Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 1
This article has been accepted for publication in The Journal of Child Sexual Abuse. Copyright notice: This article
may not exactly replicate the final published version. It is not the copy of record. Please do not duplicate or
distribute any part of this article without the permission of the author.
Disclosures of Sexual and Physical Abuse Across Repeated Interviews
Mitchell L. Eisen1 Gail S. Goodman2 Jessica Diep1
Marianne Lacsamana1Lauren Ristrom1 Jian Jian Qin 3
1California State University, Los Angeles
2University of California, Davis
3California State University, Sacramento
Acknowledgements: This work was supported by the National Center on Child Abuse and
Neglect. All correspondence reading this paper should be directed to Professor Mitchell Eisen,
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 2
This study examined the recorded interviews of 132 children between 3 and 16-years of age who
were involved in a forensic investigation evaluating allegations of sexual and/or physical abuse.
As part of this investigation, two interviews were conducted over a 5-day period. The interviews
were analyzed to examine how frequently these children disclosed substantiated allegations of
abuse when asked directly about these experiences in one or both interviews. Results revealed
that 39.2% of children with substantiated sexual abuse and 55.6% of those with substantiated
physical abuse denied these experiences in one or both interviews. The denial rate was highest
among school aged children (6- to-10-year- olds), as over a third of the girls and more than half
of the boys in this age group denied the substantiated allegations in one or both interviews.
Recantations were also relatively common, as 24% of the children who disclosed sexual and/or
physical abuse prior to the assessment denied the allegations in one or both interviews. The
youngest children (3- to 5-year-olds) were most likely to be inconsistent in their
denials/disclosures across interviews. Custody status and relation to the perpetrator also
predicted denials/disclosures of sexual, but not physical abuse. Overall, the data suggest that
denials of both sexual and physical abuse are quite common and higher than rates found in most
archival file reviews.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 3
Jane Doe was a 9-year-old Mexican-American girl who was admitted to a special hospital
unit in Chicago designed to investigate allegations of abuse. Jane had told her mother that her
father orally copulated her, and then she repeated this disclosure to a physician when she was
admitted to the unit. The case seemed straightforward enough; she had genital gonorrhea, her
father had oral gonorrhea, and her statements fit well with the physical evidence. The assessment
team at the hospital scheduled a forensic interview that was to be observed in real-time through a
one-way mirror by the prosecutor, investigating officer, and child services caseworker. Jane did
not speak English, and although the interviewer spoke fluent Spanish, she was not a Latina.
Rather, she was a light-skinned young woman of European descent who grew up in Costa Rica.
To everyone’s great surprise, when Jane was asked directly about the sexual abuse, she denied
that it had occurred, and when pressed about her earlier statements, Jane recanted the allegations
altogether. Immediately after leaving the interview room, Jane searched the area for her
Mexican-American social worker, who she had grown quite attached to; and when she found this
woman, Jane ran up to her, leaped into her arms, and told her in Spanish, “I didn’t want to talk to
that white lady.” A few days later, Jane was re-interviewed in a new context where the trusted
social worker was present, and in this safe environment, Jane gave a full report of the sexual
abuse that matched her earlier statements and was consistent with the physical evidence.
Unfortunately, her denial in the previous interview created a complicated record, which now
included a clear recantation of the allegation. This inconsistent record of disclosures and denials
scared off the district attorney from criminal prosecution and made it much more difficult for the
judge in the Dependency Court to sustain the allegation of sexual abuse. The defense argued that
Jane’s recantation called her claims of sexual abuse into question, and maintained that her
father’s gonorrhea could possibly have been transmitted to Jane innocently in the home through
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 4
mutual contact with a damp towel (The Case of Jane Doe, Mt Sinai Hospital Under the Rainbow
Denials and recantations can have a profound effect on the identification of child sexual
abuse. Previous research has shown that inconsistent child witnesses are considered to be less
believable than those who make consistent reports across interviews (Myers et al., 1999), and
that in general reluctant children give lower quality reports (Blasbalg, et al., 2018). In Jane Doe’s
case, the child recanted, because she was wary of speaking with someone who she considered to
be an outsider. In other cases, children may deny or recant true allegations of abuse for a wide
variety of reasons, including, but not limited to, fear of retaliation by the abuser, familial
pressures, fear of not being believed, experiences of embarrassment and shame, protecting the
abuser, limited understanding of why they are being interviewed, or simply not wanting to talk
about the incident with that interviewer at that time (see Lemaigre et al., 2017 for a recent
review). If the witness does not consistently report the allegation, this will ultimately result in a
checkered record of disclosures, followed by denials and retractions.
Of course, not all denials and recantations involve reluctant witnesses who have
experienced actual abuse but are unwilling to talk about it. Denials and recantations can also
occur in cases that involve false allegations driven by suggestive interviews or made at the
behest of another party who has coached the child to make baseless claims to further their own
personal aims (Bruck & Ceci, 2004; Lyon et al., 2008; O’Donohue et al., 2018). Some
investigators have asserted that a substantial portion of denials and recantations reported in
previous studies were likely due to the erroneous inclusion of false allegations, which they argue
are more likely to be recanted (Bruck & Ceci, 2004). These investigators have argued that, in
reality, denials and recantations are relatively uncommon, and that when children who have truly
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 5
been abused are asked directly, they will usually report it (London et al., 2005). Based on this
assumption, these theorists have argued that if child does not readily disclose abuse when asked
directly, the risks involved in pushing forward with more direct and potentially suggestive
questions far outweigh any potential benefits in getting reluctant victims to make statements that
they otherwise would not have provided (Bruck & Ceci, 2004).
