Abuse Characteristics and Individual Differences Related to Disclosing Childhood Sexual,Physical,and Emotional Abuse and Witnessed Domestic Violence

Abstract

Many adult survivors of childhood abuse hide their victimization, avoiding disclosure that could identify perpetrators, end the abuse, and bring help to the victim. We surveyed 1,679 women undergraduates to understand disclosure of childhood sexual, physical, and emotional abuse, and, for the first time, witnessed domestic violence, which many consider to be emotionally abusive. A substantial minority of victims failed to ever disclose their sexual abuse (23%), physical abuse (34%), emotional abuse (20%), and witnessed domestic violence (29%). Overall, abuse-specific factors were better predictors of disclosure than individual-level characteristics. Disclosure of sexual abuse was related to experiencing more frequent abuse (by the same and by multiple perpetrators), being more worried about injury and more upset at the time of the abuse, and self-labeling as a victim of abuse. Disclosure of physical abuse was related to experiencing more frequent abuse (by the same and multiple perpetrators), being less emotionally close to the perpetrator, being older when the abuse ended, being more worried and upset, and self-labeling as a victim. Disclosure of emotional abuse was associated with being older when the abuse ended, and being more worried and upset. Disclosure was unrelated to victim demographic characteristics or defensive reactions (dissociative proneness, fantasy proneness, repressive coping style, and temporary forgetting), except that among physical and emotional abuse victims, repressors were less likely to disclose than non-repressors. Disclosure of witnessing domestic violence was not significantly related to any factors measured.

Keywords

disclosure child sexual abuse child physical abuse child emotional abuse witnessing domestic violence

The societal goal of identifying and preventing child maltreatment is thwarted when children fail to disclose abuse. Although much attention has focused on false reports of childhood abuse, the problem of non-disclosure of abuse has received relatively little empirical attention (Pipe, Lamb, Orbach, & Cederborg, 2007). We examine the prevalence and correlates of disclosure of childhood sexual and physical abuse as well as the less-studied psychological traumas of experiencing emotional abuse and witnessing domestic violence. As such, the present article is largely and purposefully descriptive, aiming to identify the role of abuse characteristics, as well as victim individual differences (i.e., demographics and defensive reactions), in the disclosure of each abuse type.

Estimates of child sexual abuse disclosure vary widely, from 16% to 96% with a modal rate of 33% (for reviews, see London, Bruck Ceci, & Shuman, 2007; Lyon, 2007, 2009), due to methodological differences among studies (e.g., definition of sexual assault, immediate vs. delayed disclosure, childhood vs. adult disclosure, etc.). The few studies of child physical abuse disclosure reveal that 27% to 39% of survivors fail to disclose (Foynes, Freyd, & DePrince, 2009; Hershkowitz, Horowitz, & Lamb, 2005; Mazza, Dennerstein, & Ryan, 1996), and the one study of child emotional abuse disclosure that we know of reported that 58% of victims never disclosed (Foynes et al., 2009).

We have found no research about the disclosure of witnessing domestic violence, an understudied but prevalent (25% to 36%, Bureau of Justice Statistics, 2009; McCloskey & Walker, 2000; see Osofsky, 2003, for a review) childhood experience associated with negative outcomes similar to those of emotional and physical abuse (Burman & Duffy-Feins, 2007; Maker, Kemmelmeier, & Peterson, 1998; Teicher, Samson, Polcari, & Greenery, 2006). A meta-analysis revealed that children who witnessed domestic violence had worse psychological outcomes than children who did not (Kitzmann, Gaylord, Holt, & Kenny, 2003). Because of its prevalence, harmful effects, and the lack of information about its disclosure, we included witnessing domestic violence in our exploration of child abuse disclosure patterns, to inform and encourage further research on the topic.

Next, we discuss the factors we investigated, including prevalence and characteristics of disclosure (targets and outcomes), as well as predictors of disclosure: abuse characteristics (age, abuse frequency and emotional impact, and perpetrator identity) and individual differences in victims’ tendency to self-label as victims and their defensive emotion-regulation reactions. To preview, for the disclosure of sexual and physical abuse, our method allowed us to study all these factors; for the disclosure of emotional abuse and witnessed domestic violence, we examined all except identity of and emotional closeness to perpetrator, abuse frequency, and victim self-labeling.

Characteristics of Disclosure: Targets of Disclosure and Outcomes

Children tend to disclose sexual abuse to peers, parents, and other trusted adults (Hanson, Resnick, Saunders, Kilpatrick, & Best, 1999; Kogan, 2004; Lamb & Edgar-Smith, 1994), but not to authority figures who are likely to intervene (only 12% of all disclosures, Hanson et al., 1999; Smith et al., 2000). Therefore, even when child abuse is disclosed, it is unlikely to lead to an official investigation, nor, perhaps, to the end of the abuse. We found no studies of disclosure targets for the other forms of abuse, but it is reasonable to suspect the patterns are similar. In fact, because physical and emotional abuse are often considered either acceptable (i.e., corporal punishment) and/or not very damaging, they are probably even less likely than sexual abuse to be reported to legal or other authorities, and in turn, less likely to elicit formal interventions. Therefore, we predicted that, for all forms of abuse, victims would be more likely to disclose to family and friends than to authority figures, and that disclosure would rarely lead to investigations and end the abuse, especially for abuse forms that are perceived as more acceptable (e.g., corporal punishment), not serious (emotional abuse), or not directed at the child (domestic violence).

Predictors of Disclosure

We predicted that various situational and individual difference factors would foster or hinder abuse disclosure. Next, we discuss each, relying primarily on existing literature on sexual abuse disclosure.

Abuse Characteristics: Victim Age, Abuse Frequency and Emotional Impact, and Perpetrator Identity

Compared with younger children, older children are less likely to delay disclosing sexual abuse (Smith et al., 2000) and have higher rates of abuse disclosure (Goodman et al., 2003), perhaps because they are better able to understand the abuse experience and are, therefore, less likely to forget it, and because they have more access to people outside their immediate family, such as peers or teachers (London, Bruck, Ceci, & Shuman, 2005). Such access might be particularly instrumental when the perpetrator is a family member. In fact, London and colleagues point out that studies of disclosure during adolescence (when most victims disclosed to friends) generally uncover higher disclosure rates than studies of childhood disclosure (when most victims disclose to family members). Thus, for all types of abuse, we expected to find more disclosure from victims who were older, rather than younger, at the time of the abuse.

Abuse severity in general, and the frequency of abuse in particular, can affect disclosure, but the exact effect is not clear from existing studies (e.g., Hanson et al., 1999; Smith et al., 2000). On one hand, frequent abuse might mean the perpetrator has increased access to and control of the child as well as a closer emotional connection (e.g., a family member or other trusted adult), both likely to deter or delay disclosure (Smith et al., 2000). Arata (1998) suggested that the longer children avoid disclosure, the more they become concerned about getting blamed for not disclosing, further discouraging disclosure. This might be exacerbated in situations involving frequent (vs. infrequent) abuse. On the other hand, theoretically, frequent abuse might increase the child’s motivation to end the experience, increase the likelihood of remembering the abuse well (Goodman et al., 2003), or create opportunities for others (i.e., family members, peers, teachers, or doctors) to notice physical or emotional sequelae and to prompt the child to disclose. Given the latter, we predicted that childhood abuse that occurs with greater (vs. lesser) frequency would be more likely to result in disclosure.

It is reasonable to expect children’s emotional reactions to abuse, such as worrying about being injured or more generally feeling upset, to affect disclosure, yet the existing literature is mixed on the direction of such effects. Lamb and Edgar-Smith (1994) and Smith et al. (2000) found no significant relation. Paine and Hansen (2002) found that perpetrators’ use of tactics that made victims fear for their or others’ lives deterred abuse disclosure, yet Hanson and colleagues (1999) found that perceptions of life threat and fear of physical injury were related to higher disclosure rates. Most studies have not distinguished among sources of worry; for example, did perpetrators threaten children to secure their compliance, or to ensure their silence after the assault? It is possible that emotions associated with threats to secure compliance increase disclosure, as children seek help and comfort after abuse, while threats aimed at silencing the child after abuse decrease disclosure (London et al., 2005). In summary, on one hand, the more fearful victims are, the more likely they might be to disclose and thereby seek help and protection. On the other hand, the more fearful they are about potential retribution from the perpetrator, the less likely they might be to disclose. A likely scenario is that threats and fear might deter victims from disclosing in the short run, but over time motivate victims to disclose. We predicted that children who were more worried at the time of the abuse were more likely to disclose the abuse sometime in their lives.

