Does Encouragement by Others Increase Rape Reporting? Findings from a National Sample of Women

Abstract

Our study explores the role of victims' consultation with others about whether or not to report their rape to police. Three groups were observed within this sample of 435 rape victims from a national telephone household probability sample of women: those who did not consult with anyone about reporting (n = 364), those who consulted with someone and were encouraged to report to police (n = 40), and those who consulted with someone and were not encouraged to report (n = 31). Descriptive analyses indicated that the encouraged group was more likely to report to police than either of the other two groups (which did not differ from each other). Because there were no differences between the two consulting groups on demographic or rape-related variables, they were combined in subsequent analyses. Consulting with others about whether to report, peri-traumatic fear of injury or death, assault perpetration by a stranger, and concerns about contracting a sexually transmitted disease were significant predictors of reporting to police after controlling for other significant predictors in a multivariate regression analysis. Implications of these findings are discussed, including the benefits and consequences of formal rape reporting for victims, and the role that disclosure recipients may have in assisting victims post-rape (e.g., encouragement of reporting, emotional support).

Nearly one-in-five (18%) U.S. women report experiencing a completed rape – defined as unwanted oral, anal, or vaginal penetration without the victim's consent, including when she is unconscious or too drunk or high to know what she is doing or control her behavior – during their lifetime(Kilpatrick, Resnick, Ruggiero, Conoscenti, & McCauley, 2007). Notably, only a minority of these rapes is ever reported to the police. Wolitzky-Taylor and colleagues (2011) analyzed data from the National Women's Study-Replication (NWS-R) and found that 15.8% of women's only, or most recent, rape experiences were reported to the police, a rate that is similar to those found in other national surveys (Bachman et al., 1998; Chen & Ullman, 2010). This low reporting rate is especially troubling, given that formal reporting is the only means through which prosecution of perpetrators can occur, leaving the vast majority of perpetrators going undetected (Kilpatrick et al., 1992).

The reporting process necessarily entails a loss of anonymity and possible recrimination and stigmatization, making the reporting process potentially overwhelming to victims (Allen, 2007), and research has identified a number of factors that affect reporting behaviors. Results from national surveys show that the most common reasons given for not reporting a rape are that it is a private matter (Allen, 2007) and that victims fear reprisal (Allen, 2007; Bachman, 1998; Felson & Pare, 2005; Wolitzky-Taylor et al., 2011). Additional reasons include a fear of negative reactions from others, particularly blame (Heath, Lynch, Fritch, McArthur, & Smith, 2011; Kilpatrick et al., 1992; Wolitzky-Taylor et al., 2011), self-blame, and fear of embarrassment (Felson & Pare, 2005; Heath et al., 2011; Patterson, Greeson, & Campbell, 2009; Starzynski, Ullman, Filipas, & Townsend, 2005). Concerns about the legal system (e.g., belief the police would not help, not knowing how to file a report) have also been identified as reasons for not reporting (Allen, 2007; Felson & Pare, 2005; Patterson et al., 2009; Wolitzky-Taylor et al., 2011).

Characteristics of the rape also affect reporting behaviors. Victims who said they had experienced a stereotypical rape (i.e., a rape perpetrated by a stranger who used a weapon or physical force and caused injury; Patterson et al., 2009) were more likely to report the assault (Allen, 2007; Bachman 1998; Felson & Pare, 2005; Starzynski et al., 2005; Wolitzky-Taylor et al., 2011). Consistent with this pattern of behavior, some victims indicated that they did not report the assault because they were not sure that it constituted a rape or that they were deserving of post-assault assistance (Patterson et al., 2009). Finally, findings regarding demographic variables are mixed, showing differential effects of race (cf. Allen, 2007; Bachman, 1998; Felson & Pare, 2005; Wolitzky-Taylor et al., 2011), age (cf. Felson & Pare, 2005; Wolitzky-Taylor et al., 2011) and education on reporting behaviors (cf. Felson & Pare, 2005; Wolitzky-Taylor et al., 2011).

Although relatively few women formally report their rape, some data indicate that approximately two-thirds of rape victims may disclose their assault to someone else (Fisher, Daigle, Cullen, & Turner, 2003; Starzynski et al., 2005). These disclosures are most often made to an informal support source (e.g., friends, family: Ahrens, Campbell, Ternier-Thames, Wasco, & Sefl, 2007; Filipas & Ullman, 2001; Starzynski et al., 2005; Ullman, 1996a; 1999) and may be done in order to receive emotional support and tangible assistance (Ahrens et al., 2007). Rape victims often disclose to these informal support sources first (Ahrens et al., 2007).

