of statistical significance was set at .05. Residuals identify which cells depart most from the model of independence. The residuals were analyzed to 25 determine which cells differed from their expected values and therefore contributed to any statistically significant associations. The standard residual is calculated to test the significance of individual cells. It is obtained by dividing the cell?s residual by the square root of its expected count and is an approximate fit to the normal standard distribution. This method, however, may exclude some cells that deviate significantly from the model of independence. Therefore, adjusted residuals were used in this study. The adjusted residual is obtained by dividing the standard residual by the estimated standard error and is read roughly as a z score. Therefore, values equal to or less than -2 or equal to or greater than +2 are considered significant at the 95% confidence level (?1.96, rounded to ?2) and depart markedly from the model of independence (?Analysing Tables?, n.d.). Additionally, residuals significant at the 90% confidence level (?1.65) are reported as trends. Hypothesis Testing Hypothesis one. CACEA forensic interviewing will be statistically more likely to be associated with obtaining an indictment and charging cases than forensic interviewing by the DHR/law-enforcement group. A chi-square analysis of this hypothesis yielded a value of 19.447 with p<.001 (N=717, df=1). Because the obtained chi-square value reaches significance at the 99% confidence level, the null hypothesis can be rejected. There is a statistically significant association between the interview type and whether the case is indicted and charged. In order to further analyze the observed association, the adjusted residuals were examined. CACEA interviews appear to be twice as likely to yield indictments and thus 26 charged cases than traditional interviews done by DHR and/or law enforcement. 23.1% of cases interviewed by the CACEA forensic interviewer were charged compared to 10.7% of cases interviewed by DHR/law enforcement (See Table 1). Table 1 Crosstabulation of Agency and Criminal Charge DHR CACEA Observed N 52 53 Expected N 71.5 33.5 Percent within agency 10.7% 23.1% Standard Residual -2.3 3.4 Charged Adjusted Residual -4.4 4.4 Observed N 436 176 Expected N 416.5 195.5 Percent within agency 89.3% 76.9% Standard Residual 1 -1.4 Not Charged Adjusted Residual 4.4 -4.4 Hypothesis two. CACEA forensic interviewing will be statistically more likely to be associated with successful legal action (pleas and trial convictions) than DHR/law- enforcement, whereas DHR/law-enforcement forensic interviewing will be statistically more likely to be associated with dropped cases and acquittals than the CACEA. To test this hypothesis, the primary dependent variable, criminal charge, was controlled while a chi-square crosstabulation was conducted on the association between interview type and disposition. The chi-square analysis yielded a value of 8.021 with a significance of .046 (N=105, df=3). Therefore, the null hypothesis can be rejected because there is a statistically significant association (at the 95% confidence level) 27 between the agency responsible for interviewing and disposition of cases that were charged. In order to investigate the observed association, follow up analyses of the adjusted residuals were conducted to determine which variables account for the statistically significant association. The adjusted residuals indicate that the CACEA interviews are nearly four times more likely to yield trial convictions than traditional interviews done by DHR and/or law enforcement. 20.8% of cases interviewed by the CACEA interviewer and charged ended in a trial conviction, whereas 5.8% of DHR interviews that were charged ended in a trial conviction. There are other trends; however, they do not meet the criteria for significance at the 95% confidence level when analyzing adjusted residuals (See Table 2). These trends are addressed further in the Discussion section. Therefore, the null hypothesis can be rejected in terms of convictions, but it cannot be rejected in terms of guilty pleas, acquittals and nolle prosses. 28 Table 2 Crosstabulation of Agency and Legal Outcome DHR CACEA Observed N 34 26 Expected N 29.7 30.3 Percent within agency 65.4% 49.1% Standard Residual 0.8 -0.8 Guilty Plea Adjusted Residual 1.7 -1.7 Observed N 3 11 Expected N 6.9 7.1 Percent within agency 5.8% 20.8% Standard Residual -1.5 1.5 Conviction Adjusted Residual -2.3 2.3 Observed N 2 6 Expected N 4.0 4.0 Percent within agency 3.8% 11.3% Standard Residual -1.0 1.0 Acquittal Adjusted Residual -1.4 1.4 Observed N 13 10 Expected N 11.4 11.6 Percent within agency 25.0% 18.9% Standard Residual 0.5 -0.5 NOLLE PROS Adjusted Residual 0.8 -0.8 * More than 20% of the cells have expected counts less than 5, thus the validity of the results is questionable. 