In contrast, other researchers and practitioners who specialize in the forensic interviewing
of children, have argued that many children who are initially reluctant to disclose will in fact
benefit from some additional prompting with more direct questions (Goodman, 2006;
Hershkowitz & Terner, 2007), or even additional interviews (Katz & Hershkowitz et al., 2013);
and that the youngest children are most likely to benefit from more direction questioning
(Goodman, 2006). The notion that use of more direct questionings and/or additional interviews
might be needed for the youngest children has raised concerns among some investigators, since
this is precisely the group that is most vulnerable to suggestive influence (Bruck & Ceci, 2004).
The current study was designed to examine disclosures of abuse in two different interviews
conducted during a forensic assessment of possible child physical and/or sexual abuse. Video
tapes of the interviews were analyzed to examine how frequently these children and adolescents
disclosed substantiated allegations of abuse in one or both interviews when asked directly about
these experiences. Before describing the study in detail, we first review previous research
examining factors found to be related to disclosures of abuse.
Relationship to the Perpetrator: Denials and recantations
Previous research has shown that children are less likely to disclose intrafamilial sexual
abuse (Hershkowitz et al., 2005; Smith et al., 2000; see Azzopardi et al., 2017 for a recent
review). Malloy et al. (2007) proposed a filial dependency model to explain how children’s
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 6
denials and/or recantations of abuse are often influenced by their susceptibility to influence from
the adults whom they depend on. According to this model, the greater the children’s dependence
on their parental figures, the greater their fears of betraying their parents and/or being punished
for disclosing abuse. In the current study we examined disclosures of alleged abuse by biological
relatives versus non-related individuals.
There is also reason to believe that placement status at the time of the evaluation may be
an important factor to consider when understanding children’s willingness to disclose abuse.
Indeed, regardless of whether the alleged abuse was committed by a biological relative or not, if
the custodial parent is not supportive of the investigation, they could influence the child not to
cooperate with authorities (Malloy et al., 2016). In the current study, we were able to directly
examine denials of children who were in foster placement versus those who were in their
parents’ custody at the time of the assessment.
What Do We Know About Denials and Recantations from Case File Reviews?
Although there is widespread agreement that many children do not disclose sexual abuse
when it occurs, and often keep these experiences a secret into adulthood (Lamb & Edgar-Smith,
1994; London et al., 2005; Smith et al., 2000), there is still considerable debate over how
consistently children will disclose sexual abuse when asked directly about their experiences
during a forensic investigation. This issue has been addressed in several published studies that
involved reviews of case files at clinics that specialize in the assessment of allegations of sexual
abuse (e.g., Bradley & Wood, 1996; Gordon & Jaudes, 1996; Keary & Fitzpatrick, 1994; Lawson
& Chaffin, 1992). Rates of disclosures as assessed through these case file reviews vary greatly,
with some studies indicating that less than half of their participants disclosed known sexual
abuse, whereas others report that most children disclosed when asked directly.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 7
London et al. (2008) reported that although denials and recantations occurred in 36% of
cases overall, the rates varied widely across studies, ranging from as low as 4% to as high as
74%. These investigators point specifically to data reported by Bradley and Wood (1996), which
showed that only 4% of the sexual abuse cases that were substantiated by social services
involved documented recantations. Based on this, London et. al., (2005, 2008) argued that when
only methodologically sound studies were examined denials and recantations are very
uncommon. This position has gained some traction in the courts. Specifically, the New Jersey
Supreme Court recently ruled to limit expert testimony describing why children may deny sexual
abuse (New Jersey, 2017). Lyon (2020) points out that in making this ruling, the court relied
heavily on the arguments proposed London et al., (2005, 2008), asserting that denials and
recantations are quite rare, and that when asked directly most children will disclose.
Lyon et al., (2020) review of the literature indicated that denials and recantations are not at
all uncommon. For example, Hershkowitz et al., (2014) examined 400 children in which abuse
was suspected and that some form of corroboration was available. These investigators reported
that half of the children in their sample either denied the allegations. Similarly, Malloy et al.
(2007) reviewed 257 randomly selected cases of sexual abuse from case files of a dependency
court located in a major metropolitan area in the US. Twenty-three percent of the children
recanted to someone at some point in the investigation (during formal or informal interviews).
Moreover, Malloy and her colleagues noted that even when restricting the review to formal
interviews conducted by the authorities and members of the assessment team, 19.5% of the
sample recanted, which was nearly five times greater than the 4% recantation rate reported by
Bradley and Wood (1996). Indeed, wide discrepancies in the rates of denials and recantations
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 8
reported across these various file review studies raise concerns over the utility of these data in
informing our understanding of how often children will deny or recant true allegations of abuse.
The Limits of Case File Reviews.
Although archival case file reviews provide the most available source of data examining
denials and recantations, reliance on clinical and court records can sometimes be problematic.
Lyon (2002) observed that clinical reports are often crafted with a bias to help children obtain
needed services and to protect them from the potential dangers of being returned to the home of a
suspected perpetrator. Lyon argued that this bias can sometimes keep clinicians and
investigators from adequately documenting denials and recantations. Moreover, Lyon et al.,
(2020) point to the problem of disclosure substantiation bias; noting that allegations are most
likely to be substantiated when a disclosure is made; which can make it appear that denials rarely
occur in substantiated cases. In the present study, rather than relying on the post hoc examination
of second-hand reports written by clinical investigators and caseworkers, we were able to review
the children’s actual taped forensic interviews. Specifically, we examined children’s abuse
disclosures in two consecutive interviews conducted during a 5-day evaluation program designed
specifically to assess allegations of maltreatment.
The Current Study
The current study utilizes the same data set as used by Eisen et al. (2007). This study
involved 328 3- to 16-year-olds engaged in a forensic investigation of abuse and neglect. During
this evaluation, the investigators were able to record two interviews designed to assess
allegations of abuse. In addition, Eisen and his colleagues had the unique opportunity to assess
these children’s level of receptive language development. Language development is obviously an
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 9
important consideration when examining how children perform when being interviewed about
allegations of abuse, particularly for very young children (Eisen et al., 2007).