In addition to the support that these individuals can provide, disclosure recipients can also play a critical role in encouraging formal rape reporting (Kilpatrick et al., 2007). Findings based on interviews with victims indicated that encouragement from others to report the rape was perceived as helpful by some in overcoming initial concerns about reporting (Patterson & Campbell, 2010). Research has shown that disclosure recipients' responses influenced whether or not crime victims contacted the police (Ruback, Greenberg, & Westcott, 1984). Similar results were also found with respect to adolescent sexual assault victims who continued their participation in the criminal justice system due to support from others (Campbell, Greeson, Bybee, Kennedy, & Patterson, 2011). Further, research shows that rape victims who felt that significant others (e.g., partner, friends) wanted them to report to the police were more likely to do so than those without this perceived social expectation (Feldman-Summers & Norris, 1984), and a national survey showed that victims with greater social support were more likely to report their rape (Allen, 2007). However, much of this research is limited by samples that were often small or non-representative of the general female U.S. population.

It appears that positive disclosure experiences may help to attenuate negative post-assault outcomes for victims (Allen, 2007; Resnick et al., 2000; Ullman, 2000; Ullman & Filipas, 2001a; Young, Bracken, Goddard, & Matheson, 1992) and that encouragement to formally report the rape may increase reporting behaviors. Thus, it is critical to identify factors associated with reporting to the police because this action has the potential to stop perpetrators from continuing to rape (Bachman, 1998; Kilpatrick et al., 1992; Wolitzky-Taylor et al., 2011). To this end, our study addresses a notable gap in the literature by assessing victims' interactions with others about reporting. More specifically, rape victims from a large U.S. national telephone household probability sample of women were assessed to determine the correlates of consulting with others about whether to report their rape to police. To clarify, the current report does not address broad disclosure of rape to others, but focuses on consultation with others about reporting to police specifically and is hereafter referred to as “consultation.” The terms “encouraged” and “non-encouraged” indicate whether women who consulted with others about reporting to police were encouraged or not encouraged to report.

Initial exploratory aims of our study included assessing patterns of consultation and responses to this interaction. Given the aforementioned mixed findings regarding demographic correlates of rape reporting, as well as limited information regarding correlates of consultation and/or subsequent encouragement to report, we did not make specific hypotheses about demographic variables and rape-related concerns. With respect to incident characteristics, first, we hypothesized that victims of stereotypical assaults (e.g., stranger perpetrator, sustained injury), as well as those who acknowledged the incident as rape, would be more likely to consult with others about reporting, consistent with previous research (Allen, 2007; Chen & Ullman, 2010; Starzynski et al., 2005; Ullman & Filipas, 2001b; Wolitzky-Taylor et al., 2011). Second, we hypothesized that, among consulting women, those who experienced assaults with stereotypical rape characteristics would be more likely to receive encouragement to report the assault compared to victims who did not experience stereotypical assaults. Third, given the role that disclosure recipients can play in influencing victims' decision to report the rape (Feldman-Summers & Norris, 1984; Kilpatrick et al., 2007; Patterson & Campbell, 2010), we hypothesized that consulting victims would engage in greater rates of reporting to the police than non-consulting victims after controlling for other predictors of reporting. With respect to the three groups of victims, it was expected that encouraged victims would be more likely to report their assaults than non-encouraged or non-consulting women; specific hypotheses about the reporting rates of these latter two groups were not made, given limited rationale for doing so.

Method

Participants and Procedure

Data were from the National Women's Study-Replication (NWS-R), a U.S. national telephone household probability sample of 3,001 women aged 18 to 76 years old (M = 42.89, SD = 15.81). The study comprised two national cross-section samples: 1,998 women aged 18-34 and 998 women aged 35 and older; five participants did not confirm their age. The samples were geographically stratified in order to account for population distribution and weighted to be consistent with 2006 U.S. Census estimates. Participants completed a 20-minute structured phone interview conducted by trained female interviewers at SRBI Inc. The procedure was approved by the Institutional Review Board at a major medical university. For a detailed description of the methodology, see previous research published using the NWS-R dataset (e.g., Kilpatrick et al., 2007; Wolitzky-Taylor et al., 2011).

A total of 435 women (14% of the sample of 3,001), who indicated a most recent rape incident at age 14 or older and who did not indicate that someone else reported the incident to the police, were asked a series of questions regarding their consultation about whether or not they should report the rape. Of these 435 participants in our final sample, the majority was White, non-Hispanic (n=330, 76%), followed by Black, non-Hispanic (n=66, 15%), Hispanic (n= 21, 5%), and Asian or Native American (n = 14, 3%), with 1% (n = 5) not specified or missing. Approximately half (n = 210, 48%) were married, 55% (n = 240) were high school graduates, and 33% (n = 145) were college graduates. Income distributions were as follows: 29% (n = 114) had an income greater than $60,000, 44% (n = 176) had an income $20-60,000, and 25% (n = 109) had an income less than $20,000.