29 DISCUSSION To date, no empirical evidence has been published to indicate that the forensic interviewing model developed by APSAC and used by NCAC (or CACEA) is any different, much less more effective, than the traditional model and/or techniques used by social workers and law enforcement personnel. The statistically significant difference in whether cases were charged suggests that there is a difference between the forensic interviewing model used by the CACEA and the traditional DHR/law enforcement interviews. Further, the direction of the association, as evidenced by the residual values, would suggest that CACEA interviews are more effective in leading cases to being charged. The literature indicates that 94% of cases charged end in a conviction (Cross et al., 2003), but that only about 24% to 28% of alleged perpetrators are charged (Finkelhor, 1983; Stroud et al., 2000; Cross et al., 2003). This suggests that the most important indicator of whether or not the forensic interviewing at the CACEA is more effective than traditional DHR or law enforcement interviewing is the rate at which the District Attorney prosecutes cases referred to him. This study found that once the CACEA became responsible for forensic interviewing, the percentage of charged cases increased significantly (to 23.1%), essentially equal to the reported 24% to 28% found in the literature (Finkelhor, 1983; Stroud et al., 2000; Cross et al., 2003). In comparison, when DHR was responsible for the interviewing, the rate of charges was less than half (10.7%) 30 of the literature reported rate. Therefore, the increased number of charged cases in Lee County, AL during the time period the CACEA was responsible for interviewing suggests that interview-type is important to cases charged. This would further suggest that forensic interviewing, as developed by APSAC, is more effective in leading cases to prosecution. Faller and Henry (2000) reported an increase of criminal charges in a community that had a functioning multidisciplinary team. However, the effect found in this study cannot be explained by the collaboration of a multi-disciplinary team because the multi- disciplinary team in Lee County was functional during both Groups? five year time- spans. Further, the Victim?s Service Officer (VSO) in the District Attorney?s office remained constant throughout the ten years included in the study. The VSO in Lee County was largely responsible for attending MDT meetings and providing input regarding which cases to charge. Conversely, a new District Attorney was elected in 1999, and his prosecution style may have differed from that of the previous District Attorney. If the later DA was a more aggressive prosecutor, it would be expected that more cases were charged during the time period he was in office. However, the later DA was the Chief Assistant District Attorney prior to his election, and may have had a significant influence on which cases were charged prior to his own election. Further, attending the MDT meetings and deciding which cases to prosecute was not the responsibility of the DA, rather one of his Assistant District Attorneys. It is unknown whether the DA?s prosecutorial style affected which cases were chosen by the ADA to charge. The possible effects of having different DAs for the two time periods in this 31 study may account for some of the results, but this cannot be substantiated or refuted in the current study. Future research should focus on controlling for variance due to differing prosecutorial style. This may partially be done by analyzing the rate of successfully charged cases (indicted cases) to cases presented to the grand jury but no-billed. In the current study, no-billed cases are included among the cases that were ?not charged.? This is because the Lee County District Attorney reported that he does not have any compiled data on how many cases were indicted versus no-billed by the grand jury during the years included in the study. Asserting that grand jury proceedings are considered secret, permission was not granted to review grand jury dockets to obtain this information (N. Abbett, personal communication, April 15, 2005). As such, the research does not indicate how many cases were pursued for prosecution by the DA, only how many were indicted. The number of indicted versus no-billed cases would be important to know because if the CACEA group has a statistically significantly higher number of no-billed cases than the DHR group, the data would suggest that the effects found in the current study were more associated with prosecutorial style. If, however, there is no statistical significance in the difference between groups, or if the CACEA has fewer no-bills, the test of the first hypotheses would be less confounded and the results more meaningful. Since the CACEA has been responsible for the forensic interviewing, there has been some statistically significant change in the dispositions of court cases. When analyzing the residual values, it becomes apparent that though there are only two cells (one complete level of the secondary dependent variable) that statistically significantly contribute to the effect at the 95% confidence level, there are trends at the 90% 32 confidence level in two other cells. As hypothesized, the CACEA is associated with significantly more convictions than expected, whereas DHR is associated with fewer convictions than expected. These figures represent the circumstances in which a trial jury found the alleged perpetrator guilty of the highest charge possible for the abuse committed. Only one of the cases (in the CACEA group) carried two charges with different dispositions, one acquittal and one conviction. Nevertheless, the CACEA cases that ended in convictions more than tripled the DHR cases that ended in convictions (20.8% versus 5.8%). The CACEA also has a trend toward less guilty pleas, while DHR has a trend toward more guilty pleas. The higher number of CACEA convictions and lower number of guilty pleas than expected indicates that more cases during the CACEA time period went to trial as opposed to being pled. This may support the assumption that the DAs office was more aggressive during the later time period and attempted to obtain more trial convictions than plea bargains. Whether or not this resulted in more overall successful legal action can be seen by collapsing the values of the data. When guilty pleas and convictions are combined (and acquittals and nolle prosses, or dropped cases, are combined) and analyzed, the chi-square does not reach significance (N=105, df= 1, ? 