For the current study, access to this wealth of data allowed us to examine a variety of
questions that could not be assessed through second-hand archival reviews of case files. First and
foremost, we were interested in how consistently children and adolescents would disclose
substantiated allegations of abuse when asked directly about these experiences over two recorded
interviews conducted within a short time frame. Specifically, how often would these children and
adolescents deny substantiated abuse in one or both interviews when asked directly about their
Based on the research reviewed above, several hypotheses were set forth. It was predicted
that older (compared to younger) children would be less likely to deny substantiated sexual and
physical abuse. It was also predicted that males (compared to females) would be more likely to
deny sexual abuse but not necessarily physical abuse. Moreover, it was predicted that children in
custody of their parents (compared to custody of the state/court) would be more likely to
cooperate with the investigation and would therefore be less likely to deny substantiated sexual
and physical abuse. We also expected that children would be more likely to deny substantiated
sexual and physical abuse when the perpetrator is a parent/caretaker. Finally, we predicted that
children who had (versus had not) disclosed sexual and/or physical abuse before the assessment
would be more likely to consistently disclose and less likely to deny the substantiated abuse in
one or both interviews.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 10
This sample was drawn from 328 children and adolescents who participated in a previous
study conducted by Eisen et al. (2007). The data were originally collected between 1994 and
1996. The participants were all engaged in a forensic investigation related to allegations of
maltreatment conducted by the authorities. As part of this investigation, the youth were admitted
to a special evaluation center designed specifically to evaluate these allegations.
Of the 328 cases, 132 were included in the current study. Cases of neglect were not
considered in the current effort, because judgements regarding what constitutes neglect are not
always clearly defined, and in many instances, the charge of neglect relates to actions of the
parents that are outside of the child’s awareness. Also, non-abuse controls from the larger data
set were not included. In the end, we only included cases if we had adequate data available to be
absolutely confident that the child disclosed or denied substantiated physical and/or sexual abuse
in each of the interviews (e.g., the taped interviews were available for review, DCFS
documentation of substantiation was available). The final sample of 132 children and adolescents
(79 females) included 54 cases of substantiated sexual abuse only (44 females), 58 cases of
substantiated physical abuse only (23 females), and 20 cases that involved documented histories
of both sexual and physical abuse (12 females).
Most of the children came from low socioeconomic status (SES) families who lived on the
south side of Chicago. Seventy-two percent of the sample were African-Americans (n = 95),
15.9% (n = 21) were Hispanic, and 12.1% (n = 16) were White. The children ranged in age from
3- to 16-years of age (M = 8.53, SD = 2.03), and for the purposes of analyses, the sample was
stratified into three age groups: 3- to 5-year-old (n = 30, M = 4.02, SD = .79), 6- to 10-year-olds
(n = 68, M = 7.51, SD = 1.18), and 11- to 16-year-olds (n = 34, M = 12.32, SD = 1.07).
Other Relevant Information About the Sample and the Population It Was Drawn From.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 11
The sample in this experiment was drawn from a larger group of 328 children and
adolescents who took part in a 5-day assessment for allegations of maltreatment conducted
between 1994 and 1996. This larger sample has been described in previous publications (e.g.,
Chae et al., 2011; Eisen et al., 2007). That said, this is the first time that the hundreds of taped
interviews collected in this effort were examined to study denials of abuse across the two
interviews. All cases were initiated by a call to the DCFS hotline, and all children were admitted
to the hospital at the earliest stage of the DCFS investigation. The inpatient approach used at this
facility was designed to isolate children from outside influences during the assessment process
and to provide children a safe refuge from alleged family and community violence. Moreover,
the inpatient approach helped provide emergency short-term placements for children who had
been removed from their homes.
The Forensic Interviews
Each child took part in two forensic interviews during the 5-day period in which they were
asked directly about the allegations. Both interviews were designed to assess allegations of abuse
but had a different secondary focus (see below). Both interviews were videotaped and the two
interviews were never conducted by the same person.
Forensic Interview #1: The Psychological Consultation. A trained clinician questioned
the children about the alleged abuse. In addition to asking about the allegations, this clinician
also assessed the child's emotional and cognitive functioning. The interviewers always opened
with rapport building, reviewed ground rules, and informally assessed the child’s language
ability (e.g., knowledge of concepts like in and out, over and under). Interviewers all used the
funnel technique in which the child was first asked open questions about the suspected abuse to
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 12
see if he or she would provide a narrative report with minimal prompting. When this did not
result in a disclosure, more direct yes/no questions were asked. Although some children were
willing reporters and voluntarily disclosed the abuse with little prompting, all children who
denied abuse were directly asked about the allegation (e.g., Did anyone ever touch you on your
Forensic Interview #2:The Child Development Interview. Despite the name, this
interview did not involve any special assessment of the child’s development. Rather, this was a
typical forensic interview designed specifically to evaluate the allegations of abuse. These
interviews were often observed though a one-way window by prosecutors, local law
enforcement, and officials from children’s services. Like the psychological consultation, the
child development interviews always consisted of a rapport building phase, followed by an
assessment of the child’s ability to tell the difference between truth and lie, and the establishment
of ground rules. The funnel method was also used. In addition, body charts were always used,
and anatomical dolls were sometimes employed. Moreover, the officials observing the interview
had the opportunity to suggest additional questions to ask based on the child’s statements.
Each interview was coded as either a denial or a disclosure. Denials were only coded if
the child denied the abuse at all points during the interview. Thus, if a child denied the abuse one
or more times but then disclosed the abuse at the end, this was coded as a disclosure. Similarly, if
the child disclosed, but then attempted to retract the disclosure later in the same interview, this
was still coded as a disclosure. Disclosures varied greatly from substantial narrative descriptions
of abuse to simple acknowledgements. Inconsistency within the interviews was not coded.
Disclosure Rates in the Interviews
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 13
Disclosure rates in the two different interviews were quite comparable. In the child
development interview, 29.2% denied past sexual abuse, and in the psychological consultation,
35.5% denied it. For physical abuse, comparable denial rates were also evident.