Measures

Demographic information

Women were asked their current age (at the time of the interview), race/ethnicity (White, Black, Hispanic, other), marital status, highest level of education (some high school, high school graduate, college graduate, or graduate school degree), and estimated personal yearly income.

Rape experiences

Several questions were used to assess women's most recent or only rape. This section was prefaced with information about sexual violence, including that the victim did not have to tell anyone about the assault in order for it to be considered a rape, that the perpetrator could be anyone (e.g., family member, former partner, stranger), and that the assaults could have happened at any time during the participants' life. This preface was followed by close-ended behaviorally-specific questions to determine specific elements that would meet criteria for a rape incident, an approach that has been successfully used in numerous previous studies as a sensitive approach to assessment of rape in a previous national survey (Kilpatrick, Saunders, Best, & Von, 1987; see Resnick, Kilpatrick, Dansky, Saunders, & Best, 1993, for a description).

For the purpose of our study, and consistent with the age cut-off specified by others (e.g., Sexual Experiences Survey; Koss et al., 2007), rape was defined as penetration of the victim's vagina, mouth or rectum without consent at age 14 or older. Cases were defined as forcible rape if the victim indicated use of force, threat of force, or injury during the assault. Cases were defined as drug-or-alcohol-facilitated/incapacitated rape if the woman said she was intoxicated and incapacitated via voluntary or involuntary consumption of drugs and/or alcohol during the incident. Rape tactic categories were non-mutually exclusive, and cases that involved elements of both forcible and drug-or-alcohol-facilitated/incapacitated rape were coded as positive for both types of rape. Women were classified as having a history of multiple rapes if they endorsed an additional rape experience that occurred at any age (n = 213, 49% of victims).

Several rape incident characteristics were assessed for this most recent/only rape incident. First, relationship to the perpetrator was coded based on victims' description of their relationship to the assailant using the following categories: father or stepfather, other relative, husband/boyfriend or ex-husband/boyfriend, friend, other non-relative, or stranger. Stranger categorization was coded 1 (yes, a stranger) versus 0 (no) based on either the response to an initial close-ended question asking if the victim had ever seen the perpetrator before, or if the victim identified the perpetrator as a stranger in a later question. Second, peri-traumatic fear was assessed with the question “During this incident, were you afraid that you might be killed or seriously injured?” and affirmative responses were coded 1 (yes) or 0 (no). Third, injury was assessed with the question “Did you suffer serious physical injuries, minor injuries, or no physical injuries as a result of the incident?” and participants who endorsed minor or serious injuries were coded as 1 (yes) versus 0 (no).

Finally, whether the victim acknowledged the incident as a rape was assessed using the question “Looking back on what happened, which best describes how you feel about the incident? It was an unpleasant incident but not a crime; it was some type of crime but not a rape; it was a rape.” Fully 272 (63%) victims viewed the incident as a rape, 94 (21%) viewed it as a crime, but not rape, and 50 (11%) did not view it as a crime or rape; limited cell sizes prevented between-group comparisons beyond coding participants' acknowledgement of the incident as a rape or not (i.e., “viewed the rape as a crime, but not as rape” or “neither a rape nor a crime”), coded 1 (yes) versus 0 (no).

Post-rape concerns

Participants used a 4-point scale to rate their degree of concern regarding the following post-assault issues: (a) “getting a sexually transmitted disease, other than AIDS or HIV”;(b) “getting AIDS or HIV”;(c) “your family knowing that you had been assaulted”;(d) “persons outside your family knowing that you had been assaulted”;(e) “people thinking that it was your fault or that you were responsible”; and (f) “getting pregnant as a result of the assault.” Responses were dichotomized as 0 (not really or a little concerned) and 1 (somewhat or extremely concerned), consistent with previous research (Zinzow et al., 2012). For all but the STD- and HIV-concern variables, this approach was comparable to a median split with a range of noted concern (i.e., coded as 1) from 47% (n = 196) for pregnancy to 58% (n = 247) for concerns about others thinking they were responsible for the assault. A relatively lower percentage noted concern regarding STDs other than HIV and HIV (n= 175, 40% and n= 159, 36%, respectively). The post-rape concern items demonstrated good reliability, with Cronbach's alpha of .71 for the six dichotomously coded items. Dichotomously coded items were consistent with the bimodal distribution of these data.

Consultation about reporting

Participants were asked, “Before deciding whether or not to report, did you consult with anyone about the possibility of reporting this incident to police?” Participants who replied “yes” were asked “Who did you discuss this with?” and “Did this person (any of these persons) encourage you to report the incident to the police?”The non-consulting group was coded as “0,” the non-encouraged group was coded “1,” and the encouraged group was coded “2.” Participants were also asked “Did you or someone else report the incident to the police, or was it not reported to the police?” Participants who reported to police themselves were coded as “1,” those who did not report were coded as “0,” and those who indicated that the incident was reported by someone else were excluded from the analyses.