Perpetrator relationship with the child might also affect the likelihood of disclosure. When the perpetrator is a close family member, children might be less motivated to disclose and have less opportunity to do so (for review, see Lyon, 2009). Close family members are in a better position to control the child, and children might be afraid to disrupt family relationships and get loved ones in trouble by disclosing (Bottoms, Goodman, Schwartz-Kenney, & Thomas, 2002; Kogan, 2004). In support, some have found that the likelihood of disclosure of sexual abuse increases when perpetrated by a stranger or non-familiar perpetrator rather than by a family member or another emotionally close person (Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003; Hanson et al., 1999; Kogan, 2004; Smith et al., 2000), yet others have not (Arata, 1998; Lamb & Edgar-Smith, 1994; Ullman, 2007). In the only work we know of to address this question regarding physical and emotional abuse, Foynes and colleagues (2009) found that a closer relationship between the victim and perpetrator (i.e., parents, step-parents, or siblings) led to less disclosure and more delay in disclosure. Given the somewhat stronger theoretical rationale underlying a prediction of less disclosure associated with closer perpetrators, and the bulk of the findings to date, we expected that victims of all forms of abuse would be less likely to disclose when perpetrators were family members and were emotionally close than when not.

Individual Differences: Victim Self-Labeling and Defensive Coping Mechanisms

Some individuals fail to self-label as victims even though their experiences match legal and theoretical definitions of abuse (Martin, Anderson, Romans, Mullen, & O’Shea, 1993). Self-labeling is impeded when abuse occurs within a family setting because perpetrators explain away the abuse, or because abuse is a constant occurrence and thus not perceived as unusual (Browne, 1991). Victims of emotional abuse and witnessing domestic violence might be especially unlikely to self-label as victims, because of the societal and legal failure to define these experiences as abusive. This issue has been seldom discussed in the disclosure literature nor has it been empirically investigated to our knowledge, even though intuitively it should predict disclosure. We included objective and subjective measures of victimization, expecting that victims who do not self-label would be less likely than others to disclose their experiences. A failure to self-label might also reflect a victim’s attempt to distance herself defensively from the experience, which we also measured in the following ways.

We investigated the relation between disclosing abuse and victims’ defensive emotion-regulation reactions, that is, their attempts to avoid acknowledgment of painful traumatic experiences and efforts to evade affect, memories, and retrieval cues associated with traumas (Briere, 2002). We reasoned that evading thoughts about trauma would lead to avoiding disclosure and discussion of trauma, and that people who are prone to using avoidant coping would be less likely than others to disclose their abuse. Our measures included extent to which victims used three types of avoidant coping: (a) repressive coping, defined by Weinberger (1990) as the tendency to avoid acknowledging experiences of negative affect and to avoid thoughts, information, and memories that are negative and stressful (see also Bonanno & Singer, 1990); (b) dissociative coping, which is a lack of integration of thoughts, feelings, and experiences (Bernstein & Putnam, 1986), and disturbances in memory, awareness, and identity (Nemiah, 1980); and (c) fantasy proneness, which is a tendency to be deeply involved in fantasy and imagination, and which has been linked to avoidance of trauma memories, which are, for the highly fantasy prone, like vividly re-experiencing the traumatic events (Wilson & Barber, 1983). We also measured victims’ experience of temporarily forgetting their abuse.

In support, Bonanno, Noll, Putnam, O’Neill, and Trickett (2003) found that victims who used repressive coping were less likely to disclose sexual abuse, which they explained as due to repressors’ active (yet not necessarily conscious) attempts to avoid the negative thoughts associated with the trauma, reflected in their reduced cognitive effort when instructed to retrieve personal negative memories. Bonanno et al. also found that chronically dissociative participants were more likely to disclose, but Goodman et al. (2003) found the opposite: Dissociative (vs. non-dissociative) participants were less likely to disclose child sexual abuse. These differences could be explained by the somewhat paradoxical effects of dissociation on memory: Although dissociative tendencies impair memory, they can also result in heightened recall of traumatic events through obtrusive memories (Bower & Sivers, 1998), and thus encourage disclosure. We know of no research addressing the relation between repressive and dissociative coping styles and disclosure of other forms of abuse, nor between fantasy proneness and the disclosure of any form of abuse.

Finally, we expected that non-disclosure would be related to temporary periods of forgetting the abuse, which is reported by a substantial minority of abuse victims (e.g., Epstein & Bottoms, 1998, 2002). Victims might experience temporary periods of forgetting because memories for traumatic events might operate through different processes than memories for non-traumatic events, with forgetting—even temporary—serving a defensive, protective function (e.g., Goodman et al., 2003; Williams, 1994). Alternatively, temporary forgetting is often reported by victims of sexual abuse, physical abuse, and other traumas for other reasons that are still potentially defensive reactions to painful memories: because they experienced a time during which they labeled the event as non-abusive or intentionally directed their thinking away from the event (Epstein & Bottoms, 2002). Thus, we expected that victims who reported temporary forgetting (for any reason) would be less likely than others to disclose, because during the time of forgetting they could not have discussed it, and because it is reasonable to assume that victims who are motivated (consciously or not) to avoid memories via temporary forgetting would probably also be motivated to avoid discussing the event with others.

Summary of Hypotheses

In summary, based on prior research and theory, we hypothesized that victims would be more likely to disclose to family and friends than to persons perceived as having official authority. We also predicted that disclosure rates would be higher for victims who (a) were older (vs. younger) at the time of abuse, (b) suffered more frequent instances of abuse, (c) were more worried about injuries at the time of the abuse, (d) were abused by someone other than a family member and were less emotionally close to the perpetrator, (e) self-labeled as abuse victims, and (f) did not use defensive mechanisms to avoid memories of the abuse, especially repressive coping style and temporary forgetting of the experience. We had no specific predictions about the effects of dissociative and fantasy proneness, due to the conflicting data and theoretical implications of prior studies. Finally, even though there has been little to no similar prior research with forms of abuse other than sexual abuse, especially emotional abuse and the witnessing of domestic violence, we expected similar patterns to hold across all four types of events for variables measured, given that all are potentially traumatic childhood experiences.

Method

Participants

Participants were 1,679 college women who completed our measures as part of a larger anonymous survey in return for course credit. They were diverse in terms of where they attended school (81% in Illinois, 13% in California, and 6% in Virginia, N = 1,679), race/ethnicity (21% African American, 22% Asian American, 36% White, 15% Hispanic, and 6% Other), and socioeconomic background as estimated by parental income, ranging from 15% of participants’ parents earning less than US$20,000 per year to 17.5% earning US$80,000 or more. Participants were between 17 and 60 years of age, M = 21, SD = 4.83. One participant who was 17 had parental consent for participation.

Measures

Women completed an anonymous survey with the three following sections:

Demographics

Women indicated their age, race/ethnicity, and parental income. Our measure of parental income was used as an approximation of family’s socioeconomic status, because it was not possible to assess parental income during the participants’ childhood objectively.

Abusive experiences

Three separate subsections of the survey measured abuse. The first measured sexual abuse, using a definition modeled after others’ (e.g., Finkelhor, 1979) based on behaviors experienced and ages of victim and perpetrator (matching laws in Illinois and California). The measure reads, “When you were 17 years old or younger, did you ever have any of the following experiences where someone at least 5 years older than you . . . (Note: this could mean that you did these things to someone or someone did them to you)?” Respondents chose one or more of the items in Table 1. The Kuder–Richardson 20 reliability coefficient for dichotomous response options was .73, indicating the inter-relatedness of the characteristics. Such an objective checklist measure is more accurate than subjective measures requiring women to self-label as victims of abuse without specifying experiences. (As discussed below, we also used self-labeling as a disclosure predictor.)

Table 1.

Percentage of Abuse Victims Who Chose Each of the Specific Forms of Abuse Listed in the Abuse Screening Questions.

% of Victims (% of Victims Who Answered Disclosure Questions)	Item From Abuse Screening Question
Sexual abuse, n = 467 (414)
6 (6)	Viewed or took part in child pornography
30 (32)	Exhibitionism (inappropriately exposed to adult’s genitals)
76 (77)	Fondling (touching) genitals, breasts, or buttocks directly or through clothing
22 (21)	Oral sex (mouth/genital contact)
6 (6)	Anal sex (penetration of anus with genitals, fingers, or other object)
29 (29)	Attempted vaginal intercourse (with penis, fingers, or other object)
30 (26)	Completed vaginal intercourse (penetration with penis, fingers, or object)
14 (15)	Other sexual abuse
Physical abuse, n = 510 (469)
74 (74)	You were spanked, whooped, or whipped, and it resulted in welts, bruises, bleeding, or other physical injuries
28 (29)	You were slapped or choked, and it resulted in welts, bruises, bleeding, or other physical injuries
19 (20)	You were punched, kicked, or beaten up, and it resulted in welts, bruises, bleeding, or other physical injuries
42 (42)	You were hit with an object, and it resulted in welts, bruises, bleeding, or other physical injuries
26 (25)	Other physical abuse

Note. The percentages listed in this table are based on only the subsample who reported abuse, not on the entire sample. Furthermore, the percentages listed parenthetically indicate the proportion of victims who endorsed each of the abuse items out of the participants who answered the disclosure question with either yes or no (i.e., the sample of interest in the study).

The second section assessed issues related to physical abuse, measured with a question based on work by Straus and Gelles (1988):

When you were 17 years old or younger, did you ever have any of the following experiences where someone at least 5 years older than you used excessive physical force on you that resulted in welts, bruises, bleeding, or other physical injuries?

Respondents chose one or more items in Table 1. A Kuder–Richardson 20 reliability coefficient of .70 indicates the inter-relatedness of these checklist characteristics.