Several months after data collection was completed, records from DCFS were obtained.
Only substantiated allegations were included in the analyses. This means that DCFS, in
conjunction with the dependency court decided that the allegations of abuse were founded. We
did not code for severity of the sexual or physical abuse. That said, there was great deal of
variability in allegations of both sexual abuse (from intercourse to fondling) and physical abuse
(from minor bruises to serious injuries).
Admission records were examined to determine if each child had reportedly disclosed the
alleged sexual and/or physical abuse to anyone before the evaluation. These prior disclosures
were either from informal interviews by non-officials or from interviews conducted by police or
social services. Admission records indicated that 76% of the children in the sexual abuse group
and 32.9% in the physical abuse group disclosed the allegations to someone prior to the
Receptive Language Ability
Receptive language ability was assessed by the Peabody Picture Vocabulary Test, Revised
(PPVT-R: Dunn & Dunn, 1981). The PPVT-R was administered to all children. The PPVT-R
correlates significantly with other indices of verbal intelligence (Newmark, 1989). PPVT-R
standard scores for receptive language ability were calculated for analyses.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 14
The first set of analyses examined denials of substantiated instances of sexual and/or
physical abuse. Denials were defined as denying past experiences of abuse when asked directly
about the allegations in one or both interviews. Thus, the denial group included both consistent
deniers (those who in both interviews denied past abuse), and inconsistent deniers/disclosers
(those who in one interview denied past abuse while disclosing it in the other interview). Table 1
shows that 39.2% of the children with substantiated sexual abuse and 44.4% of those with
substantiated physical abuse denied having these past experiences in one or both interviews.
A series of binary logistic regressions were conducted to examine predictors of
denials/disclosures of sexual and physical abuse. For each regression regarding sexual abuse, the
following predictors were included: age (3- to 5-year-olds, 6- to 10-year-olds, 11- to 16-year-
olds), gender (male, female), substantiated abuse (sexual only versus both sexual and physical),
custody status (DCFS custody versus parent custody), and PPVT-R receptive language standard
scores. Since age was a three-level variable, categorical comparisons for each age group against
the others were made using the 11- to 16-year-olds as a reference. Preliminary analyses revealed
that none of the predicators were significantly correlated, (rs < .22, ps > .05). In the first model,
all main effects were entered, and in subsequent models, interactions were examined. We first
present the analyses related to denials/disclosures of sexual abuse.
The first regression examined denials of sexual abuse expressed in one or both
interviews. The first model was significant X2 (5, n = 72) = 20.41, p = .002. As predicted, the
younger children were more likely, in one or both interviews, to deny past sexual abuse.
Specifically, the 3- to 5-year-olds were significantly more likely to deny the abuse than the 11- to
16-year-olds (B = 2.49, SE = .90, Wald’s χ2(1) = 7.67, p = .006, OR = 12.01, (95% CI [2.68,
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 15
69.73]). Custody status also predicted denials, as children in DCFS custody (48.9%) were more
likely to deny abuse than those in the custody of their parents (21.4%), B = 1.25, SE = .62,
Wald’s χ2(1) = 4.04, p = .04, OR = .29, (95% CI [.09, .97]). Also, those who had substantiated
experiences of both sexual and physical abuse (55%) were more likely to deny sexual abuse than
those in the sexual abuse only group (33.3%), B = 1.37, SE = .67, Wald’s χ2(1) = 4.23, p = .04,
OR = 3.94, (95% CI [1.07, 14.55]). However, gender did not predict disclosure of sexual abuse.
The second model was not a significant improvement, and no interactions emerged.
Consistent Denial of Sexual Abuse
The next logistic regression examined predictors of consistent denials of sexual abuse as
the dependent variable (denial of past sexual abuse in both interviews) using the same set of
predictors. The first model was significant X2 (6, n = 69) = 18.63, p = .005. As predicted, boys
(47.1%) were significantly more likely to consistently deny being sexually abused than girls
(17.0%), B = 1.63, SE = .73, Wald’s χ2(1) = 5.03, p = .03, OR = 5.11, (95% CI [1.23, 21.22]).
Custody status also predicted consistent denials, as children in DCFS custody (36.6%) were 7.9
times more likely to deny abuse than those in the custody of their parents (7.1%), B = 2.11, SE
= .92, Wald’s χ2(1) = 5.24, p = .02, OR = .12, (95% CI [.02, .74]). Surprisingly, when
considering, no significant age differences emerged. Also, no interactions were detected.
Inconsistent Denial of Sexual Abuse
The next logistic regression examined inconsistent denials of sexual abuse (in one
interview, denying past abuse, but in the other interview, disclosing it) using the same set of
predictors. The first model was significant X2 (6, n = 68) = 15.58, p = .02. Girls (47.1%) were
significantly more likely to be inconsistent in their denials than boys (17.0%), B = 2.72, SE =
1.33, Wald’s χ2(1) = 4.18, p = .04, OR = .07, (95% CI [.005, .90]). Age also predicted
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 16
inconsistent denials, as 3- to 5-year-olds were significantly more likely to be inconsistent in their
denials (31.58%) when compared to the 11- to 16-year-olds (5%), B = 3.10, SE = 1.35, Wald’s
χ2(1) = 5.20, p = .02, OR = 21.70, (95% CI [1.54, 306.02]). The children who had substantiated
experiences of both sexual and physical abuse were more likely to be inconsistent in their
disclosures (31.6%) than those who had only experienced sexual abuse (11.8%), B = 2.40, SE
= .98, Wald’s χ2(1) = 5.96, p = .02, OR = 10.99, (95% CI [1.60, 75.35]). No interactions
Consistent Disclosure of Sexual Abuse
The next logistic regression examined consistent disclosures of sexual abuse (disclosing
past sexual abuse in both interviews). The first model was significant X2 (6, n = 68) = 20.54, p
= .002. Age predicted consistent disclosures, as the 11- to 16-year-olds were significantly more
likely to consistently disclose sexual abuse (85%) when compared to the 3- to 5-year-olds
(36.8%), B = 2.35, SE = .91, Wald’s χ2(1) = 6.66, p = .01, OR = .10, (95% CI [.02, .58]). Also,
children who were in their parents’ custody at the time of the evaluation were more likely to
consistently disclose in both interviews (78.6%) than those who were in DCFS custody (46.3%),
B = 1.42, SE = .63, Wald’s χ2(1) = .5.03, p = .03, OR = 4.14, (95% CI [1.20, 14.34]). In addition,
those in the sexual abuse only group (64.7%) were more likely to consistently disclose in both
interviews than those in the physical and sexual abuse group (42.1%). B = 1.44, SE = .69, Wald’s
χ2(1) = 4.32, p = .04, OR = .24, (95% CI [.06, .92]). No interactions emerged.