Data Analyses

First, descriptive analyses are provided in order to characterize the frequency of consultation, encouragement, and reporting among participants. Second, correlations between potential predictor variables were conducted in order to provide further descriptive information regarding these behaviors. Third, we conducted preliminary analyses to determine if there were differences on demographic and rape-related variables (i.e., assault characteristics, rape-related concerns) between consulting (i.e., encouraged and non-encouraged) women. Given that there were no differences between these groups, they were combined into a single consulting group for subsequent analyses. Next, study hypotheses were tested using univariate and multivariate analyses to compare all consulting women with non-consulting women on demographic and rape-related variables, as well as actual reporting to police. Finally, descriptive chi-square analyses were conducted to evaluate whether there were overall and/or specific group differences between the three groups of interest (i.e., non-consulting, encouraged, and non-encouraged) in prevalence of reporting to police.

Results

Consultation, Encouragement, and Reporting

Descriptive statistics regarding consultation, encouragement of reporting (among those who consulted), and reporting behavior are presented in Table 1. Women were considered to be a member of one of the three following groups: (a) victims who did not consult with anyone about reporting (i.e., non-consulting women), (b) victims who consulted with others and were encouraged to report (i.e., encouraged women), and (c) victims who consulted with others and were not encouraged to report (i.e., non-encouraged women). In general, women did not consult with others about reporting, and the vast majority of non-consulting women did not report their rapes. Among those who did consult, approximately half were encouraged to report, and these encouraged women engaged in substantially higher rates of actual reporting. Of the 16.5% (n = 72) of victims who consulted with someone else, participants were most likely to consult with a friend (48%, n = 35), followed by a parent (24%, n = 17), other family member (14%, n = 10), or husband, boyfriend, or partner (8%, n = 6).

Table 1. Consultation, Encouragement, and Reporting (N = 435).

	Non-Consulting	Consulting
Total n	364	71
%	83.5%	16.5%
		Encouraged	Non-Encouraged
n		40	31
% of Total		9.3%	7.2%
% of Consulting		56.4%	43.6%
Reporting
n	32	23	5
% of Total	7.2%	5.2%	1.1%
% of Non/Consulting	8.7%
% of Not/Encouraged		56.2%	14.7%

Note. Due to weighting, numbers within some sub-groups may vary slightly for some comparisons.

Associations among Potential Predictor Variables

Pearson's correlations, point-biserial correlations, and chi-square analyses were used to examine the relations among potential predictor variables. Acknowledgment of the assault as a rape was positively associated with sustained injury, peri-traumatic fear, forcible rape, drug- and alcohol-facilitated and incapacitated rape, multiple rape history, concern about contracting an STD and HIV, concern about becoming pregnant as a result of the rape, consultation with others about reporting, and actual reporting behaviors (see Table 2). Actual reporting was negatively associated with age at the time of the interview and positively associated with sustained injury, peri-traumatic fear, forcible rape, rape perpetrated by a stranger, concerns about contracting an STD and HIV, concerns about non-family members finding out about the rape, and consultation with others about reporting. There were also significant differences in acknowledgement with respect to education. Individuals with some high school education (74%) reported the greatest level of acknowledgement, followed by high school graduates (70%) and graduate school graduates (66%); the lowest rate of acknowledgement was among college graduates (53%), χ²(df = 3, n = 416) = 11.26, ϕ = .16, p = .01.

Table 2. Correlations among Study Variables (N = 435).

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Age	---
2. Injury	.14**	---
3. Fear	.12*	.39**	---
4. FR	.27**	.37**	.43**	---
5. DAFR/IR	-.37**	.24**	.40**	.65**	---
6. Stranger perpetrator	-.08	-.02	.07	-.02	.11*	---
7. Age at rape	.36**	.09	.12*	.10*	-.08	.02	---
8. Acknowledgment	.09	.41**	.43**	.36**	.30**	-.03	-.07	---
9. Multiple rape history	-.06	.24**	.20**	.17**	-.07	.03	.11*	.19**	---
10. STD concerns	-.39**	.09	.21**	.01	.10*	.11*	-.06	.19**	.19**	---
11. HIV concerns	-.26**	.14**	.19**	.03	.10*	.10*	.11*	.16**	.25**	.67**	---
12. Family finding out	-.14**	.17**	.34**	.13**	-.04	.05	-.09	.09	.18**	.19**	.22**	---
13. Non-family finding out	-.24**	.12*	.21**	.05	.03	.13**	-.23**	.08	.13**	.21**	.15**	.58**	---
14. Blame from others	-.16**	.08	.05	.08	.08	.09	-.16**	.03	.12*	.11*	.08	.38**	.51**	---
15. Pregnancy concerns	-.05	.19**	.20**	.01	.05	.10*	-.21**	.20**	.18**	.26**	.22**	.27**	.31**	.25**	---
16. Consultation	-.26**	.10*	.09	.02	.06	.03	-.10*	.14**	.02	.15**	.10*	.08	.22**	.10*	.03	---
17. Reported to police	-.14**	.18**	.25**	.13**	-.07	.16**	.03	.20**	.07	.23**	.23**	.05	.15**	-.07	.04	.32**

Note. FR = forcible rape; DAFR/IR = drug- and alcohol-facilitated and incapacitated rape. Consultation = consulted with someone about reporting most recent/only incident to police.

