How do children tell? The disclosure process in child abuse
The disclosure process in child sexual abuse
By April R. Bradley and James M. Wood
Child Abuse & Neclect. Vol. 20, No. 1). pp. 881 891, 1996
Copyright © 1996 Elsevier Science Ltd
Pergamon Printed in the USA. All rights reserved. No part of this article may be reproduced in any way without the publishers express permission.
Reproduced here with the kind permission of the publisher.
Department of Psychology, The University of Texas at EI Paso, EI Paso, TX, USA
Research indicates that 15% to 40% of females, and up to 10% of males, are victims of child sexual abuse (Salter. 1988). In a large proportion of validated sexual abuse cases, the victim's statement constitutes the only evidence that abuse has occurred (Rieser, 1991). According to some reports abused children commonly deny that abuse has occurred, or later recant their allegations (Sorenson & Snow, 1991; Summit, 1983). Such denials and recantations tend to compromise the child's credibility and complicate the legal response to sexual abuse allegations (Myers. 1992).
Theorists and researchers have attempted to explain why sexually abused children deny or recant sexual abuse. Koverola and Foy (1993) have reported that victims of sexual abuse often display symptoms of Post Traumatic Stress Disorder (PTSD). According to Koverola and Foy (1993), "children suffering from PTSD often enter an avoidance phase," in which they deny abuse or recant because they cannot cope with the anxiety aroused by traumatic memories. According to Koverola and Foy (1993), anxiety about court appearances or a change in the home environment may lead to denial or recantation.
Rieser ( 1991 ) has suggested that children may recant if they feel isolated from their natural support systems. Other writers speculate that recantation and denial may be a way for victims to "make it all go away" (Gonzalez, Waterman, Kelly, McCord, & Oliveri. 1993). In a summary of anecdotal evidence from psychotherapists. Fontes (1993) found that certain cultural groups may recant more often than others due to cultural taboos regarding sexual abuse. Research suggests that loyalty to family members. or fear of their reaction to abuse allegations. may also contribute to some denials, recantations, and reluctance to disclose (Farrell, 1988. Lawson & Chaffin, 1992; Sauzier. 1989).
Some authors have argued that denial and recantation are part of a process experienced by many victims of sexual abuse (Sorenson & Snow. 1991; Summit. 1983). In an influential article, Summit (1983; see also Summit. 1992) has described the disclosure process as the "Child Sexual Abuse Accommodation Syndrome" (CSAAS). The CSAAS consists of five stages: (1) secrecy; (2) helplessness: (3) entrapment and accommodation; (4) delayed, unconvincing disclosure; and (5) retraction. According to Summit (1983), children retract their statements as part of a process of dealing with sexual abuse victimization.
Summit (1983) did not present research to support his description of CSAAS. However, an empirical study of the disclosure process by Sorenson and Snow 1991 has produced findings consistent with Summit's (1983) views. Sorenson and Snow 1991 examined 117 cases in which a finding, of sexual abuse was supported by medical evidence, perpetrator confession, or criminal conviction. Sorenson and Snow (1991) found that most disclosures of abuse were accidental (74%) and that many victims (22%) recanted their statements only to reaffirm them later (93% of recantations). Seventy-two percent of victims initially denied abuse, and 78% were reluctant to discuss the abuse.
Sorenson and Snow (199 1. p. 14) concluded. "These findings may reinforce the importance of viewing disclosure as a process with phases that. like developmental tasks, can be resolved." According to Sorenson and Snow (1991), the disclosure process includes five stages: (1) initial denial that the abuse occurred; (2) tentative disclosure or reluctance to discuss the abuse: (3) active disclosure or a complete statement about the abuse; (4) recantation of the allegations; and (5) reaffirmation of the allegations.