Denial of Physical Abuse
The next set of logistic regression analyses used the same predictors to examine physical
abuse denials. In the first model, all main effects were entered, and in subsequent models,
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 17
interactions were examined. We first analyzed denial in one or both interviews. The first model
was not significant, X2 (6, n = 77) = 10.25, p = .11, and no significant interactions were found.
Consistent denial of physical abuse
The regression analyses were repeated with consistent denials of physical abuse as the
dependent variable (in both interviews, denial of past physical abuse) using the same set of
predictors. However, for this analysis, the two older age groups were collapsed to form a single
group. This was done because no one in the 11- to- 16-year-old group consistently denied the
physical abuse. The first predictor model was not significant, X2 (5, n = 52) = 1.18, p = .95, and
no significant interactions emerged.
Inconsistent Denials of Physical Abuse
The regression analyses were repeated with inconsistent denials of physical abuse as the
dependent variable (in one interview, denial of past physical abuse, whereas in the other
interview, reporting past physical abuse) with the same set of predictors as above, and again, the
first model was not significant, X2 (5, n = 51) = 4.39, p = .50, and there were no significant
Consistent Disclosure of Physical Abuse
For the final logistic regression analyses, which examined consistent disclosure of
physical abuse, the first model was not significant X2 (6, n = 66) = 8.73, p = .19. Although the 11-
to 16-year-olds (82.4%) were significantly more likely to consistently disclose physical abuse
than the 6- to 10-year-olds (72.2%), this difference did not reach statistical significance B = 1.57,
SE = .95, Wald’s χ2(1) = 2.71, p = .10, OR = 21, (95% CI [.03, 1.35]).
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 18
Prior Disclosures of Abuse
Seventy six percent of the children had disclosed sexual abuse before the assessment, and
44.3% had disclosed physical abuse before the evaluation. To examine the relation between prior
disclosure of sexual abuse and denials during the interviews, a 2 (Prior disclosure) x 2 (Denials
in one or both interviews) chi-square analyses was conducted. Results revealed that children who
disclosed sexual abuse before being admitted to the unit were significantly less likely to deny the
allegations in one or both interviews (91.1%) than those who had no record of prior disclosures
(8.9%), X2(1, 69) = 14.85, p < .001, Φ = .46. These analyses were repeated for denials of
physical abuse. Surprisingly, these analyses revealed no significant difference in denials among
children who disclosed physical abuse before being admitted to the unit (40%) and those who did
not (46.3%), X2(1, 79) = .28, p = .60, Φ = .06.
Relationship to the Perpetrator
To examine the relationship between disclosing sexual abuse in one or both interviews
and the perpetrator being a primary caretaker, a 2 (primary caretaker) x 2 (denials in one or both
interviews) chi-square test was conducted. Among the sexual abuse cases, 40% of the
perpetrators were primary caretakers. Results revealed no significant difference in denials
between cases when the perpetrator was a primary caretaker (i.e., biological parent, foster parent,
step-parent, or grandparent) and when the perpetrator was not X2(1, 65) = 2.02, p = .16, Φ = .18.
Rather than looking simply at primary caretakers, who are often step-parents and foster parents, a
new variable was formed to compare perpetrators who were biologically related to the child
(e.g., biological parents, siblings, grandparents) to perpetrators who were not biologically related
to the child (e.g., step-parents, foster parents, and partners of the biological parents). For sexual
abuse, 42.4% were biologically related. As expected, these analyses revealed that children were
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 19
significantly more likely to deny substantiated instances of sexual abuse when the perpetrators
were a biological parent, grandparent, or sibling (48.3%), compared to when there was no
biological relation (51.7%), X2(1, 74) = 2.86, p = .09, Φ = .39. Specifically, children were
equally likely to deny the allegations in one or both interviews when the biological relation was a
parent (50%), sibling (52.9%), or grandparent (50%).
For physical abuse, the perpetrator was always some type of caretaker, and 81.3% were
biologically related. Chi-square analyses were conducted to see if children were more likely to
deny the abuse when the perpetrator was a biological relative versus a non-relative caretaker.
These analyses revealed that children were no more likely to deny physical abuse when the
perpetrator was a biological relative, X2(1, 81) = .78, p = .38, Φ = .07.
This study was designed to examine how readily children and adolescents disclosed
substantiated instances of sexual and/or physical abuse when asked directly about these
allegations in two different interviews conducted over a 5-day period. In addition, we examined
relations between age, gender, substantiated abuse history (sexual, physical, both), custody status
(DCFS versus parental custody), and receptive language ability as predictors of denials (both
consistent and inconsistent) and disclosures. We also examined if prior disclosures and
relationship to the perpetrator were related to denials of physical and/or sexual abuse.
Rates of Denial
Results of this study revealed that more than a third of the sample denied experiencing
documented physical and sexual abuse when asked directly about the allegations in two separate
interviews conducted over a 5-day period. The 39.2% denial rate for allegations of sexual abuse
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 20
and 44% denial rate for physical abuse was generally in line with data reported by Hershkowitz
et al., 2014).