For both of these sections, if victims indicated multiple discrete instances of sexual or physical abuse (i.e., with different perpetrators), they were asked to complete the survey about the most traumatic experience. Then, participants were asked, “When you were 17 years old or younger, were you a victim of childhood [sexual/physical] abuse?” (yes or no). Respondents who chose yes were also considered to be abuse victims even if they had not chosen any of the objective checklist items.

Witnessing domestic violence and emotional abuse were assessed in the third section (“other traumas”) modeled after a survey designed by Falsetti (1996). Respondents were asked, “When you were 17 years old or younger, did you ever have any of the following experiences?” and were offered 10 alternative responses. In this study, we included only those who chose “witnessed domestic violence” or “emotional or verbal abuse,” and/or those who later indicated experiencing those events in response to an open-ended descriptive question. This section of our survey included all measures described herein except for (a) the identity of and emotional closeness to perpetrator, (b) frequency of the abuse, and (c) self-labeling as a victim, which, therefore, are not included in the results section.

Circumstances and outcomes of abuse disclosure

For each type of abuse, participants who had indicated that they experienced that particular type of abuse answered the disclosure question, “Did you ever tell anyone about your [abuse/traumatic] experience?” (yes or no). Otherwise, they were instructed to skip the disclosure question and all subsequent questions about the circumstances of disclosure. We also asked victims who disclosed their abuse to indicate the people to whom they disclosed (response options were friend, parent, other relative, significant other, therapist, teacher/clergy, legal authorities, and other) and what happened as a result (response options were abuse stopped, legal action ensued, parents divorced, abuser confessed, nothing, that is, abuse continued, and other).

Abuse characteristics

We measured victims’ age at the time the abuse started and stopped for physical and sexual abuses. For domestic violence and emotional abuse, participants reported their age at the “time the traumatic experience happened.” We asked victims to indicate the frequency with which they experienced the abuse: 1 (once), 2 (twice), 3 (3-5 times), 4 (6-10 times), 5 (11-20 times), and 6 (21 or more times). The resulting scale was used as a continuous measure of abuse frequency.

To measure perceived emotional impact of the abuse, we asked, (a) “When you experienced your (sexual abuse/physical abuse/traumatic event), did you worry about being seriously injured or killed?” (b) “At the time it occurred, was your (sexual abuse/physical abuse/traumatic) experience emotionally upsetting or distressing to you?” and (c) “Is your (sexual abuse/physical abuse/traumatic) experience emotionally upsetting or distressing to you now?” All were answered on scales ranging from 1 (no, not at all) to 7 (yes, very).

Victims also indicated their relationship with their abuser (response options: parent; step-parent; family member; non-family member, friend, or trusted adult; and acquaintance, stranger), and how emotionally close they felt to the perpetrator prior to the incident on a scale from 1 (not close at all) to 7 (extremely close).

Self-labeling

To discover whether victims labeled themselves as victims, after each of the separate checklists measuring sexual and physical abuses described earlier, the participants were asked, “When you were 17 years old or younger, were you a victim of childhood [sexual/physical] abuse?” (yes or no).

Repressive coping

As is standard in the field of personality and social psychology, we measured repressive coping by combining two measures that each have good reliability and construct validity (Weinberger, 1990). Specifically, the short form of the Marlowe–Crowne Social Desirability Scale (MCSDS; Strahan & Gerbasi, 1972) measured defensiveness (i.e., the tendency to protect self-esteem through rigid self-control) with 20 true/false statements (e.g., “I’m always willing to admit it when I make a mistake”; Kuder–Richardson 20 = .70 in this sample). The Taylor Manifest Anxiety Scale–Short Form (TMAS; Bendig, 1956) included 20 true/false statements such as “I feel anxiety about something or someone almost all the time” (α = .83 in this sample). Mean scores were 9.74 (SD = 3.50) on the MCSDS and 8.87 (SD = 4.68) on the TMAS.

When used together, the two scales identify four distinct categories of people: repressors (low anxiety, high defensiveness), low anxious (low anxiety, low defensiveness), high anxious (high anxiety, low defensiveness), and defensive high anxious (high anxiety, high defensiveness). Because only the comparison of repressors with non-repressors was of interest to us, we followed Weinberger’s (1990) suggestion of performing median splits on both scales and using the resulting categories to identify repressors as those low in anxiety and high defensiveness (n = 397 or 24% of the sample responding to this item) and non-repressors as all others (n = 1,282 or 76%).

Dissociative proneness

We measured this coping style by the percentage of time (0%-100%) victims had experienced each of the 28 items of the Dissociative Experiences Scale (e.g., “Some people have the experience of finding themselves in a place and having no idea how they got there,” Carlson & Putnam, 1993). The measure has good reliability and validity (e.g., Bernstein & Putnam, 1986); in our sample, α = .94, with overall average dissociation scores ranging from 0% to 68% (M = 13%, SD = 12%), consistent with scores in young non-clinical samples (Carlson & Putnam, 1993).

Fantasy proneness

Participants answered “yes” or “no” to experiencing the 52 items of the Inventory of Childhood Memory and Imaginings (ICMI; Wilson & Barber, 1983, for example, “As an adult, I occasionally pretend I am someone else”), which assesses imaginativeness, types of fantasies, and the extent to which childhood imaginings influence adult functioning. Summed fantasy proneness scores ranged from 0 to 44 (M = 16.82, SD = 7.25, Kuder–Richardson 20 = .84), consistent with scores in other non-clinical populations (Rhue, Lynn, Henry, Buhk, & Boyd, 1990-1991).

Temporary forgetting

We asked victims, “Was there ever a time when you could not remember your (sexual abuse/physical abuse/traumatic) experience?” (yes or no), a measure modeled after that used by Briere and Conte (1993) and in other studies (e.g., Epstein & Bottoms, 1998, 2002; Melchert & Parker, 1997).

Procedure

All participants received the same instructions and survey materials either as a take-home exercise or during an experimental session in a large room where respondents could not see each other’s words. The ordering of questions about abuse/trauma experiences was rotated across surveys according to a Latin square design, and the order of the defensive reactions individual difference measures versus the rest of the survey was counterbalanced. In keeping with the Institutional Review Board protocol, participants signed informed consent forms that were returned separately from surveys, surveys were not marked with any identifying information, and participants were assured that their answers would remain confidential and anonymous.

Results

Of the 1,679 participants, 853 (51%) had experienced some form of abuse. Because some victims reported experiencing more than one form of abuse, we could not perform direct statistical comparisons between the different forms of abuse. First, we present preliminary analyses to rule out effects of demographic and background factors. Second, for each type of abuse separately, we present descriptive profiles of the abuse characteristics and circumstances of disclosure. Third, for each type of abuse separately, we present the results of exploratory intercorrelations among all disclosure predictors that were measured. In our final set of analyses, we test our main hypotheses: For sexual and physical abuse, we present the results of separate logistic regression analyses comparing participants who disclosed to those who did not in terms of (a) abuse characteristics (i.e., victim age, abuse frequency and emotional impact, and perpetrator identity), and then (b) the individual differences (self-labeling, the three types of defensive coping, and temporary forgetting). For emotional abuse and witnessed domestic violence, we present the results of similar separate logistic regressions, but without the predictors of perpetrator identity, emotional closeness to perpetrator, abuse frequency, and self-labeling, which were not measured.

Characteristics of Abuse and Disclosure Circumstances

Twenty-eight percent (n = 467) of all participants reported at least one instance of sexual abuse (see Table 1), consistent with prior studies (e.g., Finkelhor, Hotaling, Lewis, & Smith, 1990, 27%; Martin et al., 1993, 25%), and 414 of those answered the disclosure question. Of those, 77% (n = 319) reported that they had disclosed their abuse. Sexual abuse was perpetrated by parents in 5% of cases, step-parents in 3%, acquaintances in 19%, and strangers in 9% of cases. Most abuse was perpetrated by other relatives (34%) and trusted adults (30%). Sexual abuse victims were most likely to report disclosing to friends (67%), followed by parents (45%), significant others (31%), other relatives (28%), and therapist (16%). Very few victims disclosed to a recognized authority figure (teacher or clergy, 7%; legal authorities, 9%). In terms of outcome, disclosure ended the abuse for only 29% of the victims; legal action (8%) and perpetrator confession (7%) were rare. Nothing happened and the abuse continued in 16% of cases, parents divorced in 1%, and some “other” unspecified outcome happened in 60% of cases.

Of the total sample, 30% (n = 510) reported physical abuse (Table 1), and 469 answered the disclosure question. Of those, 66% (n = 311) reported disclosing their abuse. Trusted family members usually perpetrated physical abuse (i.e., parents, 77%; step-parents, 3%; and other family members, 12%). The perpetrator was rarely some other trusted adult (4%), acquaintance (3%), or stranger (1%). Victims were most likely to disclose to friends (79%), followed by other relatives (36%), parents (31%), significant others (27%), and therapist (18%). Again, few disclosed to an authority figure such as teacher or clergy (8%) or legal authorities (7%). Disclosure ended the abuse for 24% of victims. Legal action (5%) and perpetrator confession (4%) were rare, and nothing happened with the abuse continuing in 42% of cases, parents divorced in 3%, and some “other” unspecified outcome happened in 31% of cases.