Other researchers have also reported on the rate of recantations in sexual abuse cases. Jones and McGraw (1987) found an 8% recantation rate among 309 reliable sexual abuse cases seen by a child protection agency. Gonzalez and colleagues (1993) found a substantially higher recantation rate of 27% among 63 sexually abused children in therapy. However, the subjects of Gonzalez and colleagues (1993) were younger children (mean age = 6.5) who had been identified as victims of ritual abuse, and therefore were not typical of the victims seen by child protection agencies and therapists.
At present, research findings regarding disclosure of sexual abuse are scanty and inconsistent. Only one study (Sorenson & Snow, 1991) has examined the frequency with which victims deny or are reluctant to report abuse to interviewers. The three studies that have examined recantation have yielded rather discrepant results, ranging from an 8% recantation rate (Jones & McGraw. 1987) to a 27% rate (Gonzalez et al., 1993). Clearly more research is needed before theories about disclosure can be accepted or rejected on a scientific basis. The present study set out to shed additional light on patterns of disclosure in child sexual abuse.
Two hundred forty-nine validated sexual abuse cases from the years 1992 and 1993 were selected from the files of the Department of Protective and Regulatory Services (DPRS) in El Paso, Texas. Some files from those years were unavailable for the study because (a) the cases were still open for services: (b) in accordance with state law, the files had been destroyed when the victim reached 18 years of age or (c) the files could not be located for other reasons. The cases in the present study constituted 67% of the DPRS validated cases for 1992, and 56% for 1993.
Demographic variables and case characteristics
Demographic variables and case characteristics recorded from case flies included the following (a) age, gender, and ethnicity of the victim, caretaker and perpetrator; (b) the nature of the current allegations and any past allegations of abuse; (c) duration. Frequency, and types of abuse; (d) any disclosures by the child prior to contact with DPRS or police; (c) frequency and types of DPRS and police contacts with the victim; (f) DPRS services provided to the victim or his/her family; (g) information regarding the safety of the child after the case was opened.
Cases studied by Sorenson and Snow (1991) met at least one of the following three strict inclusion criteria: (1) positive medical evidence; (2) confession by the perpetrator; or (3) criminal conviction. These variables were also scored in the present study, and a sub-analysis was performed on cases that met the Sorenson and Snow (1991) inclusion criteria.
Scoring of the disclosure process
The crucial issue in this study was the nature of the disclosure process. Each interview of the victim by DPRS or the police was scored for the following disclosure events: (1) denial, (2) general reluctance and specific reluctance (Sorenson and Snow's (1991) "tentative disclosure"); (3) first disclosure and new disclosure (Sorenson and Snow's (1991) "active disclosure"); (4) repetition of the same information; (5) total recantation and partial recantation; (6) reaffirmation: (7) DPRS or police contact without any other disclosure event.
Operational definitions of each disclosure event are listed in Appendix A. DENIAL was scored if a victim denied that abuse took place. RELUCTANCE/GENERAL and RELUCTANCE/SPECIFIC were scored if the victim was reluctant to discuss the general topic of abuse or specific details of abuse incidents, respectively. DISCLOSURE/FIRST was scored for the interview in which the victim first disclosed sexual abuse to a DPRS or police interviewer, and DISCLOSURE/NEW was scored for any interviews in which the victim revealed new details about the abuse after the first disclosure. REPEAT was scored if the child repeated previous disclosures of sexual abuse. for example, when making statements to the police. RECANTATION/ALL and RECANTATION/PART were scored if the victim recanted all previous disclosures of sexual abuse or specific details of the abuse, respectively. REAFFIRMATION was scored if the victim reaffirmed a previous allegation, following recantation. NONE was scored for an interview in which no other disclosure event occurred.