The majority of children in this study reportedly disclosed sexual abuse to someone in a
formal or informal interview prior to being admitted to the unit (76%). Notably, 24% of the
children who disclosed sexual abuse before the assessment denied the abuse in one or both
interviews. Essentially, they recanted the previous disclosure. This is nearly identical to the 23%
recantation rate reported by Malloy et al. (2007), and is consistent with data reported by
Hershkowitz et a., (2014) who found that 29% of children who disclosed abuse prior to being
assessed, denied the allegations when interviewed (i.e., recanted). The recantation rate observed
in this study and the others discussed above run directly counter to assertions made by some
investigators who have argued that recantations are rare when the allegations are true (London et
Denials of physical abuse looked very different. Surprisingly, when considering physical
abuse denials, children who disclosed before the assessment were not less likely to deny the
allegations in one or both interviews. That said, the denial rate for physical abuse disclosures was
much higher than with sexual abuse, as 40% of the children who made a documented disclosure
of physical abuse before the assessment denied the allegations when asked directly about it in
one or both interviews.
Some researchers have argued that repeated interviewing can be a useful tool to facilitate
disclosures, particularly for young children who may have been reluctant to disclose in the first
interview (Katz & Hershkowitz, 2013; La Rooy et al., 2009). Thus, it was of interest to see if
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 21
those children who were inconsistent in their disclosures across the two interviews were more
likely to make a disclosure in the second interview. The data from this study did not support this
proposition, as the second interview did not increase the prevalence of sexual or physical abuse
disclosure. Overall, 17% of the sample disclosed sexual abuse in one interview but then denied it
in the other. For the 12 children who had substantiated experiences of sexual abuse and who
disclosed in one interview and then denied in the other, 6 of them disclosed only in the first
interview and 6 only disclosed in the second. For the 16 children who were inconsistent in their
disclosures of physical abuse, 11 only disclosed in the first interview, while 5 only disclosed in
the second. Moreover, there was no difference in age for those who disclosed sexual or physical
abuse in the first versus second interviews.
Inconsistent Denials of Physical Abuse
Although the school aged boys were very consistent in their disclosures/denials of sexual
abuse across the two interviews, they were surprisingly inconsistent in their reporting of physical
abuse across interviews. When considering the 6- to- 10-year-old group, 35% of the boys
disclosed physical abuse in one interview and then denied it in the other. The relatively high
rates of inconsistent denials observed in the physical abuse cases may be related at least in part to
the less definitive nature of what constitutes a disclosure of physical compared to sexual abuse.
Although some forms of sexual abuse are difficult to define, for the most part, any genital touch
by an adult that is not related to bathing or daily care can be readily classified as sexual abuse. In
contrast, physical abuse is not always defined by the behavior of the adult, but rather is often
dependent on the results of the adult’s actions. For example, in many cases, the child may
disclose being hit with a belt, cord, or switch but will not acknowledge that this resulted in
marks, bruises, or other injuries. If the child acknowledged being hit, but denied injury, and no
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 22
marks, bruises, or scars that are directly attributable to the physical discipline can be detected in
the medical exam, then the allegation of physical abuse would most likely not be substantiated.
Relationship to the Perpetrator
As expected, children were more likely to deny substantiated instances of sexual abuse in
one or both interviews when the perpetrator was a biological relative. Interestingly, the denial
rate for biological relatives was very close to 50% regardless of whether the perpetrator was a
parent, a grandparent or a sibling. This finding is consistent with previous research which has
consistently shown that denial rates are higher when the perpetrator is alleged to be a biological
parent (Azzopardi et al., 2019; Malloy et al., 2007). Moreover, these data suggest that the
determining factor for relation to the perpetrator was not whether they were a primary caretaker
or not, but rather if they were a close biological relative. The 15.8% denial rate in cases when the
perpetrator was not a parent, grandparent or sibling is also telling and suggests that denials of
sexual abuse are far less common when the perpetrator is not a close family member. Taken
together, this pattern of results supports the argument raised by London et al. (2008), that studies
which include greater proportions of parent perpetrators are likely to have lower disclosure rates.
This is an area where physical abuse disclosures look very different than sexual abuse
disclosures. Previous research has found that children are not less likely to report physical abuse
by parent (Rush et al., 2014). Indeed, when considering physical abuse denials, there was no
difference between cases in which the perpetrator was a parent or grandparent versus a non-
relative caretaker. As noted earlier, most children are likely aware of the wrongfulness of sexual
activity with adults. Whereas, because physical abuse is usually related to some form of common
physical discipline, there is arguably less prohibition against discussing these acts with others.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 23
The denial rates for both sexual and physical abuse in this sample were lower for those
children who were in the custody of at least one biological parent at the time of the assessment.
Malloy et al. (2007) found that children in foster care were less likely to deny abuse after they
had previously disclosed, because they were less likely to be influenced by familial pressures to
not talk about the abuse. Indeed, there is no reason to doubt that unsupportive parents increase
the chances of denials and recantations. This apparent discrepancy between our data and
Malloy’s can be explained by differences in the methods used. Most notably, Malloy and her
colleagues examined case files spanning the life of the investigation, while we only examined
disclosures across two interviews conducted over a short time frame. Logically, continued access
to the child by family members who are unsupportive of the investigation would increase
opportunities for familial influence. However this would not be apparent in the current study.
Moreover, the mere fact that a child was admitted to the unit for assessment by a biological
parent who retains custody suggests the parent was supportive of the assessment. At minimum,
children’s services judged them to be more fit caretakers than those who had lost custody.
Age differences in denials.