Eleven percent of participants (n = 191) reported emotional abuse. Of those who answered the disclosure question (n = 182), 80% (n = 146) reported that they had disclosed their abuse. They were most likely to report disclosing to friends (81%), followed by parents (44%), other relatives (36%), significant others (25%), and therapists (19%). Disclosure to an authority figure was rare (teacher and clergy, 14%; legal authorities, 6%), as was legal action (2%) and perpetrator confession (1%). The abuse ended after disclosure in only about 13% of cases. Nothing happened and the abuse continued in 60% of cases, parents divorced in 1%, and there was some “other” unspecified outcome in 33% of cases.

Eight percent of participants (n = 133) reported witnessing domestic violence. Of those who answered the disclosure question (n = 128), 71% (n = 91) reported that they disclosed the experience. Victims were most likely to disclose to friends (70%), followed by other relatives (53%), significant others (35%), parents (28%), and therapists (8%). Only 7% disclosed to a teacher or clergy and 7% to legal authorities. Disclosure ended the abuse in only about 13% of cases. No perpetrators confessed because of the disclosure and legal action ensued in only 7% of the cases. The child’s parents divorced in 12% of the cases. Nothing happened and the abuse continued in 49% of cases, and there was some “other” unspecified outcome in 33% of cases.

Predictors of Disclosure¹

Mean responses of victims to all predictor variables, as a function of disclosure, are presented in Table 2 separately for each abuse type.

Table 2.

Disclosers and Non-Disclosers Comparative Means and Frequencies for All Predictors.

	Disclosed Sexual Abuse		Disclosed Physical Abuse		Disclosed Emotional Abuse		Disclosed Witnessing Domestic Violence
Predictor	Yes	No	Yes	No	Yes	No	Yes	No
Assault characteristics
Age abuse	9.92	9.75	8.12	7.59	11.68	10.03	9.11	9.43
Started/happened	(4.2)	(4.4)	(4.2)	(3.3)	(3.9)	(3.3)	(4.1)	(3.6)
Frequency of abuse	2.95 (1.8)	2.47 (1.5)	3.94 (1.7)	3.13 (1.6)	N/A	N/A	N/A	N/A
Worry	2.45 (1.9)	1.44 (0.9)	3.16 (2.1)	2.13 (1.6)	2.53 (1.8)	2.17 (1.6)	3.07 (2.1)	2.92 (2.1)
Upset	4.73 (2.3)	3.66 (2.3)	5.45 (1.8)	4.39 (2.2)	6.44 (1.0)	5.89 (1.5)	6.23	6.38
Closeness to perpetrator	3.40 (2.1)	3.27 (2.0)	4.72 (2.0)	5.18 (1.8)	N/A	N/A	N/A	N/A
Family perpetrator
Yes	79%	21%	66%	34%	N/A	N/A	N/A	N/A
No	75%	25%	67%	33%	N/A	N/A	N/A	N/A
Individual differences
Temporary forgetting
Yes	69%	31%	68%	32%	70%	30%	71%	29%
No	78%	22%	63%	37%	81%	19%	71%	29%
Label as victim
Yes	65%	35%	50%	50%	N/A	N/A	N/A	N/A
No	36%	64%	24%	76%	N/A	N/A	N/A	N/A
Repressor
Yes	76%	24%	52%	45%	62%	38%	74%	26%
No	77%	23%	69%	31%	81%	19%	70%	30%
Dissociative coping	13.90 (13.2)	11.39 (10.9)	13.40 (12.5)	14.28 (13.6)	13.20 (12.2)	16.25 (14.7)	12.33 (12.1)	13.84 (14.4)
Fantasy proneness	18.09 (7.3)	17.66 (6.8)	18.72 (7.2)	18.14 (6.9)	18.46 (6.6)	17.44 (6.9)	17.17 (7.7)	19.16 (7.9)

Note. For continuous predictors (i.e., age abuse started, frequency of abuse, worry, upset, closeness to perpetrator, dissociative coping, and fantasy proneness) the numbers in each cell denote the means for disclosers versus non-disclosers. For categorical predictors (i.e., family perpetrator, temporary forgetting, labeling as a victim, and repressor), the numbers denote the percentage of disclosers and non-disclosers who endorsed each predictor. Where means are reported, the standard deviations are noted parenthetically. NA = not applicable.

Preliminary analyses

There were no significant differences between victims who disclosed and those who did not in terms of victim race/ethnicity for cases of sexual abuse, χ²(3, N = 388) = 3.36, ns; physical abuse, χ²(3, N = 439) = 2.35, ns; emotional abuse, χ²(3, N = 168) = 6.38, ns; or witnessing domestic violence, χ²(3, N = 120) = 1.30, ns. Nor did disclosers and non-disclosers differ in terms of parental income, age at the time of the survey, or education, all ts (0.13-1.91) ns. Thus, we collapsed across these demographic predictors in other analyses.

Intercorrelations among all predictor variables

We performed exploratory correlations among all disclosure predictors (Pearson’s r coefficients for continuous variables, point-biserial coefficients for dichotomous-continuous variables, and Spearman’s coefficient for dichotomous variables) to examine potential inter-relationships. Analyses were done separately for each of the four types of abuse. All statistically significant findings are reported here, using a conservative probability level of p < .01, given the large number of correlations.

For child sexual abuse, the more likely a victim was to self-label as a victim, the more frequent the abuse (.31), the closer the perpetrator (.16) and the more likely the perpetrator was to be a family member (.62), the more she felt upset (.38) and worried about danger (.25), and the younger she was at onset of abuse (−.38). In addition, the older the victim at the onset of abuse, the lower the frequency of the abuse (−.16), the less likely the perpetrator was a family member (−.36), and the less likely the victim was to experience temporary forgetting (−.32). Children who were abused by a family member felt closer to the perpetrator (.33), were abused more frequently (.21), and were more upset by the abuse (.17). Finally, victims who were more upset by the abuse were also more worried (.40) and less likely to temporarily forget the abuse (−.13).

For child physical abuse, self-labeling as a victim was more likely for participants who experienced abuse that happened more frequently (.39) and was perpetrated by a family member (.54)—yet who felt less close to the perpetrator (−.21). Those who self-labeled felt more upset (.36) and worry at the time of the abuse (.40), and were more likely to have forgotten the abuse at some point (.14), and were younger rather than older at the time the abuse started (−.13). Self-labeling was more likely for those who were fantasy prone (.13) and less for those who were repressors (−.13).

Furthermore, for physical abuse victims, being older when the abuse started was associated with suffering abuse not perpetrated by a family member (−.31) and that happened less frequently (−.43), and being a repressor (.14). Children who suffered more frequent abuse were more worried (.30) and upset (.24), more likely to be abused by a family member (.14), more fantasy prone (.14), and less likely to be repressors (−.17). Children abused by family members felt closer to the perpetrator (.11), were less worried (−.21), were less likely to be repressors (−.11), and were more fantasy prone (.13). Children who were more worried at the time of the abuse were also more upset (.43), more fantasy prone (.11), and less likely to be repressors (−.12).

There were no significant intercorrelations for witnessing domestic violence, and fewer for emotional abuse than for sexual and physical abuse: Those who were more upset by the abuse were more worried (.15), more dissociative prone (.16), and less likely to be repressors (−.27).

Multivariate prediction of sexual abuse disclosure

We included all abuse-characteristics predictors into a logistic regression predicting whether sexual abuse victims disclosed their abuse (Nagelkerke R² = .13, see Table 3). This revealed that sexual abuse disclosers (vs. non-disclosers) had been abused more frequently, were more worried about being seriously injured or killed at the time of the abuse, and more upset at the time of the abuse. A similar regression with individual differences as predictors (Nagelkerke R² = .11) revealed that disclosers (vs. non-disclosers) were less likely to forget the abuse, and more likely to self-label as victims.

Table 3.

Logistic Regression Results for Predicting Disclosure of Sexual and Physical Abuse.

	Disclosure of Child Sexual Assault (Yes)				Disclosure of Child Physical Assault (Yes)
Predictor	B	Wald	OR	OR 95%CI [LL, UL]	B	Wald	OR	OR 95%CI [LL, UL]
Assault characteristics
Age abuse started	0.02	0.31	1.02	[0.84, 1.13]	0.07	3.78*	1.07	[1.00, 1.15]
Frequency of abuse	0.20	5.50*	1.22	[0.96, 1.42]	0.24	9.66*	1.27	[1.09, 1.48]
Worry	0.38	11.11*	1.46	[1.44, 1.79]	0.13	3.82*	1.14	[1.00, 1.30]
Upset	0.12	4.15*	1.13	[1.01, 1.27]	0.15	6.74*	1.17	[1.04, 1.31]
Closeness to perpetrator	−0.01	0.02	0.99	[0.86, 1.14]	−0.13	4.70*	0.88	[0.78, 0.99]
Family perpetrator	0.04	0.01	1.04	[0.56, 1.88]	0.29	0.48	1.34	[0.59, 3.05]
Individual differences
Temporary forgetting	−0.70	4.55*	0.50	[0.26, 0.95]	−0.43	1.39	0.65	[0.32, 1.33]
Label as victim	1.22	22.01*	3.40	[2.04, 5.66]	1.16	23.9*	3.20	[2.01, 5.10]
Repressor	0.14	0.14	1.15	[0.56, 2.36]	−0.66	5.59*	0.52	[0.30, 0.89]
Dissociative coping	0.01	1.41	1.01	[0.99, 1.03]	−0.02	4.30*	0.98	[0.96, 1.00]
Fantasy proneness	0.01	0.07	1.01	[0.97, 1.04]	0.01	0.39	1.01	[0.98, 1.04]

Note. The table includes unstandardized coefficients (B), Wald statistics, odds ratios, and 95% confidence intervals for odds ratios for two logistic regression models measuring the likelihood of abuse disclosure as a function of (a) assault characteristics and (b) individual differences. All predictors are continuous except family perpetrator (no = 0, yes = 1), temporary forgetting (no = 0, yes = 1), label as victim (no = 0, yes = 1), and repressor (no = 0, yes = 1). OR = odds ratio; CI = confidence interval.

p < .05.