A hypothetical abuse case may illustrate the scoring system. Consider a hypothetical 13 year old victim who has recently been diagnosed with genital warts by her physician. DPRS is contacted and an interview conducted. In the first interview, the victim denies any sexual abuse and says she does not know how she contracted the sexually transmitted disease (DENIAL). A second interview is conducted 3 days later by DPRS and the victim refuses to discuss the subject of sexual abuse in any manner (RELUCTANCE/GENERAL). She states that she does not want to talk and asks DPRS to leave her alone. Four days later the victim's mother brings her to the DPRS office and the victim states that her father has been fondling her unclothed genitals for 2 years (DISCLOSURE/FIRST). The next day DPRS comes to the victim's home and asks her how she thinks she contracted the genital warts. At this time victim states that her father has been having intercourse with her for 1 year (DISCLOSURE/NEW). DPRS contacts the police and the victim is brought to the police station to make a statement. During the police interview, the child states that she was never abused (RECANTATION/ALL). She states that she lied because she was angry with her father. Later in the same interview, the child admits that her father did sexually abuse her, and that everything she told DPRS was true (REAFFIRMATION). However, she refuses to give details of the abuse (RELUCTANCE/SPECIFIC). One week later DPRS reinterviews the victim. She states that she does not want to discuss intercourse with her father. However, she does describe him fondling her (RELUCTANCE/SPECIFIC). Toward the end of the same interview, the victim again describes intercourse with her father (REPEAT). One month later the case worker makes contact with the victim to see how she likes counselling (NONE).
In this hypothetical abuse case, there are 10 disclosure events and eight interviews. All disclosure events occur in this case at least once.
The measurement tool used in the present study allowed raters to record which disclosure events occurred, and in which interviews they occurred. Raters were the first author and three advanced undergraduate assistants. The first author provided one-on-one training to the assistants and monitored their adherence to the scoring rules throughout the project.
Case files were scored for 249 victims. Fifteen victims had not been interviewed by DPRS or police, usually because the child was too young or had moved to another locale. These cases were eliminated from the analyses, leaving a sample of 234 (82% female; 18% male).
Twenty percent of victims were 1 to 5 years old, 36% were 6 to 10, 36% were 11 to 14, and 7% were 15 to 18 (mean age = 10.0; SD = 3.8). Seventy-six percent of victims were Hispanic, 15% were non-Hispanic White and 9% were either African American or a member of another ethnic group. Victim ethnicity was approximately representative of the population of EI Paso.
In 96% of the sample, the primary caretaker was an immediate family member, most often the mother (92%). Annual family income was between $0 and $8,999 in 38% of cases, between $9,000 and $17,999 in 33%, between $18,000 and $33,999 in 23%, and between $34,000 and $62,999 in 7%. Most households contained three to five members (88%). Most families in the sample would be considered lower middle class or just above the poverty level. Additional information regarding demographics and other case characteristics is provided in Bradley (1995).
Characteristics of the sexual abuse allegations
Most cases involved one victim (78%) and/or one perpetrator (95%). No case involved more than two perpetrators. The number of abuse incidents was one in 19% of cases, two to five in 26%, six to nine in 16%, and 10 or more in 40%. The perpetrator was a member of the household (residing with family for more than 6 months) in 59% of cases. The perpetrator was an immediate family member in 34% of cases, and an extended family member in an additional 32%.
Forty-nine percent of cases involved penetration (anal, vaginal, oral, or unspecified). Fondling of unclothed genitals was the most common type of abuse (55%) and fondling of clothed genitals the second most common (36%). Percentages do not add to 100% because some cases involved more than one type of sexual contact.
In 27% of the sample there was evidence that the victim had not been effectively isolated from the perpetrator while the case was open. This figure included cases in which the perpetrator was present in the home during unexpected home visits by the case worker, or the victim reported contacts with the perpetrator after the case was opened.
Disclosures prior to DPRS or police contact
Very few victims (6%) directly contacted DPRS or police to make the initial report of abuse. However, most victims (72%) had disclosed abuse to someone else prior to contact with DPRS or police. Victims most often made a first disclosure to an immediate family member (35%), extended family member or friend (16%), or school official (13%). Case files did not indicate whether the victims' initial disclosures were accidental or intentional.
Frequency of disclosure events
Ninety-six percent (226) of victims in the present study made a partial or full disclosure of abuse during at least one interview with DPRS or police. Eight cases were validated without a disclosure by the victim.