^*p < .05.

^**p < .01.

Encouraged versus Non-Encouraged Women

Our second study hypothesis predicted differences between encouraged versus non-encouraged consulting women. To address this hypothesis, and to also potentially simplify the analytic strategy, preliminary chi-square analyses were conducted to determine whether these groups might be combined in subsequent analyses examining predictors of consultation group category (see Table 3). There were no differences between these groups with regard to demographic or rape-related variables (i.e., rape characteristic and post-rape concerns), with a Bonferroni correction applied (p = .02). Thus, the second hypothesis was not supported, and these groups were combined to form a single group (i.e., consulting women) for the purpose of subsequent analyses.

Table 3. Comparisons between Encouraged and Non-Encouraged Victims (n = 71).

	Non-encouraged % (n) or M (SD)	Encouraged % (n) or M (SD)	χ2 or F	p
Age	35.21 (10.14)	32.76 (9.69)	1.07	.30
Rape characteristics
Injury	43.8% (14)	67.5% (27)	4.09	.04
Peri-traumatic fear	57.1% (16)	66.7% (26)	0.63	.43
Forcible rape	83.9% (26)	90.0% (36)	0.59	.44
DAFR/IR	38.7% (12)	30.0% (12)	0.59	.44
Acknowledgement as rape	77.4% (24)	80.5% (33)	0.10	.75
Multiple rape history	41.9% (13)	58.5% (24)	1.39	.24
Post-rape concerns
Contracting HIV	41.9% (13)	52.5% (21)	0.78	.38
Contracting STD	58.1% (18)	56.1% (23)	.028	.87
Family finding out	66.7% (20)	62.5% (25)	0.13	.72
Non-family member finding out	71.0% (22)	80.0% (32)	0.78	.38
Blame from others	77.4%(24)	62.5% (25)	1.82	.18
Pregnancy concerns	50.0% (16)	48.8%(20)	0.01	.92

Non-Consulting versus Consulting Groups

Demographics and rape-related variables

To test our first hypothesis (that consulting women would be more likely than non-consulting women to have both experienced rapes with stereotypical characteristics and acknowledge their assaults as rape), univariate and multivariate logistic regression analyses were used to examine differences between the groups on demographic characteristics and rape-related variables. Results from univariate analyses indicated that consulting women were younger at the time of the interview and at the time of the rape compared to non-consulting women (see Table 4). Consistent with the first hypothesis, consulting women were more likely than non-consulting women to report peri-traumatic injury (a stereotypical rape characteristic) and to acknowledge the incident as a rape. They were also more likely to report concerns about contracting a STD and HIV, non-family members finding out about the rape, and being blamed by others. Aside from sustained injury, other stereotypical rape characteristics (i.e., perpetration of the rape by a stranger, forcible rape) did not differ between these groups.

Table 4. Results of Logistic Regression Analysis Predicting Rape Victims' Consultation with Someone about Reporting to Police (N =435).

	Non-Consulting	Consulting	Univariate		Final Model
Predictors	%	%	OR	95% CI	OR	95% CI
Age (M(SD)	44.69(15.83)	33.83(9.90)	F=31.26***		0.95***	0.92-0.97
Age at rape (M(SD)	22.36 (8.65)	20.18(6.34)	F = 3.99*		1.00	0.96-1.05
Injury	43	57	1.74*	1.04-2.90	1.55	0.84-2.83
Peri-traumatic fear	50	63	1.64	0.96-2.80
Forcible rape	85	87	1.16	0.55-2.45
DAFR/IR	28	35	1.39	0.81-2.38
Stranger perpetrator	11	14	1.31	0.63-2.74
Acknowledged as rape	62	80	2.37**	1.28-4.39	2.35*	1.16-4.79
Multiple rape history	48	51	1.11	0.67-1.85
STD concerns	37	57	2.25**	1.34-3.76	0.95	0.52-1.76
HIV concerns	35	49	1.72*	1.02-2.88
Concerned family finding out	53	64	1.60	0.94-2.72
Concerned non-family finding out	47	76	3.63***	2.02-6.53	2.34*	1.17-4.67
Concerned blame from others	56	69	1.72*	1.00-2.97	0.96	0.49-1.86
Pregnancy concerns	46	50	1.17	0.70-1.94

Note. Reference group is non-consulting women. OR = odds ratio; CI = confidence interval; FR = forcible rape; DAFR/IR = drug- and alcohol-facilitated and incapacitated rape.