Multivariate prediction of physical abuse disclosure

Our abuse-characteristics logistic regression (Nagelkerke R² = .16) revealed that emotional abuse disclosers (vs. non-disclosers) had been older when the abuse started, were abused more frequently, were less close to the perpetrator, and were more worried about injury and upset at the time of the abuse (see Table 3). A similar regression with individual differences as predictors (Nagelkerke R² = .12) revealed that disclosers (vs. non-disclosers) were more likely to self-label as victims, less likely to be repressors, and less dissociative prone.

Multivariate prediction of emotional abuse disclosure

The abuse-characteristics logistic regression (Nagelkerke R² = .09) revealed that emotional abuse disclosers (vs. non-disclosers) had been older when the abuse started and were more upset at the time the abuse took place (see Table 4). A similar regression with individual differences as predictors (Nagelkerke R² = .11) revealed that disclosers (vs. non-disclosers) were less likely to be repressors and less dissociative prone.

Table 4.

Logistic Regression Results for Predicting Disclosure of Emotional Abuse and Witnessing Domestic Violence.

	Disclosure of Child Emotional Abuse (Yes)				Disclosure of Witnessing Domestic Violence (Yes)
Predictor	B	Wald	OR	OR 95% CI [LL, UL]	B	Wald	OR	OR 95%CI [LL, UL]
Assault characteristics
Age abuse happened	0.12	4.73*	1.13	[1.01, 1.25]	−0.02	0.16	0.98	[0.89, 1.08]
Worry	0.04	0.12	1.04	[0.82, 1.34]	0.06	0.40	1.07	[0.88, 1.30]
Upset	0.31	4.08*	1.37	[1.01, 1.85]	−0.20	1.06	0.82	[0.56, 1.20]
Individual differences
Temporary forgetting	−0.47	0.38	0.62	[0.14, 2.80]	0.06	0.01	1.06	[0.20, 5.79]
Repressor	−1.42	8.44*	0.24	[0.09, 0.63]	0.15	0.10	1.17	[0.44, 3.08]
Dissociative coping	−0.04	5.29*	0.96	[0.93, 0.100]	−0.00	0.02	1.00	[0.97, 1.03]
Fantasy proneness	0.03	0.88	1.03	[0.97, 1.10]	−0.03	1.30	0.97	[0.92, 1.02]

Note. The table includes unstandardized coefficients (B), Wald statistics, odds ratios, and 95% confidence intervals for odds ratios for two logistic regression models measuring the likelihood of abuse disclosure as a function of (a) assault characteristics and (b) individual differences. All predictors are continuous except temporary forgetting (no = 0, yes = 1) and repressor (no = 0, yes = 1). OR = odds ratio; CI = confidence interval; LL = lower limit, UL = upper limit.

p < .05.

Multivariate prediction of the disclosure of witnessing domestic violence

Neither the logistic regression models with abuse characteristics as predictors (Nagelkerke R² = .02) nor the regressions with individual differences as predictors (Nagelkerke R² = .02) revealed any significant relations between the predictors and likelihood of disclosure, all ps > .05 (see Table 4).

Discussion

We investigated disclosure patterns for not only sexual abuse but also physical abuse, emotional abuse, and, for the first time, witnessing domestic violence. The sexual and physical abuse disclosure rates we found are similar to those found by prior researchers (e.g., sexual abuse, Arata, 1998; physical abuse, Foynes et al., 2009), but our emotional abuse disclosure rate (80%) is higher than Foynes et al.’s (42%), perhaps because those researchers excluded victims who delayed disclosure for 1 year. The disclosure rate for witnessing domestic violence was even lower, at 71%. Non-disclosure of this type of experience, which could be considered emotional or vicarious abuse, may result from victims feeling ashamed of the violence in their home and fear of retaliation or disruption of the family, motivational factors that future work should explore.

For all types of abuse, most victims disclosed to friends, followed by parents and other relatives, consistent with prior research (e.g., Lamb & Edgar-Smith, 1994). Some sexual, physical, and emotional abuse victims also disclosed to a therapist, suggesting that these abuses have a significant impact on victims’ emotional well-being. Few participants who witnessed domestic violence reported to a therapist even though they reported the experience to be emotionally distressing. Perhaps because victims were not a direct target of the violence, they did not acknowledge this as directly harmful in a way that would warrant therapy.

The low percentage of all victims making a formal report of abuse to legal authorities (< 10%) is very close to rates reported by others (Arata, 1998, 10%; Smith et al., 2000, 12% for discussion, see Lyon, 2009) and is troubling: When child abuse is disclosed, it does not usually lead to an official investigation. As a result, abuse may not stop and perpetrators are undeterred from future abuse. Only 16% of sexual assault victims said “nothing” happened as a result of disclosure, but that number was about half for disclosers of other abuse. Physical and emotional abuse, as well as witnessing domestic violence, seem to be perceived as less deserving of active interventions, perhaps because corporal punishment is still widely accepted in American society (e.g., Straus & Gelles, 1988), and emotional abuse is discounted as less severe than physical forms of abuse.

Of course, disclosure is a complicated issue, and from the subjective perspective of the child, it might often be problematic. Although disclosure appears to be beneficial and is associated with lower levels of posttraumatic stress disorder (PTSD; Arata, 1998), people sometimes feel worse after the disclosure of a personal, traumatic experience of abuse (Kelly, Coenen, & Johnston, 1995). We also know that the effects of disclosure on subsequent well-being greatly depend on the reactions children receive when they disclose (e.g., Ullman, 2007).

Predictors of Disclosure

Overall, abuse characteristics were better predictors of disclosure than individual-level victim characteristics, and nothing we measured significantly predicted the disclosure of domestic violence.

Abuse characteristics

As predicted, indicators of more severe abuse such as frequency, worry about being injured, and being emotionally upset were related to higher likelihood of disclosure for sexual and physical abuse. Sexual and physical abuse disclosers (vs. non-disclosers) were more worried about serious injury and more upset, unlike some studies (Arata, 1998), but consistent with others (e.g., Hanson et al., 1999). This difference could be due to the conceptualizations (and measures) of disclosure. Specifically, Arata (1998) asked survivors whether they had disclosed at the time of the abuse; we, however, asked them whether they had ever disclosed. It is possible that frequency, fear of injury, and increased distress could lead to both delayed disclosure and higher disclosure rates: Fear-inducing tactics might be effective for short-term deterrence of disclosure, but over longer periods, fear might motivate victims to seek help to stop the abuse, or to seek emotional support even after the abuse ends (Hanson et al., 1999; Ullman, 2007). Those who were more upset and worried were also more likely to label themselves as abuse victims, which could also contribute to the motivation to disclose. Similarly, emotional abuse disclosers were more likely than non-disclosers to be upset at the time of abuse, but not more worried about physical injury, which is not surprising given the non-physical nature of this type of abuse.

Alternatively, frequent abuse and emotional distress might also be more likely to attract the attention of others and, in turn, lead to disclosure, even if that disclosure is not an active outcry from the victim, a possibility also suggested by Hanson and colleagues (1999) and Bottoms, Shaver, and Goodman (1996). Future research should include more specific questions about the nature of disclosure, such as whether the child received medical attention as a result of the abuse or whether injuries were noticed by persons such as teachers or friends.

We also found that disclosers of physical and emotional abuse were older than non-disclosers at the time of the abuse. Because physical abuse was mostly perpetrated by family members, who are usually the receivers or targets of disclosure for younger children (Lamb & Edgar-Smith, 1994), it is not surprising that children who were younger at the time of abuse did not have the opportunity to disclose. Older children, in contrast, might have been more likely to develop peer or other relationships that foster disclosure when the perpetrator is a family member. Our correlational analyses offer an alternative explanation: The older the age of onset for physical abuse, the less likely it was that the perpetrator was a family member, and the less frequent the abuse—suggesting that some of the abuse reported by older children was perpetrated outside the family, which perhaps reduced their motivation to hide the abuse. It is also possible, as Hershkowitz et al. (2005) suggested, that younger children lack the cognitive and linguistic skills to understand and communicate their abusive experiences to others. Whatever the reason, we have confirmed prior findings documenting the direct correlation between age and sexual abuse disclosure, and extended this finding to physical and emotional abuse disclosure.