Previous studies have found that disclosures of sexual abuse increase with age, until age
11 (Leach et al., 2017). In the current study, half of the children under 11-years of age denied
substantiated experiences of sexual abuse in one or both interviews. Although the 6- to 10-year-
olds were generally less likely to deny the sexual abuse than the preschoolers and more likely to
deny than the older group, after controlling for gender, language ability, and custody status, these
group differences were not statistically significant. As expected, the 11- to 16-year-olds were
significantly less likely to deny the sexual abuse than the 3- to 5-year-olds.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 24
The results of the analyses examining physical abuse denials looked quite different than
what was observed for sexual abuse. When considering physical abuse denials, the 6- to- 10-
year-olds showed nearly identical denial rates as the 3- to- 5-year-olds. As expected, both groups
were generally more likely to deny the abuse than their older counterparts, but the age
differences were not statistically significant. Notably, none of the 11- to 16-year-olds
consistently denied the substantiated allegations of physical abuse in both interviews, while 10%
of children in this age group consistently denied sexual abuse.
The age differences in disclosures of physical versus sexual abuse are interesting. At the
time of the assessment, most of the children were likely aware that sexual acts perpetrated by an
adult were deemed to be unacceptable by societal standards. Indeed, many of the adolescents
likely understood that this type of behavior could constitute a criminal act. By comparison, most
forms of physical abuse are related to relatively common disciplinary behaviors. It is likely that
the relatively normal actions of their adult caretakers that often constitute physical abuse were
easier to talk about because these behaviors were not overtly deviant or related to obvious
Gender differences in denials.
As expected, males were more than twice as likely as females to consistently deny
allegations of sexual abuse. Table 1 shows that this effect was driven primarily by the 6- to 10-
year-old males, who consistently denied these allegations in both interviews over 50% of the
time. The current data suggest that sexual abuse may be most difficult for school-aged boys to
acknowledge, perhaps because homosexual experiences may be more embarrassing to share with
others (Gagnier et al., 2016). This explanation would account for why gender differences were
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 25
not observed in the youngest group who were less likely to have been socialized to be
embarrassed about homosexuality, or more generally about societal sexual taboos.
Despite girls’ greater willingness to discuss substantiated experiences of sexual abuse when
asked directly about the allegations, 17% of girls still consistently denied the substantiated
claims of sexual abuse in both interviews. Moreover, when considering denials more broadly by
examining denials in one or both interviews, 36.8% of females denied the sexual abuse when
asked directly about the experience in at least one interview. The denial rate for females was
most concerning in the youngest age group, where more than half of the girls denied the
allegations in one or both interviews. In contrast, only 1 of 18 girls in the 11- to 16-year-old age
group consistently denied the substantiated sexual abuse in both interviews.
When considering physical abuse denials, the same trends emerged, as boys were generally
more likely to deny abuse than girls. Table 1 shows that the most pronounced differences were in
the 6- to 10-year-old range, as 63% of the 6- to 10-year-old males denied physical abuse in one
or both interviews, compared to only 35% of the girls in this age group.
It is important to note that the majority of the participants in this study were African
American children from impoverished communities on the Southwest side of Chicago. As noted
earlier, these children came from communities who are generally distrustful of the authorities.
Indeed, this may have affected their willingness to cooperate with the authorities. It is also
important to note that the data from this study were collected between 1994 and 1997. Although
the interviewers at this facility kept up to date on new trends in research and practice related to
interviewing children, they were not using well-articulated versions of the then nascent NICHD
protocol. Indeed, after being prompted with open questions, if children did not volunteer full
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 26
disclosures, all of the interviewers at this facility asked about the allegations using direct yes-no
questions that could be considered leading. Also, the forensic interviewers used body charts, and
in some cases, anatomical dolls. Of course, since the major concern with using these tools is
eliciting more false disclosures, the denial rates observed were not likely related to the use of
these potentially suggestive tools and strategies. Also, although all the cases examined involved
substantiated allegations, corroborating physical evidence was not available in all instances; and
in the absence of such evidence, ground truth is often unknown. Indeed, this is a limitation of this
study and most all field studies and archival case reviews.
These data provide a snap shot into what can be expected when children are interviewed
about substantiated instances of abuse in two different interviews conducted during a forensic
investigation over a short time frame. Moreover, these findings illustrate some important
differences in sexual abuse versus physical abuse disclosures that had not been previously
examined. The fact that the second interview did not increase disclosures, even in young children
who initially denied allegations, demonstrates the limited utility of interviewing children a
second time over a short time frame when denials occur. Also, the 24% recantation rate among in
substantiated sexual abuse cases is particularly notable, as it closely matches the 23% recantation
rate found by Malloy et al., (2007) and is consistent with the 29% denial of prior disclosures
reported by Hershkowitz et al., (2014). These data run directly counter to the notion that
children rarely deny or recant sexual abuse, and highlight the challenges faced by the child
welfare and criminal justice systems in differentiating between recantations driven by the child’s
reluctance to disclose previously reported abuse and denials of prior false-disclosures which may
have been the product suggestion.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 27
Azzopardi, C., Eirich, R., Rash, C. L., MacDonald, S., & Madigan, S. (2019). A meta-analysis of
the prevalence of child sexual abuse disclosure in forensic settings. Child abuse and Neglect.
93, 291-304. https://doi.org/10.1016/j.chiabu.2018.11.020
Blasbalg, U., Hershkowitz, I., & Karni-Visel, Y. (2018). Support, reluctance, and production in
child abuse investigative interviews. Psychology Public Policy and Law, 24 (4), 518–527.
Bradley, A. R., & Wood, J. M. (1996). How do children tell? The disclosure process in child
sexual abuse. Child Abuse & Neglect, 20(9), 881-891. https://doi.org/10.1016/0145-
Bruck, M., & Ceci, S. (2004). Forensic developmental psychology: Unveiling four common
misconceptions. Current Directions in Psychological Science, 13(6), 229-232. https://doi.org/
Chae, Y., Goodman, G. S., Eisen, M. L., & Qin, J. (2011). Event memory and suggestibility in
abused and neglected children: Trauma-related psychopathology and cognitive functioning.