^*p < .05,

^**p < .01,

^***p < .001

Variables that were significant at the univariate level were included in a final multivariate model. HIV concerns were not included due to multicollinearity with the STD concerns variable (r = .67, p < .01); STD concerns were retained because they represent a broader construct and were more strongly related to the variable of consultation. Results from the multivariate logistic regression analysis indicated that, compared to non-consulting women, consulting women were younger in age at the time of the interview, more likely to acknowledge the incident as a rape, and more concerned about non-family members finding out about the rape (see Table 4).

Predicting reporting

In order to assess the third hypothesis and determine if consultation served as a unique predictor of reporting, a multivariate logistic regression was used to examine the influence of consultation on actual reporting behavior, while controlling for study variables that were significantly correlated with reporting (see Table 5). Peri-traumatic injury was not included in the model because injury was used as a criterion for defining forcible rape—a variable that was included in this model. Results indicated that consultation was a predictor of reporting to the police after taking other related variables into account, consistent with the third hypothesis. More specifically, non-consulting women were less likely than consulting women to report the rape to police (OR = 5.23, p < .001). Additional predictors of reporting included peri-traumatic fear, perpetration of the assault by a stranger, and concerns about contracting a STD.

Table 5. Logistic Regression Analysis Predicting Rape Victims' Reporting Behavior and Encouragement of Other Victims to Report to Police.

	Reported to Police (N = 435)
Predictors	OR	95% CI
Age	0.99	0.96-1.02
Peri-traumatic fear	3.29**	1.37-7.90
Forcible rape	2.08	0.33-13.30
Stranger perpetrator	2.99**	1.36-6.57
Acknowledgment	2.26	0.87-5.90
STD concerns	2.19*	1.06-4.52
Non-family finding out concern	0.94	0.46-1.90
Consultation	5.23***	2.53-10.80

Note. Consultation = consulted with someone about reporting most recent/only incident to police. OR = odds ratio; CI = confidence interval

^*p < .05.

^**p < .01.

^***p < .001.

In order to further assess our third hypothesis and explicate the significant relation between consulting and reporting, we conducted post-hoc chi square analyses that compared all three study groups (i.e., non-consulting, encouraged, non-encouraged). These analyses indicated a difference among groups with respect to reporting, χ²(df, = 2N = 435) = 69.73, ϕ = .40, p < .001. Follow-up group comparisons revealed that encouraged women were more likely to report their rapes to the police (56.2%) than both non-consulting (8.7%), χ²(df, = 1, n = 405) = 70.27, ϕ = .42, p < .001, and non-encouraged women (14.7%), χ²(df, = 1, n = 73) = 12.45, ϕ = .41, p < .001. Non-consulting and non-encouraged women did not differ in prevalence of reporting, χ²(df, = 1, n = 396) = 1.62, ϕ = .06, p = .17.

Discussion

When do Women Report their Rapes?

The results of our study indicate that there may be differences in decisions about whether or not to report a rape that are influenced by the reactions of others. Consultation with someone else about reporting the rape was a significant predictor of actual reporting behavior, even after controlling for other predictors of reporting, consistent with our third hypothesis. A final set of descriptive analyses indicated that women who were encouraged were the most likely to report their rapes, whereas non-encouraged and non-consulting women did not differ in their reporting rates. Importantly, less than 10% of non-consulting women reported their rapes, whereas slightly more than half of encouraged women reported to the police. Taken together with the finding that consulting itself predicts reporting behaviors, these findings underscore the influence that others may have on victims' post-rape behaviors. Research should further assess the reasons that victims choose to consult with others about whether or not to report, as well as whether such individuals had already considered reporting, and whether recommendations by others influence victims' actual reporting behaviors. Such information will help to clarify the potential impact of consultation with, and recommendations by, others in victims' decision-making processes about formal rape reporting.

Additionally, women who reported fear that they might be killed or seriously injured during the rape, and those whose rapes were perpetrated by a stranger, were more likely to report their rapes. These findings are consistent with others from a previous report looking at all women in the NWS-R sample regardless of age at the time of the assault (Wolitzky-Taylor et al., 2011), as well as previous research (Allen, 2007), underscoring the influence of stereotypical rape characteristics on post-rape behaviors. Finally, concern about contracting an STD was also associated with reporting, consistent with the aforementioned findings about consultation. Given that participants with this concern are more likely to seek out medical services post-rape (Zinzow et al., 2012), it is possible that they are also more likely to engage with other agencies post-rape including police.

When do Women Consult about Reporting?