Six percent (13) of victims initially denied to DPRS or police that abuse had occurred. This figure was significantly lower (Zprop = 13.1, p < .001) than the 72% rate of initial denials reported by Sorenson and Snow (1991). In over one-half of the cases involving denial, the victim made a complete disclosure of abuse within the same interview as the denial, or in the next interview.
Ten percent (24) of victims in the present study displayed reluctance to discuss the topic of abuse, or specific aspects of the abuse, during one or more interviews with DPRS or police. This figure was significantly lower (Zprop = 12.7, p < .001) than the 78% reluctance rate ("tentative disclosure") reported by Sorenson and Snow (1991).
Three percent (8) of victims in the present study recanted their allegations to DPRS or police. This figure was significantly lower (Zprop = 5.5, p < .001) than the 22% recantation rate reported by Sorenson and Snow (1991). When the eight victims who never disclosed (and therefore could not recant) were eliminated from the analysis, the recantation rate was 4%, a figure significantly lower (Zprop = 5.4, p < .001) than the rate reported by Sorenson and Snow ( 1991 ). Each of the eight cases of recantation is described in detail below.
Chi-square analyses were performed on the relation of denial, reluctance, and recantation with perpetrator confession or conviction, medical evidence, victim age, type of abuse, use of threats, and primary caretaker protectiveness. No statistically significant results were found, perhaps because of the rather low frequency of denial, reluctance, and recantation in the present sample.
Cases meeting the inclusion criteria of Sorenson and Snow (1991)
One hundred twenty-one (52%) of the present cases met the inclusion criteria employed by Sorenson and Snow (1991). Specifically, (a) in 14% of the 234 cases, medical evidence indicative of abuse was present; (b) in 33% of the cases, the perpetrator gave a full or partial confession to DPRS (38% of confessions), police (14% of confessions), or other individuals (47% of confessions); and (c) in 9% of the cases, the perpetrator was convicted in a criminal court. These figures do, not sum to 52% because some cases met more than one inclusion criterion.
Disclosure events were analyzed for the subset of 121 cases that met the inclusion criteria of Sorenson and Snow (1991). In this subset, disclosures occurred in 95% (115) of cases, initial denials occurred in 7% (9) of cases, reluctance in 12% (15), and recantations in 3% (4). Six of the 121 victims never disclosed, and therefore could not recant. When these six victims were eliminated from the analysis, the recantation rate was 3%. None of these figures were significantly different from those for the entire sample of 234 subjects.
Description of cases involving recantation
Victim 1. This was a 14 year old. Hispanic female who was fondled and digitally, vaginally penetrated by a 57 year old perpetrator, her stepfather, who lived in the home. After a full disclosure of abuse, the victim recanted all allegations in two subsequent interviews, stating that she had made them up.
During the DPRS investigation the step father had been living in a hotel. Following the investigation and validation of the abuse allegations, however, DPRS allowed him to move back into the home. It was after the step father's return that the child recanted. The mother in this case repeatedly told DPRS that the stepfather's absence from the home was financially and emotionally destructive to the family. The case worker suspected that the mother encouraged the daughter to recant her allegations. No legal action was taken against the step father.
Victim 2. This was a 13 year old non-Hispanic White female who was fondled by a 38 year old perpetrator, her father, who lived in the home. The victim was interviewed by DPRS and gave a full disclosure of abuse in the DPRS office. She then stayed in a waiting room with her family while her sister was interviewed. After the sister's interview, the 13 year old victim recanted the allegations she had made shortly before. When asked why she had changed her story, she stated that her mother had told her to. Apparently the mother, frightened by possible repercussions to her husband, had pressured her daughter to recant. This victim reaffirmed the allegations within the same interview that she recanted. No legal action was taken against the perpetrator.