2 =.076, sig.=.782). Overall, there is no significant difference between agencies in the success of the charged cases. So, although the second null hypothesis can be rejected, the collapsed (overall) numbers help explain why the value of the post-hoc is so low. Moreover, when looking at the collapsed values, 71.2% of DHR cases that were charged ended in a guilty plea or conviction, and 69.9% of CACEA cases that were charged ended in a guilty plea or conviction. These rates, though not significantly 33 different from each other, differ greatly from the reported 94% average across 21 studies (Cross et al., 2003), indicating that Lee County is not representative of other counties in the U.S. This finding, together with the fact that the total number of guilty pleas and convictions did not change significantly after the CACEA became responsible for the interviewing, underscores the above assertion that rate of cases charged is the greatest indicator of interview effectiveness. Nonetheless, it should be reiterated that three times as many defendants were found guilty of the highest charge possible during the CACEA time period, rather than pleading guilty to a lesser charge. This difference may be due to increased credible disclosures while the CACEA was conducting interviews. Limitations and Future Directions An important limitation to this study is its size. When testing the second hypothesis, more than 20% of the cells in the crosstabulation had an expected count lower than five. As such, the results of this hypothesis are questionable. Future studies should include larger sample sizes to avoid this limitation. The largest limitation to this study is the data available. Besides not knowing how many cases were presented to the Grand Jury and no-billed, the data also do not reflect some of the circumstances of the cases that would be important to consider. The data from the reports generated by the Lee County Justice Center are purely quantitative of the specific charges included in the study and exclude some important qualitative facts. For example, if a defendant is charged with abusing more than one child, he/she may have plead guilty to abusing one child in return for having the charges on the other child(ren) dropped. The data available do not indicate when such things may have 34 happened. This would affect the outcomes of both groups because a case considered dropped may actually have been part of a guilty plea. The data also do not reflect the following: the family of the victim decides not to prosecute; the defendant is later pardoned; the large number of nolle prosses may be due to a plea to a charge not included in the study (e.g., contributing to the delinquency of a minor) or defendant?s death. All of these are factors that confound the results of the current study. Future studies should focus on obtaining more detailed information from court files. As already noted, future studies should also control for the effects of varying prosecutorial styles. As mentioned, this can be done in part by obtaining more data regarding Grand Jury proceedings. The data do not need to include confidential information, only the numbers of cases presented to the Grand Jury and subsequently no- billed and indicted. Conclusions Currently, NCAC-trained interviewers employ the model they have been certified to use without any empirical evidence to validate their interview style. The CACEA follows suit. To date, this study is the only one that assesses APSAC forensic interviewing. It suggests that there is not only a difference between APSAC and traditional interviewing, but that there is also a difference that can increase offender accountability and hopefully deterrence of the crime. Twice as many child sexual abuse cases carried criminal charges when the CACEA became responsible for interviewing. Furthermore, CACEA interviews led to more than three times the number of trial convictions than traditional DHR/law enforcement interviews. These findings indicate 35 that CAC forensic interviewing is closely connected to more positive legal outcomes and has major implications for the role of Children?s Advocacy Centers? forensic interviewing in the criminal justice system. As a result, this study should act as a springboard for future evaluative studies of forensic interviewing. Comparative, large-scale studies can be done to further assess and validate the forensic interviewing model used by Children?s Advocacy Centers across the nation and internationally. If studies of this kind are done, and forensic interviewing is further validated, they may influence the criminal justice system in such a way that offenders are required to more fully account for their crimes. This may in turn act to deter perpetration and decrease the victimization of our children. 