Like Arata (1998), we found no evidence of a link between sexual abuse disclosure and perpetrator identity. We did find, however, that non-disclosers of physical abuse felt more emotionally close to the perpetrator than did disclosers, in line with Foynes et al. (2009) who found that emotional closeness deterred children from disclosing physical and emotional abuse. This finding confirms our hypothesis that victims tend to protect trusted abusers, who, in physical abuse cases, are most often the parent or other relatives—as was the case in our study, where victims felt more emotionally close to perpetrators who were family members. According to Betrayal Trauma Theory, abuse perpetrated by an emotionally close adult is perceived by the child as a betrayal and, therefore, less likely to be remembered when compared with other traumas, and in turn allows the child to maintain a close relationship with the adult, despite the abuse (Foynes et al., 2009). This was not supported in our sample, where closeness to perpetrator was unrelated to temporary forgetting—although it was related to feeling more upset in the case of sexual abuse. In contrast, physical abuse (i.e., corporal punishment) from an emotionally close adult might be regarded by the victim as normal, not abusive, and not warranting disclosure. In support, the closer physical abuse victims felt to the perpetrators, the less upset and worried they were about injuries, and the less likely they were to self-label as victims—which could indicate they might not even have deemed the experience as abuse at the time, and saw no reason to disclose.

Individual differences

As predicted, sexual and physical abuse victims who self-labeled as victims were more likely to disclose their abuse. Someone who does not self-label would not think of herself as having suffered abuse, and would not tell anyone about the experience. Our correlational analyses revealed that more frequent, worrying, and upsetting forms of abuse resulted in a higher likelihood of identifying as a victim, so victims may be more motivated to disclose more severe abuse to stop it. Alternatively, however, frequent and upsetting abuse is more likely to be noticed by others, which in turn can lead to conversations about the experience that could lead the victims to understand that the experience is abusive. Therefore, disclosure might lead to self-labeling, rather than the reverse. We cannot establish the temporal order of disclosure and self-labeling, but prior work conceptualizes self-labeling as a precursor, rather than outcome, of disclosure (Browne, 1991; Littleton, Axsom, Breitkopf, & Berenson, 2006). Moreover, self-labeling in our sample was related to most other predictors we considered, even those unrelated directly to disclosure (e.g., identity of perpetrator and younger age at abuse onset), which also supports our conceptualization of self-labeling as a precursor, rather than a result, of disclosure. Thus, survivors’ subjective perceptions of themselves as victims are important when it comes to disclosing abuse and perhaps seeking help. When victims’ perceptions do not fit objective definitions of victimization, abuse is likely to go undisclosed or unnoticed. Public awareness of the various forms of sexual and physical abuse is important for victims to understand when their experiences are abusive, and, perhaps, in turn, to foster disclosure.

Temporary forgetting of abuse was also associated with lower disclosure rates for child sexual abuse, a result consistent with Arata’s (1998) finding of a relation between forgetting and sexual abuse disclosure. This relation could indicate, as Arata suggests, that non-disclosure is one of many forms of avoidance used by some victims, who actively choose not to think about the abuse, which can result in temporary forgetting and avoiding discussing the topic with others. Temporary forgetters (vs. non-forgetters) were also significantly less upset, so it is also possible that abuse not perceived as severe is less salient, and, therefore, more likely to be forgotten and in turn less likely to be disclosed, even if forgetting is only for temporary periods (Goodman et al., 2003).

Chronic defensive coping styles were related to disclosure only for physical and emotional abuse victims: As predicted, victims with repressive and dissociative tendencies were less likely to disclose their abuse. But this was not true for sexual abuse victims, which is surprising in light of prior research (Bonanno et al., 2003; Goodman et al., 2003). It is possible that sexual abuse is more likely than physical or emotional abuse (which can be construed as disciplining the child) to be noticed by others and identified as abuse regardless of victims’ desire to repress negative emotions and hide the assault. In contrast, because physical and emotional assault are more ambiguous forms of abuse, it is more likely that (a) others do not become worried and prompt the child to disclose, and (b) repressors and those with dissociative tendencies are successful in subduing negative emotions or have less intense emotions related to the assault, and thus feel less need to disclose. Of note, repressive and dissociative tendencies were negatively related to each other (as found by Bonanno et al., 2003), suggesting that participants might have used these coping strategies separately to cope with the abuse, and that the strategies were different paths that both lead to avoiding abuse disclosure.

Strengths and Limitations

Studying childhood abuse scientifically is difficult, and the field must often necessarily settle for converging evidence from multiple studies rather than getting all questions answered in one study. Like others, our study has strengths and weaknesses. A central strength is its uniqueness in light of the paucity of research on patterns of disclosure and non-disclosure not only of sexual abuse, which is of course an important threat to children’s well-being, but also physical abuse, a more common form of child maltreatment, and emotional abuse which has been given almost no attention in the disclosure literature. Ours is the only study we know of to address the disclosure of witnessed domestic violence, which can also be a serious threat to children’s psychological well-being (e.g., Kitzmann et al., 2003). Our study is, therefore, timely and novel, providing the first rigorous examination of the extent to which such forms of maltreatment are disclosed, are considered worthy of self-labeling as abusive by victims, and affected their victims at the time and years later, with an emphasis on how disclosure is related to all those factors as well as victim’s defensive coping strategies. We believe we have provided foundational information that will assist further research on the disclosure of all these forms of abuse.

A particular strength of our study is our use of an anonymous survey methodology that afforded participants the confidence to provide honest answers. Our method of defining abuse objectively (with the checklist approach) and subjectively (with a question about perceived victimhood) also allowed for more accurate and inclusive identification of abuse victims. The value of multiple, focused questions that define specific behaviors is that they encourage more respondents to acknowledge the abuse, reducing, at least in part, people’s tendency to under-report (Lyon, 2009). Our method also allowed us to study the phenomenon of self-labeling and isolate its effect on disclosure, finding that victims who perceived themselves as victims were approximately twice as likely to disclose abuse than were victims who has similar experiences but did not consider themselves to be victims.

Because our participants were not aware of the topic of the study before coming to the survey session, our sample is not biased by containing an over-representation of victims who label themselves as such and who might, therefore, have been more likely to have disclosed. The likelihood of false abuse allegations is also reduced when participants do not self-select for a study related to abuse (Lyon, 2009). This approach is more likely to include participants who have never disclosed their abuse; therefore, it is less likely to over-estimate disclosure rates. In contrast, studies of reports to legal authorities or to therapists do not capture victims who have never disclosed. According to Lyon (2009), surveys of college samples offer some of the central benefits of general, versus clinical, population samples: They are unlikely to elicit false endorsements of abuse, allow for a separate assessment of disclosures to legal authorities, and are a better proxy for abuse incidence than clinical samples (i.e., samples of already-documented abuse victims).

Thus, another strength of our study is its large, diverse, non-clinical sample, which also allowed for testing relations between disclosure and race and socioeconomic status. The large number of participants allowed us to have confidence in making statistical comparisons between victims who did and did not disclose their abuse. The non-clinical sample also provided more information than clinical samples about the relations among variables in the population, and about victims who have never told anyone about their abusive experiences—even those who never thought of their experiences as abusive. Even so, our findings might not represent the experiences of individuals such as men, older people, or victims who suffered extremely severe abuse or emotional sequelae and who, therefore, are less likely to appear in a sample of highly functioning college students (e.g., Duncan, 1999).

Finally, our study is unique in examining the relation between disclosure and individual differences in the way victims defensively cope with abusive experiences, differentiating between the strategies of temporarily forgetting memories of abuse and using repressive, dissociative, and fantasy tendencies.

Of course, our methodology also has some limitations. For example, as is true for any retrospective self-report study, our conclusions must be tempered by the fact that we rely on the victims’ memories and reports, even though people are not always accurate in reflecting on their past experiences (Azar, 1997). Even so, such retrospective methodologies have been used successfully by other researchers in the field of repressed memories and other child trauma–related issues (e.g., Goodman et al., 2003; Piper, Lillevik, & Kritzer, 2008). Furthermore, based on their review of 14 empirical studies, Hardt and Rutter (2004) concluded that “when abuse or neglect is retrospectively reported to have taken place, these positive reports are likely to be correct (p. 270).” Finally, self-report studies are only one of many important pieces of converging evidence when investigating sensitive and difficult issues such as child abuse, because there are many obvious practical impediments to methodological alternatives.