Victims 3 and 4. This case included two African American sisters, ages 10 and 13, who were fondled and digitally, anally penetrated by a 36 year old perpetrator, their stepfather, who lived in the home. Both victims were interviewed by DPRS and gave a full disclosure of sexual abuse. They subsequently repeated the allegations to police. The mother had made the initial report to DPRS. However, when the stepfather was forced to leave the home, she expressed great dissatisfaction with the arrangement. When military investigators later interviewed the victims, both recanted. It is possible that the victims may have been pressured to recant by the mother, or even by the military investigators. Allegations of abuse in the military require legal action, and the superior officers of an alleged perpetrator may sometimes be reluctant to pursue charges.
In this case, the mother repeatedly allowed the stepfather to have contact with the victims, contrary to DPRS mandates. The 10year old victim reaffirmed the abuse in the next DPRS interview following the recantation. The 13year old victim refused to discuss the topic further with DPRS. The perpetrator was eventually convicted in criminal court and sentenced to a military prison.
Victim 5. This was a 4 year old Hispanic female who stated that she had been fondled by an unidentified perpetrator. The child had been found to have a sexually transmitted disease (warts on the anus) and was referred to DPRS by her doctor. She initially denied abuse, then disclosed abuse, then recanted all allegations. No perpetrator was identified by DPRS. However, the victim did identify a teenage boy living across the street as the perpetrator. There had been a previous allegation of sexual abuse of an older sister in the same family, but the perpetrator in that incident was in jail at the time of the current investigation. The alleged teenage perpetrator and all males in the victim's home were tested for sexually transmitted diseases. All tests were negative. No legal action was pursued.
Victim 6. This was a 10 year old Hispanic female who was fondled and digitally vaginally penetrated by a 39 year old perpetrator, an extended family member not living in the home. This was a case of partial recantation. The victim did not deny her previous allegations, but claimed that she could not remember two of the incidents disclosed earlier. The victim reaffirmed the abuse in the next interview following the partial recantation. Although an arrest warrant was issued for the perpetrator, the case file contained no information regarding subsequent prosecution.
Victim 7. This was an 11 year old African American female who was fondled and genitally, orally penetrated by a 32 year old perpetrator, her step father who lived in the home. The victim made a full disclosure to DPRS and the police. One month later, she recanted all allegations of abuse. According to the victim, her mother had said that there was no medical evidence, and that she did not believe the allegations, which were ruining the step father's life. After reassurances from DPRS and the police, the victim reaffirmed her allegations during the same interview that she had recanted them. Although an arrest warrant was issued for the perpetrator, the case file contained no information regarding subsequent prosecution.
Victim 8. This was a 14 year old Hispanic female who was genitally, vaginally penetrated by a 31 year old perpetrator, her mother's boyfriend, who lived in the home. This case included a validated allegation of physical abuse and the victim was removed from the home and placed with her grandparents. In the first DPRS interview the victim stated that the perpetrator had put a gun to her head and raped her. Three months later DPRS was notified that the victim was pregnant. At this time the victim recanted her allegation that the perpetrator had used a gun, and stated that the sexual contact had been voluntary. The victim was removed from DPRS counselling by the grandmother and married the perpetrator. Because the mother gave permission for the marriage and did not want to press charges, DPRS reluctantly closed the case.
Six findings in the present study seem particularly notable. First, the present results indicate that the Child Sexual Abuse Accommodation Syndrome (CSAAS) described by Summit (1983) may be rather infrequent, at least among cases seen by child protection agencies (see also Jones & McGraw, 1987). In the present study, recantations, which are characteristic of CSAAS, occurred in only 3% of all cases, and in 4% of cases involving disclosure. Partly due to the influence of Summit (1983) himself, general awareness of child sexual abuse has increased in the years since he proposed CSAAS. As adults have become less skeptical, perhaps young victims feel less pressure to recant. Alternatively, perhaps CSAAS and recantations are characteristic only of a particular subset of victims, such as teenagers who have been abused repeatedly over many years by an immediate family member.
Second, although recantations were rare in the present study, four of the eight victims who recanted apparently did so in response to pressure from a caretaker. This finding is consistent with (a) Summit's (1983) view that "retraction" is often a response by the victim to pressure from adults: and (b) Lawson and Chaffin's (1992) finding that denial of abuse by the victim is more likely if the caretaker is unsupportive.