As discussed above, consultation with others was uniquely associated with greater rates of formal rape reporting. Unfortunately, the vast majority of women (83%) chose to not consult with someone about reporting and very few women who did not consult with someone else formally reported their rape (9%). Given that as many as two-thirds of victims in other studies have been observed to disclose their assault to someone else (Fisher et al., 2003; Starzynski et al., 2005), these findings are of great interest. It may be that victims are not intentionally seeking out assistance from others when deciding whether or not to report to police (i.e., the behavior assessed in our study), but may instead be looking for other forms of assistance, like emotional support and medical care (Ahrens et al., 2007). Although these reasons may be affected by the passage of time (e.g., consulting about reporting in the immediate aftermath of the assault vs. seeking emotional support months or years later), the influence of time until disclosure and consultation was not assessed in our study.

Consistent with the first hypothesis and the literature on rape reporting, women who experienced assaults with a stereotypical rape characteristic (i.e., injury) and women who acknowledged the rape were more likely to consult with another about rape reporting (Campbell et al., 2001; Starzynski et al., 2005), although not all stereotypical characteristics were associated with consultation. It is possible that injury was the most influential of these characteristics given related post-rape needs, like medical attention, that could best be assessed and accessed by talking to another. If the sustained injuries were observable by others, this circumstance might also increase the likelihood of discussion about the assault with others, including formal reporting.

Interestingly, a variable that was related to these stereotypical characteristics in this study, acknowledgment of the rape, was the only variable that remained significant in the multivariate analysis. Although injury no longer remained significant in this model, it is notable that women who sustained an injury were one-and-a-half times as likely to consult with another about reporting. Taken together, these findings may be best explained by Allen (2007), who proposed that a stereotypical rape characteristic (e.g., injury) corroborates that the rape occurred and serves as evidence for both the victim and those to whom the victim discloses. It is feasible then that these stereotypical features alone increase the likelihood of, but do not directly lead to, consultation. It is also possible that acknowledgement mediates the relation between injury and consultation. As suggested above, because acknowledgement was positively associated with a number of stereotypical characteristics (i.e., forcible rape, sustained injury), it appears that rape-related characteristics play an important role in victims' perceptions of the assault, as well as their post-assault behaviors, consistent with previous research (Koss, 1985; Littleton, Rhatigan, & Axsom, 2007).

Interestingly, concern about non-family members finding out about the rape was also strongly associated with consultation. It is reasonable to assume that individuals who are considering reporting the assault to police may have realistic concerns that others will find out about the assault. Those victims who are not thinking about reporting would likely have fewer concerns about others learning of the assault, a possibility raised in previous research (Cohn, Zinzow, Resnick, & Kilpatrick, 2012). This interpretation may also indicate that those who consult with others may already be more seriously considering reporting prior to consultation with others. Taken together, the relations between these post-rape concerns and consultation about reporting are congruent with previous research emphasizing victims' concerns about confidentiality and privacy (Allen, 2007).

With respect to age at the time of the interview, it is possible that the difference between consulting and non-consulting women may reflect cohort effects of the social networks among younger and older women (e.g., concerns about STDs and HIV) and different ways in which these groups may respond to a rape (e.g., active seeking of support vs. personal reflection about options). Further, the inverse relation between age and concern about others finding out about the rape in our study may reflect continued fears of stigma and blame regarding rape (Ahrens, 2006; Miller, Canales, Amacker, Backstrom, & Gidycz, 2011), although this finding may also be due to an age-related cohort effect (e.g., younger women's concern about their parents finding out about the assault). Given that disclosure and consultation with others can happen at any point after the assault, these potential cohort effects should be assessed in future research.

When are Women Encouraged to Report?

Contrary to our second hypothesis, there were not significant differences in encouragement of consulting women with respect to rape characteristics or acknowledgement of the assault as a rape. Of the 17% of women who consulted with someone else about reporting the assault, approximately even numbers were encouraged to report (56%) and not encouraged to report (44%), indicating that there may be some unmeasured variables that affect recipients' encouragement of reporting (e.g., relationship to the victim, victim presentation). Of note, as discussed above, consulting women experienced more stereotypical assaults and reported greater post-rape concerns than non-consulting women. Thus, consulting women, regardless of whether or not they received encouragement to report from others, may be a more homogeneous group than non-consulting women. Further research should assess a range of potential predictors, as well as learn more from recipients about what factors affect the responses of others to victims. Due to small sample sizes for comparisons, the analyses may also have been underpowered, and future research should aim to further assess those variables associated with encouragement of reporting within larger samples.