Another limitation is that we measured fewer potential predictors of emotional abuse and witnessing domestic violence than of sexual or physical abuse. This is because these cases were pulled from a section of the survey that largely focused on traumas that do not involve perpetrators (e.g., car accident), so many of the questions asked about sexual and physical abuse (e.g., perpetrator identity, emotional closeness) would not have applied. Even with our more limited number of predictors, however, we found several important differences between disclosers and non-disclosers of emotional abuse (age, distress, and coping styles), adding valuable information to the extant sparse research on the topic. The variables we measure for witnessing domestic violence, however, failed to predict disclosure, suggesting that these variables are less relevant to this form of abuse compared to the more direct forms (i.e., sexual, physical, and emotional). The mechanisms leading to its disclosure must be different from those for other abuses, perhaps because child witnesses are not the direct targets of this abuse and might feel that it is not their secret to keep or share. Further research on potential triggers or inhibitors of witnessing domestic violence disclosure is very important given its frequency and the large emotional impact it can have on witnesses (Kitzmann et al., 2003; Maker et al., 1998; Teicher et al., 2006). That is, our data revealed that witnessing domestic violence elicited much worry about injury, as well as high rates of emotional distress at the time of the abuse and at the time of the survey.

A related concern might be that our method undercounts the true incidence of emotional abuse and witnessing domestic violence, by not providing participants with an objective list of behaviors as we did for sexual and physical abuse. Indeed, our rates for emotional abuse (11%) and witnessing domestic violence (8%) are lower than those reported elsewhere (e.g., Foynes et al., 2009; McCloskey & Walker, 2000; Osofsky, 2003). Although our study might not provide the most accurate measures of the incidence of emotional abuse and witnessed domestic abuse, our findings are still reasonable and generalizable estimates of the disclosure of these forms of abuse, which is the point of our research.

Finally, in response to a few measures, respondents chose the “other” category fairly often, an indication that the measure did not capture all the plausible answers. For example, an outcome of sexual assault disclosure other than those we anticipated was endorsed by 60% of the participants. It is not always possible to anticipate all plausible answer categories for measures, especially in “first step” research such as this study.

Conclusion

Our findings make an important contribution to a growing area of inquiry into the circumstances surrounding the disclosure and non-disclosure of childhood maltreatment. Our study reveals that a significant number of victims never disclose abuse at all, and even when they do, disclosure does not lead to formal investigation and does not bring an end to their abuse. It is of critical importance to ensure that investigation techniques guard against false disclosures of abuse; however, it is clear from the present study that techniques must also be sensitive to the huge societal problem of the under-reporting of child maltreatment. Although we would hope disclosure leads to perpetrators being apprehended, we recognize that, as our own data show, this is not always the case. A combination of research such as ours, to understand processes and outcomes of disclosure, and research on children’s eyewitness testimony is needed to ensure a future where children’s disclosure can lead to an end of abuse.

Footnotes

Acknowledgements

We thank John Briere and Sarah Ullman for comments on our survey instrument; Beth Schwartz, Gail Goodman, Jodi Quas, Simona Ghetti, and Maureen Smith for help with participant recruitment; and Alaine Kalder, Matthew Badanek, Nadine Stevoff, Catherine Pelzman, Kara Doering, and Elaine Shreder for valuable research assistance. Portions of this research were supported by a grant to Michelle Epstein from the National Institute of Mental Health.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Bette L. Bottoms is dean of the Honors College and professor of psychology at the University of Illinois at Chicago (UIC). Nationally, she is fellow of the American Psychological Association (APA) and past president of APA’s Society for Child and Family Policy and Practice and of its Section on Child Maltreatment. Her research on child abuse, children’s eyewitness testimony, and jurors’ perceptions of child victims and of juvenile offenders has been published widely, funded by the National Institute on Mental Health and the National Science Foundation, and recognized with the American Psychology-Law Society’s Saleem Shaw Early Career Award. She has edited five books and served on the editorial boards of journals such as Law and Human Behavior; Psychology, Public Policy, and Law; and Child Maltreatment. Vice Provost Bottoms has been on the faculty at UIC since 1992. She has received nine awards for excellence in undergraduate and graduate teaching and mentoring, including every such UIC award and her disciplinary field’s highest awards. She is currently helping to spearhead UIC’s unprecedented effort to improve undergraduate success.

Liana C. Peter-Hagene is a doctoral candidate in social psychology working with Dr. Bette Bottoms at University of Illinois at Chicago. Her research interests include factors that affect jury decision making, the role of emotions and self-regulation in decision-making, as well as disclosure of and recovery from childhood and adult sexual assault.

Michelle A. Epstein is a licensed Clinician Psychologist with a PhD from the University of Illinois at Chicago. She serves as an Advisory Board Member of Rebecca’s Dream (a non-for-profit dedicated to raising mental health awareness and decreasing stigma) and has a private practice in Deerfield, IL.

Tisha R. A. Wiley contributed to this paper while a graduate student at the University of Illinois at Chicago. She is currently a Program Director at the National Institutes of Health, National Institute on Drug Abuse (NIH/NIDA), where she manages a large portfolio of grants aimed at improving substance use treatment in specialty settings, including primary care, criminal justice, juvenile justice, and child welfare. She is also currently the NIDA scientist on the Juvenile Justice Translational Research on Adolescent in the Legal System (JJ-TRIALS) cooperative agreement, a large multi-site initiative that seeks to increase the adoption of evidence-based practices targeting youth substance use within the juvenile justice system.

Carrie E. Reynolds graduated from Northwestern University in 2007 with a BA in Psychology and received her MA in Social and Personality Psychology in 2011 from the University of Illinois at Chicago. She has experience in jury research and consulting, and her career currently is expanding into the market research field where she is heavily involved in multiple industries.

Aaron G. Rudnicki contributed to this paper during his time as a graduate student at the University of Illinois at Chicago. He is currently working as an Analyst in the Data Sciences group for Information Resources, Inc. in Chicago. IRI is a market research company which provides manufacturer and retailer clients with consumer, shopper, and retail market intelligence and analysis.

References

Arata

C. M.

(1998). To tell or not to tell: Current functioning of child sexual abuse survivors who disclosed their victimization. Child Maltreatment, 3, 63-71.

Azar

(1997). Poor recall mars research and treatment: Inaccurate self-reports can lead to faulty research conclusions and inappropriate treatment. The APA Monitor, 28, 1.

Bartholomew

Horowitz

L. M.

(1991). Attachment styles among young adults: a test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244.

Bendig

A. W.

(1956). The development of a short form of the Manifest Anxiety Scale. Journal of Consulting Psychology, 20, 384.

Bernstein

E. M.

Putnam

F. W.

(1986). Development, reliability, and validity of a dissociation scale. Journal of Nervous and Mental Disease, 174, 727-735.

Bonanno

G. A.

Noll

J. G.

Putnam

F. W.

O’Neill

Trickett

P. K.

(2003). Predicting the willingness to disclose childhood sexual abuse from measures of repressive coping and dissociative tendencies. Child Maltreatment, 8, 302-318.

Bonanno

G. A.

Singer

J. L.

(1990). Repressor personality style: Theoretical and methodological implications for health and pathology. In Singer

J. L.

(Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health (pp. 435-470). Chicago, IL: The University of Chicago Press.

Bottoms

B. L.

Goodman

G. S.

Schwartz-Kenney

B. M.

Thomas

S. N.

(2002). Understanding children’s use of secrecy in the context of eyewitness reports. Law and Human Behavior, 26, 285-313.

Bottoms

B. L.

Shaver

P. R.

Goodman

G. S.

(1996). An analysis of ritualistic and religion-related child abuse allegations. Law and Human Behavior, 20, 1-34.

10.

Bower

G. H.

Sivers

(1998). Cognitive impact of traumatic events. Development and Psychopathology, 10, 625-653.

11.

Bowlby

(1980). Attachment and loss. Vol. 1. Attachment. New York: Basic Books. Original work published 1969.

12.

Briere

(2002). Treating adult survivors of severe childhood abuse and neglect: Further development of an integrative model. In Myers

J.E.B.

Berliner

Briere

Reid

Jenny

(Eds.). The APSAC Handbook on child maltreatment, 2nd Edition (pp. 175-202). Newbury Park, CA: Sage Publications.

13.

Briere

Conte

(1993). Self-reported amnesia for abuse in adults molested as children. Journal of Traumatic Stress, 6, 21-31.

14.

Browne

(1991). The victim’s experience: Pathways to disclosure. Psychotherapy: Theory, Research, Practice, Training, 28, 150-156.

15.

Bureau of Justice Statistics. (2009). Profile of intimate partner violence in large urban counties. U.S. Department of Justice. Retrieved from http://www.bjs.gov/content/pub/pdf/pipvcluc.pdf.

16.

Burman

Duffy-Feins

(2007). Children witnessing parental violence. In Jackson

N. A.

(Ed.), Encyclopedia of domestic violence (pp. 156-160). New York, NY: Routledge.

17.

Carlson

E. B.

Putnam

F. W.

(1993). An update on the Dissociative Experiences Scale. Dissociation: Progress in the Dissociative Disorders, 6, 16-27.

18.

Duncan

(1999, July). Childhood maltreatment and college drop-out rates: Implications for researchers and educators. Presented at the 7th Annual Colloquium of the American Professional Society on the Abuse of Children, San Antonio, TX.

19.

Epstein

M. E.

Bottoms

B. L.

(1998). Memories of childhood sexual abuse: A survey of young adults. Child Abuse & Neglect, 22, 217-238.

20.

Epstein

M. E.

Bottoms

B. L.

(2002). Forgetting and recovery of abuse and trauma memories: Possible mechanisms. Child Maltreatment, 7, 210-225.