Third, in one case of recantation (Victim 6) the child reported that she could no longer remember some of her allegations. It is possible that this victim was in the avoidance phase of PTSD, as described by Koverola and Foy (1993), although other explanations of the memory lapse are also possible.
Fourth. denial of abuse occurred in 6% of validated cases. As this finding indicates, denial does not constitute definitive evidence that a child has not been abused. If a child initially denies abuse, additional interviewing may still be appropriate, especially if there is reliable independent evidence of abuse. In such circumstances a supportive, non-leading approach by the interviewer seems desirable.
Fifth, reluctance to disclose abuse was recorded in 10% of case files. However, a child's reluctance to describe abuse can be brief and expressed nonverbally, and thus may not be reflected in case notes. In all likelihood, the 10% figure for reluctance underestimates the frequency with which children hesitate or show discomfort before disclosing abuse.
Sixth, the present findings do not support the view that disclosure is a quasi-developmental process with "stages" that can be "resolved" (Sorenson & Snow, 199 1, p. 14). Stage theories of development, such as those of Kohlberg and Piaget, involve the assumption that children pass through an invariant sequence of stages, without skipping any of them (Loevinger, 1976). In the present study, however, most children (78%) entered the active disclosure stage at the first interview and never progressed to any other stage. Denials and recantations did not appear to follow any particular temporal or sequential pattern, as proposed by Sorenson and Snow (1991) and Summit (1983: but see Summit, 1992).
Seventh, whereas the findings of Sorenson and Snow (1991) indicate that denials and recantations are quite common among sexual abuse victims, the results of the present study suggest that these events may be rather unusual among victims seen by child protection agencies. We cannot say with certainty why the present results differ from those of Sorenson and Snow (1991). However, four reasons may be identified:
Subject characteristics. In some respects the subjects in the present study resembled those of Sorenson and Snow (1991). Specifically, both samples were of comparable age and composed primarily of intrafamilial abuse cases. However, the two samples differed markedly in ethnic background and religion. Most subjects in the present sample were Hispanic and in all likelihood Catholic (although religious background was not recorded in case files). By contrast, the subjects of Sorenson and Snow (1991) were predominantly non-Hispanic White and Mormon (B. Snow, personal communication. June 18, 1995). It is possible that the different disclosure patterns in the two studies reflect divergent cultural or religious responses to child sexual abuse allegations.
The present sample consisted only of validated sexual abuse cases. Although it is likely that some victims who denied abuse during their initial DPRS interview were "screened out" by the validation process, there are two reasons that this artefact probably does not account for the discrepancy between the present findings and those of Sorenson and Snow (1991).
First, it appears that the sample studied by Sorenson and Snow (1991) also consisted of validated sexual abuse cases and thus would have been subject to the same kind of "screening" effect described here. Second, denial of abuse by a child does not lead to "automatic" case closure by El Paso DPRS. The agency continues its investigations if there is other reliable evidence that abuse has occurred (e.g., medical evidence, disclosure by the victim to a reliable third party). In files that we reviewed, the DPRS caseworkers sometimes responded to an initial denial by scheduling additional interviews or arranging for the child to see a counselor.
It is also possible that a second selection artefact affected the present results. Specifically, only a subset of validated cases from 1992 and 1993 were examined in the present study. The remainder were unavailable, usually because the cases were still open for services. It is possible that the rate of denial. reluctance. and recantation was different for open and closed cases. However, a review of available files did not reveal a relationship between disclosure events and the length of time that cases were kept open.
The present study examined denials, reluctance, and recantations in the setting of child protection or police interviews. By contrast, Sorenson and Snow (1991) examined these disclosure events in the setting of psychotherapy. It is possible that the frequency of denials or recantations differs from one setting to another, or is affected by different social expectations or interviewer approaches in the two settings. For example, an abused child who is willing to discuss abuse during an initial DPRS interview may become reluctant to continue the discussion during multiple therapy sessions. Alternatively, it may be that over the course of therapy there are times when children recant briefly and then reaffirm.