Journal of Experimental Child Psychology, 110(4), 520–538.
Dunn, L. M., & Dunn, L. M. (1981). Peabody Picture Vocabulary Test-Revised. Manual for
Forms L and M. Circle Pines, MN: American.
Eisen, M. L., Goodman, G. S., Qin, J., Davis, S., & Crayton, J. (2007). Maltreated children's
memory: Accuracy, suggestibility, and psychopathology. Developmental Psychology, 43(6),
Gagnier, C., & Collin-Vézina, D. (2016). The disclosure experiences of male child sexual abuse
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 28
survivors. Journal of Child Sexual Abuse, 25(2), 221 241.
Gordon, S., & Jaudes, P. K. (1996). Sexual abuse evaluations in the emergency department: Is
the history reliable? Child Abuse & Neglect, 20(4), 315-322. https://doi.org/10.1016/0145-
Hershkowitz, I., Horowitz, D., & Lamb, M. E. (2005). Trends in children's disclosure of
abuse in Israel: A national study. Child Abuse & Neglect, 29(11), 1203-1214.
Hershkowitz, I., & Terner, A. (2007). The effects of repeated interviewing on children’s
forensic statements of sexual abuse. Applied Cognitive Psychology, 21(9),
Katz, C., & Hershkowitz, I. (2013). Repeated interviews with children who are the alleged
victims of sexual abuse. Research on Social Work Practice. 23(3), 210-218.
Keary, K., & Fitzpatrick, C. (1994). Children’s disclosure of sexual abuse during formal
investigation. Child Abuse & Neglect, 18(7), 543-548. https://doi.org/10.1016/0145-
La Rooy, D., Lamb, M. E., & Pipe, M.-E. (2009). Repeated Interviewing: A critical evaluation of
the risks and potential benefits. In K. Kuehnle & M. Connell (Eds.), The evaluation of child
sexual abuse allegations: A comprehensive guide to assessment and testimony (p. 327-361).
John Wiley and Sons Inc.
Lamb, S., & Edgar-Smith, S. (1994). Aspects of disclosure: Mediators of outcome of childhood
sexual abuse. Journal of Interpersonal Violence, 9(3), 307-326.
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 29
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
by diagnosis of STD. Journal of Interpersonal Violence, 7(4), 532–542.
Leach, C., Powell, M., Sharman, S., & Anglim, J. (2017). The relationship between children’s
age and disclosures of sexual abuse during forensic interviews. Child Maltreatment, 22(1), 79-
Lemaigre, C.. Taylor, E. P. Gittoes, C. (2017). Barriers and facilitators to disclosing sexual abuse
in childhood and adolescence: A systematic review. Child Abuse & Neglect, 70, 39-52,
London, K., Bruck, M., Ceci, S. J., & Shuman, D. W. (2005). Disclosure of child sexual
abuse: What does the research tell us about the ways that children tell? Psychology, Public
Policy, & Law, 11(1), 194-226. http://dx.doi.org/10.1037/1076-89220.127.116.11
London, K., Bruck, M., Wright, D. B., & Ceci, S.J. (2008). Review of the contemporary
literature on how children report sexual abuse to others: Findings, methodological issues, and
implications for forensic interviewers. Memory, 16(1), 29-47.
Lyon, T. D. (2007). False denials: Overcoming methodological biases in abuse disclosure
research. In M.-E. Pipe, M. E. Lamb, Y. Orbach, & A.-C. Cederborg (Eds.), Child sexual
abuse: Disclosure, delay, and denial (p. 41-62). Lawrence Erlbaum Associates Publishers.
Lyon, T. D., Malloy, L. C., Quas, J. A., & Talwar, V. A. (2008). Coaching, truth induction,
and young maltreated children’s false allegations and false denials. Child Development, 79(4),
Running Head: DISCLOSURES OF SEXUAL AND PHYSICAL ABUSE 30
Lyon, T., Malloy, L. C., Lyon, T. D., & Quas, J. A. (2007). Filial dependency and recantation of
child sexual abuse allegations. Journal of the American Academy of Child & Adolescent
Psychiatry, 46(2), 162-170. https://doi.org/10.1097/01.chi.0000246067.77953.f7
Lyon, T., Williams, S., Stolzenberg, S. N. (2020). Understanding expert testimony on child
sexual abuse denial. After New Jersey v. J.L.G.: Ground truth, disclosure suspicion, and
disclosure substantiation bias. Behavioral Sciences and the Law, 38, 630-647. DOI:
Malloy, L. C., Mugno, A. P., Rivard, J. R., Lyon, T. D., & Quas, J. A. (2016). Familial
influences on recantation in substantiated child sexual abuse cases. Child Maltreatment, 21,
256–261. DOI: 10.1177/1077559516650936
Myers, J. E. B., Redlich, A. D., Goodman, G. S., Prizmich, L. P., & Imwinkelried, E. (1999).
Jurors’ perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy and
Law, 5(2), 388-419. http://dx.doi.org/10.1037/1076-8918.104.22.1688
New Jersey v. J.R. (2017). 152 A.3d 180 (N.J.).
O’Donohue, W., Cummings, C., & Willis, B. (2018). The frequency of false allegations of child
sexual abuse: A critical review. Journal of Child Sexual Abuse, 27(5), 459-475.
Rush, E. B., Lyon, T., Ahern, E., & Quas, J. A. (2014). Disclosure suspicion bias and abuse
disclosure: Comparison between sexual and Physical abuse. Child Maltreatment, 19(2), 113-
Smith, D. W., Letourneau, E. J., Saunders, B. E., Kilpatrick, D. G., Resnick, H. S., & Best, C. L.
(2000). Delay in disclosure of childhood rape: Results from a national survey. Child Abuse &
Neglect, 24(2), 273-287. https://doi.org/10.1016/S0145-2134(99)00130-1