Limitations and Future Directions

Despite the important information gained from our study, there are limitations to the conclusions that may be drawn from it. First, all data collected in our study were self-report, retrospective, and cross-sectional. These issues are particularly relevant to some of the analyses conducted because participants gave information about both current characteristics (e.g., demographics) and historical experiences (e.g., concerns about post-rape consequences), and memory for, and accuracy of, these domains may be affected (e.g., greater level of concern after a non-encouraging response to consultation). This manner of assessment may have affected the results obtained, and future research should investigate whether assessment of victim consultation and reporting decisions shortly after the rape results in a different pattern of findings. Given that such research is difficult to do among nationally representative U.S. populations, potential sample differences should be considered in studies among more circumscribed populations.

The obtained rape prevalence within our study sample was similar to that found in another recent national study (Black et al., 2011), supporting the validity of the assessment of rape and key rape characteristics used in the current study. However, there is not a similar point of comparison for consulting with someone else about reporting a rape to police specifically, the endorsement of which was notably low in our sample. A potential measurement limitation related to consultation with others about reporting to police, as well as respondents' reactions to disclosure of rape by other victims, was the use of single-item questions to assess these constructs. A more detailed assessment of these variables using multiple questions (e.g., the situation in which they consulted with someone else, what the victim talked to others about) would have provided further information, and we recommend the development and evaluation of the validity of single-item as compared to multiple-item measures of these constructs in future research.

On a related note, more detailed assessment of actual encouragement to report (e.g., mentioned reporting as an option, took the victim to the police station) may reveal differences between encouraged and non-encouraged consulting women with respect to assault- and victim-related variables. Future research aimed at the development and evaluation of measures related to responses specific to reporting to police using multiple questions would also be useful, as would comparison of such measures with single-item approaches. Third, only those participants who reported an assault at age 14 or older were included in this study; a broader range of age at the time of victimization may lead to different results. Finally, participants had to have a landline phone number; however, over 85% of US households had landlines during the study period (Blumberg & Luke, 2007), so the impact of this exclusionary criterion is limited.

Practice Implications

The prevalence of reporting to the police in this sample (14%) was consistent with that found in other national surveys (Bachman et al., 1998; Chen & Ullman, 2010). As described above, this low prevalence of reporting is unfortunate, as it is the only way to identify perpetrators and bring them to trial (Kilpatrick et al., 1992). Yet, formal reporting may not be a desired option for all victims. There are potential negative consequences related to reporting, including loss of anonymity, fear of, or experienced, reprisal or pressure to not participate in the criminal justice system by the perpetrator or others, and possible exposure to other potentially harmful negative reactions (e.g., Allen, 2007; Bachman 1998; Felson & Pare, 2005; Patterson & Campbell, 2010; Patterson et al., 2009; Wolitzky-Taylor et al., 2011). Thus, formal reporting cannot be assumed to be an inherent part of the post-rape process, and the focus should be on helping victims to make an informed decision about whether or not to file a report.

Importantly, disclosure recipients can play an important role in helping to support victims as they consider their post-rape actions. Consistent with previous research, these results show that consultation itself, particularly that which results in encouragement from others to report, is associated with higher rates of reporting (e.g., Feldman-Summers & Norris, 1984; Kilpatrick et al., 2007; Patterson & Campbell, 2010), underscoring the role that disclosure recipients can play in this decision. Increasing recipients' and victims' knowledge about what is involved in filing a formal report (e.g., general procedure, steps in the legal process) could help in better informing any choice that is made in this regard. This information should include awareness of potential negative consequences of reporting so that steps can be taken to either avoid, or minimize, their effects, if possible (e.g., greater attention to potential safety concerns, increased involvement with supportive others). Finally, despite any influence that recipients have with respect to victims' reporting behaviors, the final decision about whether or not to pursue a report should be that of the victim, and that choice should be respected by those around her.

Regardless of whether or not the victim chooses to report, what is truly important is that she receives a validating response to her disclosure and is provided with appropriate assistance and support. Thus, the disclosure recipient may also assist the victim by providing him or her with information about a range of resources in addition to formal reporting (e.g., available medical services, irrespective of reporting behavior; Price, 2010), as well as other community agencies, including rape crisis centers and victim advocates. These other agencies may help to provide the aforementioned information, as well as connections to other resources and services. With this level of support and knowledge, the victim can then make an informed decision about how she would like to proceed and the disclosure recipient may support that choice. Although efforts such as these have begun (e.g., Start by Believing Campaign; End Violence Against Women International, 2012), much can still be done to educate and inform the public so that they may best support victims.

Conclusion

These findings add to the body of work attesting to the role of others in helping victims to recover post-rape, specifically with respect to increasing formal reporting through consultation about the act, as well as encouragement by others to do so. Although more work is needed to further elucidate the role of disclosure recipients in best assisting victims and improving post-assault outcomes, results suggest that the role they play is an important one. Moving forward, it is important to clarify how disclosure recipients influence victims' post-rape behaviors and outcomes, as well as how they can best support them after the rape. By working with recipients to best assist victims, it may be possible to increase the amount of information, resources, and support available to victims, and hopefully improve their post-assault outcomes.