It should be noted that Jones and McGraw (1987) examined recantations of sexual abuse in the setting of a child protection agency and reported a recantation rate of 8%, a figure close to the 34% in the present study. By contrast, Gonzalez and colleagues (1993) examined recantations in a therapy setting and reported a recantation rate of 27%, a figure not much higher than the 22% reported by Sorenson and Snow (1991). Perhaps the disclosure process differs from one context to another. This may be an issue worthy of further research.
On one hand, the DPRS case notes in the present study are problematic because they may have been written under time pressure or with an eye to legal considerations. Thus, they may have systematically underreported the denials, reluctance, or recantations of sexual abuse victims. We would not be surprised, for example, if case workers often failed to record instances in which a victim showed only brief reluctance or embarrassment about discussing abuse. On the other hand, a reliance on therapists' recollections as in the study by Sorenson and Snow (1991), is also problematic. As is generally recognized, retrospective reconstruction of facts can lead to errors and bias in scientific data.
In conclusion, we must confess that we are not sure why our results differ so strikingly from those of Sorenson and Snow (1991). Sorenson and Snow (1991) entitled their article, "How Children Tell." By contrast, the title of the present article poses a question: "How Do Children Tell?" In our opinion, this question cannot yet be answered with any certainty. When research findings conflict, as they do in the present case, the differences can be resolved only through additional research with careful methodology. Neither the present study nor that of Sorenson and Snow (1991) provides a definitive answer to the question of how children disclose sexual abuse.
The authors gratefully acknowledge the staff of the El Paso Department of Protective and Regulatory Services for their cooperation with this project, with special thanks to Eddie Wilson. Leonard Giron, lrene Perez, Connie Hewitt, Manuel Melendez, and Mariano Marinelarena. The authors wish to thank Judge Max Higgs, El Paso County Probate Court, for his ongoing sponsorship of the research, and Harmon Hosch, Martha Smithey, and Randy Whitworth for their comments on the first author's thesis. Our thanks to Barbara Snow for providing us with information regarding her own research, and to the anonymous reviewer of this article for offering the suggestions to explain the low rates of recantation, denial, and reluctance in our sample. We especially wish to express our appreciation to Lili Santoyo, Veronica de la 0, and Dawn Bayardo of Por Ninos y Familias for their assistance in scoring case files.
Bradley, A. R. (1995). How do children tell? The disclosure process in child sexual abuse. Unpublished master's thesis. University of Texas at EI Paso, EI Paso, TX.
Farrell, L. T. ( 1988). Factors that affect a victim's self-disclosure in father/daughter incest. Child Welfare. 5, 463 - 468.
Forties, L. A.(1993). Disclosures of sexual abuse by Puerto Rican children: Oppression and cultural barriers. Journal of Child Sexual Abuse. 2(1), 21-35.
Gonzalez, L. S., Waterman. J., Kelly, R. I., McCord, J. & Oliveri, M. K. (1993). Children's patterns of disclosures and recantations of sexual and ritualistic abuse allegations in psychotherapy. Child Abuse and Neglect, 17, 281-289.
Jones, D. P. H., & McGraw, J. M. (1987). Reliable and fictitious accounts of sexual abuse in children. Journal of Interpersonal Violence, 2. 2745.
Koverola. C., & Foy, D. (1993). Post traumatic stress disorder symptomatology in sexually abused children: Implications for legal proceedings. Journal of Child Sexual Abuse. 2(4), 119-128.
Lawson. L., & Chaffin, M. (1992). False negatives in sexual abuse disclosure interviews: Incidence and influence of caretaker's belief in abuse in cases of accidental abuse discovery of STD. Journal of Interpersonal Violence, 7(4), 532-542.
Loevinger, J. (1976). Ego development: Conceptions and theories. San Francisco. CA: Jossey Bass.