36 REFERENCES American Professional Society on the Abuse of Children (2002). Investigative interviewing in cases of alleged child abuse: Practice Guidelines. Chicago: Author. Analysing tables, Part V, Interpreting chi square. (n.d.). Retrieved October 23, 2005, from http://oassis.gcal.ac.uk/rms/pdf/tables5.pdf. Blodgett, N. (1987). . . .Child abuse: Alabama community finding success through cooperation. ABA Journal, 73(5), 27. Bradford, R. (1994). Developing an objective approach to assessing allegations of sexual abuse. Child Abuse Review, 3, 93-101. Bradshaw, T.L., & Marks, A.E. (1990). Beyond a reasonable doubt: Factors that influence the legal disposition of child sexual abuse cases. Crime & Delinquency, 36, 276-285. Carnes, C.N., Nelson-Gardell, D., Wilson, C., & Orgassa, U.C. (2001). Extended forensic evaluation when sexual abuse is suspected: A multisite field study. Child Maltreatment, 6, 230-242. Carnes, C.N., Wilson, C., & Nelson-Gardell, D. (1999). Extended forensic evaluation when sexual abuse is suspected: A model and preliminary data. Child Maltreatment, 4, 242-254. Connor-Linton, J. (2003). Chi square tutorial. Retrieved October 20, 2005, from Georgetown University, Faculty Web site: http://www.georgetown.edu/faculty/ballc/webtools/web_chi_tut.html Cross, T.P., Walsh, W.A., Simone, M., & Jones, L.M. (2003). Prosecution of child abuse: A meta-analysis of rates of criminal justice decisions. Trauma, Violence, & Abuse, 4, 323-340. De Jong, A.R., & Rose, M. (1991). Legal proof of child sexual abuse in the absence of physical evidence. Pediatrics, 88, 506-511. Faller, K.C., & Henry, J. (2000). Child sexual abuse: a case study in community collaboration. Child Abuse & Neglect, 24, 1215-1225. 37 Finkelhor, D. (1983). Removing the child- Prosecuting the offender in cases of sexual abuse: Evidence from the national reporting system for child abuse and neglect. Child Abuse & Neglect, 7, 195-205. Geddie, L.F., Beer, J., Bartosik, S., & Wuensch, K.L. (2001). The relationship between interview characteristics and accuracy of recall in young children: Do individual differences matter?. Child Maltreatment, 6, 59-68. Gibson, L.E., & Leitenberg, H. (2000). Child sexual abuse prevention programs: Do they decrease the occurrence of child sexual abuse?. Child Abuse & Neglect, 24, 1115- 1125. Gilstrap, M.S. (2003, June 18). Attorney General Opinion No. 2003-17, Kansas. Hodson, D., & Skeen, P. (1987). Child sexual abuse: A review of research and theory with implications for family life educators. Family Relations, 36, 215-221. Kellogg, N.D. (2002). Child sexual abuse: a marker or magnifying glass for family dysfunction?. Social Science Journal, 39, 569-582. Kendall-Tackett, K.A., Williams, L.M., & Finkelhor, D. (1993). Impact of sexual abuse on children: A review and synthesis of recent empirical studies. Psychological Bulletin, 113,164-180. Kuehnle, K. (1996). Assessing allegations of child sexual abuse. Sarasota, FL: Professional Resources Press. Lindsey, D., & Regehr, C. (1993). Protecting severely abused children: Clarifying the roles of criminal justice and child welfare. American Journal of Orthopsychiatry, 63, 509-517. Martin, M.J. (1992). Child sexual abuse: Preventing continued victimization by the criminal justice system and associated agencies. Family Relations, 41, 330-333. Myers, J.E.B., Diedrich, S., Lee, D., Fincher, K.M., & Stern, R. (1999). Professional writing on child sexual abuse from 1900 to 1975: Dominant themes and impact on prosecution. Child Maltreatment, 4, 201-216. Prior, V., Glaser, D., & Lynch, M.A. (1997). Responding to child sexual abuse: The criminal justice system. Child Abuse Review, 6, 128-140. Roberts, K.P., & Powell, M.B. (2001). Describing individual incidents of sexual abuse: A review of research on the effects of multiple sources of information on children?s reports. Child Abuse & Neglect, 25, 1643-1659. 38 Sauer, M., & Okerblom, J. (1993, September 6). Trial by therapy. National Review, pp. 30-39. Saywitz, K.J., Mannarino, A.P., Berliner, L., & Cohen, J.A. (2000). Treatment for sexually abused children and adolescents. American Psychologist, 55, 1040-1049. Shapiro, L. (1993, April 19). How to safeguard your own children. Newsweek, pg. 60. Spungen, C.A., Jensen, S.E., Finkelstein, N.W., & Satinsky, F.A. (1989). Child personal safety: Model program for prevention of child sexual abuse. Social Work, 34(2), 127-131. Stewart, A. (1999, March). Domestic violence: Deterring perpetrators. Paper presented at the 3 rd National Outlook Symposium on Crime in Australia, Mapping the Boundaries of Australia?s Criminal Justice System, Canberra, Australia. Stroud, D.D., Martens, S.L., & Barker, J. (2000). Criminal investigation of child sexual abuse: A comparison of cases referred to the prosecutor and those not referred. Child Abuse & Neglect, 24, 689-700. The Harvard Law Review Association (1985). The testimony of child victims in sex abuse prosecutions: Two legislative innovations. Harvard Law Review, 98(4), 806-827. Wood, J.M., & Garven, S. (2000). How sexual abuse interviews go astray: Implications for prosecutors, police, and child protection services. Child Maltreatment, 5, 109- 118. Woody, J.D. (2002). Media coverage of child sexual abuse: An opportunity for family therapists to help families and communities. The American Journal of Family Therapy, 30, 417-426.