21.

Falsetti

(1996). Trauma assessment for adults: Self report version. Unpublished manuscript.

22.

Finkelhor

(1979). Sexually victimized children. New York, NY: Free Press.

23.

Finkelhor

Hotaling

Lewis

I. A.

Smith

(1990). Sexual abuse in a national survey of adult men and women: Prevalence, characteristics, and risk factors. Child Abuse & Neglect, 14, 19-28.

24.

Foynes

M. M.

Freyd

J. J.

DePrince

A. P.

(2009). Child abuse: Betrayal and disclosure. Child Abuse & Neglect, 33, 209-217.

25.

Goodman

G. S.

Ghetti

Quas

J. A.

Edelstein

R. S.

Alexander

K. W.

Redlich

A. D.

. . . Jones

D. P. H.

(2003). A prospective study of memory for child sexual abuse: New findings relevant to the repressed-memory controversy. Psychological Science, 14, 113-118.

26.

Goodman-Brown

T. B.

Edelstein

Goodman

G. S.

Jones

Gordon

D. S.

(2003). Why children tell: A model of children’s disclosures of sexual abuse. Child Abuse & Neglect, 27, 525-540.

27.

Hanson

R. F.

Resnick

H. S.

Saunders

B. E.

Kilpatrick

D. G.

Best

(1999). Factors related to the reporting of childhood rape. Child Abuse & Neglect, 23, 559-569.

28.

Hardt

Rutter

(2004). Validity of adult retrospective reports of adverse childhood experiences: Review of the evidence. Journal of Child Psychology and Psychiatry, 45, 260-273.

29.

Hershkowitz

Horowitz

Lamb

M. E.

(2005). Trends in children’s disclosure of abuse in Israel: A national study. Child Abuse & Neglect, 29, 1203-1214.

30.

Kelly

A. E.

Coenen

M. E.

Johnston

B. L.

(1995). Confidants’ feedback and traumatic life events. Journal of Traumatic Stress, 8, 161-169.

31.

Kitzmann

K. M.

Gaylord

N. K.

Holt

A. R.

Kenny

E. D.

(2003). Child witnesses to domestic violence: A meta-analytic review. Journal of Consulting and Clinical Psychology, 71, 339-352.

32.

Kogan

S. M.

(2004). Disclosing unwanted sexual experiences: Results from a national sample of adolescent women. Child Abuse & Neglect, 28, 147-165.

33.

Lamb

Edgar-Smith

(1994). Aspects of disclosure: Mediators of outcome of childhood sexual abuse. Journal of Interpersonal Violence, 9, 307-326.

34.

Littleton

H. L.

Axsom

Breitkopf

C. R.

Berenson

(2006). Rape acknowledgment and postassault experiences: How acknowledgment status relates to disclosure, coping, worldview, and reactions received from others. Violence and Victims, 21, 761-778.

35.

London

Bruck

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, and Law, 11, 194-226.

36.

London

Bruck

Ceci

S. J.

Shuman

D. W.

(2007). Disclosure of child sexual abuse: A review of the contemporary empirical literature. In Pipe

M. E.

Lamb

Orbach

Cederborg

A. C.

(Eds.), Child sexual abuse: Disclosure, delay, and denial (pp. 11-40). Mahwah, NJ: Lawrence Erlbaum.

37.

Lyon

T. D.

(2007). False denials: Overcoming methodological biases in abuse disclosure research. In Pipe

M. E.

Lamb

Orbach

Cederborg

A. C.

(Eds.), Child sexual abuse: Disclosure, delay and denial (pp. 41-62). Mahwah, NJ: Lawrence Erlbaum.

38.

Lyon

T. D.

(2009). Abuse disclosure: What adults can tell. In Bottoms

B. L.

Najdowski

Goodman

G. S.

(Eds.), Children as victims, witnesses, and offenders: Psychological science and the law (pp. 19-35). New York, NY: Guilford Press.

39.

Maker

A. H.

Kemmelmeier

Peterson

(1998). Long-term psychological consequences in women of witnessing parental physical conflict and experiencing abuse in childhood. Journal of Interpersonal Violence, 13, 574-589.

40.

Martin

Anderson

Romans

Mullen

O’Shea

(1993). Asking about child sexual abuse: Methodological implications of a two stage survey. Child Abuse & Neglect, 17, 383-392.

41.

Mazza

Dennerstein

Ryan

(1996). Physical, sexual and emotional violence against women: A general practice-based prevalence study. Medical Journal of Australia, 164, 14-17.

42.

McCloskey

L. A.

Walker

(2000). Posttraumatic stress in children exposed to family violence and single-event trauma. Journal of the American Academy of Child & Adolescent Psychiatry, 39, 108-115.

43.

Melchert

T. P.

Parker

R. L.

(1997). Different forms of childhood abuse and memory. Child Abuse & Neglect, 21, 125-135.

44.

Nemiah

J. C.

(1980). Dissociative disorders. In Freedman

A. M.

Kaplan

H. I.

(Eds.), Comprehensive textbook of psychiatry (pp. 1544-1561). Baltimore, MD: Williams & Wilkins.

45.

Osofsky

J. D.

(2003). Prevalence of children’s exposure to domestic violence and child maltreatment: Implications for prevention and intervention. Clinical Child and Family Psychology Review, 6, 161-170.

46.

Paine

M. L.

Hansen

D. J.

(2002). Factors influencing children to self-disclose sexual abuse. Clinical Psychology Review, 22, 271-295.

47.

Pipe

M. E.

Lamb

Orbach

Cederborg

A. C.

(Eds.). (2007). Child sexual abuse: Disclosure, delay, and denial. Mahwah, NJ: Lawrence Erlbaum.

48.

Piper

Lillevik

Kritzer

(2008). What’s wrong with believing in repression?: A review for legal professionals. Psychology, Public Policy, and Law, 14, 223-242.

49.

Rhue

J. W.

Lynn

S. J.

Henry

Buhk

Boyd

(1990-1991). Child abuse, imagination and hypnotizability. Imagination, Cognition and Personality, 10, 53-63.

50.

Smith

D. W.

Letourneau

Saunders

Kilpatrick

D. G.

Resnik

H. S.

Best

C. L.

(2000). Delay in disclosure of childhood rape: Results from a national survey. Child Abuse & Neglect, 24, 273-287.

51.

Strahan

Gerbasi

K. C.

(1972). Short, homogeneous versions of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology, 28, 191-193.

52.

Straus

M. A.

Gelles

R. J.

(1988). How violent are American families? Estimates from the National Family Violence Survey and other studies. In Hotaling

G. T.

Finkelhor

Kirkpatric

J. T.

Straus

M. A.

(Eds.), Family abuse and its consequences: Violence in American families. New York, NY: Anchor/Doubleday.

53.

Teicher

M. H.

Samson

J. A.

Polcari

Greenery

C. E.

(2006). Sticks, stones, and hurtful words: Relative effects of various forms of childhood maltreatment. American Journal of Psychiatry, 163, 993-1000.

54.

Ullman

S. E.

(2007). Relationship to perpetrator, disclosure, social reactions, and PTSD symptoms in child sexual abuse survivors. Journal of Child Sexual Abuse, 16, 19-36.

55.

Weinberger

D. A.

(1990). The construct validity of the repressive coping style. In Singer

J. L.

(Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health (pp. 337-386). Chicago, IL: The University of Chicago Press.

56.

Williams

L. M.

(1994). Recall of childhood trauma: A prospective study of women’s memories of child sexual abuse. Journal of Consulting and Clinical Psychology, 62, 1167-1176.

57.

Wilson

S. C.

Barber

T. X.

(1983). The fantasy prone personality: Implications for understanding imagery, hypnosis, and parapsychological phenomena. In Sheikh

A. A.

(Ed.), Imagery: Current theory, research, and application (pp. 340-387). New York, NY: John Wiley & Sons.

Abuse Characteristics and Individual Differences Related to Disclosing Childhood Sexual,Physical,and Emotional Abuse and Witnessed Domestic Violence

Abstract

Keywords

Characteristics of Disclosure: Targets of Disclosure and Outcomes

Predictors of Disclosure

Abuse Characteristics: Victim Age, Abuse Frequency and Emotional Impact, and Perpetrator Identity

Individual Differences: Victim Self-Labeling and Defensive Coping Mechanisms

Summary of Hypotheses

Method

Participants

Measures

Demographics

Abusive experiences

Circumstances and outcomes of abuse disclosure

Abuse characteristics

Self-labeling

Repressive coping

Dissociative proneness

Fantasy proneness

Temporary forgetting

Procedure

Results

Characteristics of Abuse and Disclosure Circumstances

Predictors of Disclosure 1

Preliminary analyses

Intercorrelations among all predictor variables

Multivariate prediction of sexual abuse disclosure

Multivariate prediction of physical abuse disclosure

Multivariate prediction of emotional abuse disclosure

Multivariate prediction of the disclosure of witnessing domestic violence

Discussion

Predictors of Disclosure

Abuse characteristics

Individual differences

Strengths and Limitations

Conclusion

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

Notes

Author Biographies

References

Predictors of Disclosure¹