Myers, J. E. B. (1992). Legal issues in child abuse and neglect. Newbury Park. CA: Sage.
Rieser. M. (1991). Recantation in child sexual abuse cases. Child Welfare. 70(6), 611 - 621.
Salter, A. C. (1988). Treating child sex offenders and victims. Newbury Park. CA: Sage Publications.
Sauzier, M. (1989). Disclosure of child sexual abuse: For better or for worse. Psychiatric Clinics of North America, 12. 455-469.
Sorenson, T., & Snow, B. (1991). How children tell: The process of disclosure in child sexual abuse. Child Welfare, 70(1), 3 15.
Summit. R. C. (1983). The child sexual abuse accommodation syndrome. Child Abuse & Neglect. 7. 177-193.
Summit. R. C. 1992). Abuse of the Child Sexual Abuse Accommodation Syndrome. Journal of Child Sexual Abuse. 1. 153-163.
Scoring rules for each disclosure event during DPRS or Police interviews
DENDenial. Victim states that he/she has not been sexually abused. A denial that occurs after "DIS1st" should be coded as recantation (see below).
RELGEN Reluctance. Victim is reluctant or refuses to discuss the general topic of sexual abuse.
REL SPECIFIC Reluctance. Victim is reluctant or refuses to discuss certain details or topics that seem related to sexual abuse.
DISIst Disclosure: First time. Victim states that he/she has been sexually abused, or describes events that constitute sexual abuse. Scored only once per victim.
REPEAT Victim repeats allegations of sexual abuse or repeats prior details regarding abuse in a new interview.
(a) REPEAT is typically scored if the victim repeats in a "New Interview" what he/she said in a previous interview.
(b) REPEAT is NOT scored if the victim repeats the same information twice in the "Same Interview".
(c) REPEAT in NOT scored if victim restates an allegation that has been recanted. Restatements of recanted allegations are always coded as REAFF.
DIS/NEW Disclosure: New. All four of the following must be present to score this category:
(a) Victim describes events that constitute sexual abuse,
(b) The case file does not indicate that the victim has described these events to DPRS or police before.
(c) The child has made at least one disclosure before, and
(d) Either the new disclosure is made during a "new interview" or the victim has denied, recanted, or refused to discuss a prior disclosure.
RECALL Recants all sexual abuse. After having disclosed, the victim denies ever having been abused or denies all events that have been previously disclosed.
REC PART Recants only part of alleged sexual abuse incidents. After having disclosed that sexual abuse occurred, the victim denies that a particular incident or type of abuse occurred. but does not deny everything that has been said, or make general statement that he/she has never been abused.
REAFF Victim reaffirms an allegation that has previously been recanted. Must include (a) and (b):
(a) The victim already disclosed the information (i.e., it was scored as DIS1st or DIS/NEW)
(b) The victim later recanted the same information (i.e., it was scored as RECALL or REC/PART).
In addition, the following rules are relevant:
(c) REAFF is NOT typically scored more than once in the "Same interview." For example, if a victim reaffirms allegations twice in the same interview. REAFF is usually scored only once.
(d) REAFF may be scored more than once for the same interview only if the victim has recanted or been reluctant since the prior scoring of REAFF in the same interview. For example, if a child reaffirms, then recants, then reaffirms again, all in the same interview, the scoring would be REAFF, RECALL, REAFF for that interview.
(e) If a child reaffirms an allegation in one interview and then reaffirms the same allegation in a second interview, the first interview is scored as REAFF and the second interview is scored as REPEAT.
NONE None of 19. Score only for a new interview with child in which none of 19 applies.
1st Interview First interview is the first interview child has with DPRS or Police.
SAME INTERVIEW If an interview is not the first interview or a new interview, then it is the same interview. NEW INTERVIEW. An interview is considered new if it:
(a) involves a new interviewer,
OR< BR>(b) was done by the same interviewer. but not on the same day
(c) was done by the same interviewer on the same day, but the victim has been interviewed by another interviewer in the meantime.