Abstract
Although victim blaming in the context of sexual assault is often emphasized, little research has compared rates of victim blaming following sexual assault relative to other forms of victimization. This research investigated whether there is a crime-specific bias toward blaming victims of sexual assault. Victim blaming was assessed via different methods from the observer perspective in vignette-based studies, as well as survivors’ accounts of social reactions they received. In Study 1, participants were asked to rate how much the survivor was to blame in three vignettes, each with a different randomized crime outcome: rape, physical assault, or theft. Study 2 assessed blame for a vignette that either ended in rape or theft, via a causal attribution statement. Study 3 asked interpersonal trauma survivors who had experienced at least two forms of victimization (i.e., sexual assault, physical assault, or theft) to report the social reactions they received following disclosure of each of these crimes. Across all three studies, victim blaming occurred following multiple forms of victimization and there was no evidence of a particular bias toward blaming survivors of sexual assault more so than other crimes. However, results of Study 3 highlight that, following sexual assault, survivors receive more silencing and stigmatizing reactions than they experienced after other crimes. Interpersonal traumas (i.e., sexual or physical assault) also resulted in more egocentric responses compared to theft. Altogether, there does not appear to be a crime-specific bias for victim blaming; however, crime-specific bias is apparent for some other, potentially understudied, social reactions. Implications of these findings highlight the value of victim blaming education and prevention efforts through trauma-informed services and outreach following victimization. Furthermore, service providers and advocates might especially seek to recognize and prevent silencing and stigmatizing reactions following sexual assault disclosures.
Introdcution
Although 90% of people will experience at least one traumatic event in their lifetime, a small fraction of these individuals will develop psychopathology related to the event (Kilpatrick et al., 2013). There are a variety of risk factors for the development of psychopathology following trauma exposure (Brewin et al., 2000). One risk factor is negative social reactions from professional or personal others when the survivor discloses their experience (Dworkin et al., 2019). These negative social reactions include stigmatization, egocentricity, distraction/discouragement from talking (which we term “silencing”), turning against, or controlling, as well as victim blaming (Ullman, 2000). Victim blaming is the communication of perceived fault or responsibility to the survivor either for the traumatic event itself or for failing to prevent it from happening and may be especially pernicious for its potential to be internalized as self-blame by the survivor (Bonnan-White et al., 2015; Dworkin et al., 2019). Another risk factor is the type of event itself. Specifically, compared to noninterpersonal traumas (e.g., motor vehicle accident, natural disaster), negative psychopathology outcomes are more common following interpersonal traumas, which are those that involve an aggressor (e.g., physical assault; Breslau et al., 1998; Resnick et al., 1993), particularly when the trauma is sexual (Kessler et al., 1995). Furthermore, self-blame is a symptom of post-traumatic stress disorder (American Psychiatric Association, 2013) and both victim blame and self-blame are cited by sexual violence survivors as major deterrents from reporting the crime to the police (Reich et al., 2021).
Considerable attention has been afforded to victim blaming specifically for sexual assault survivors in past research (Suarez & Gadalla, 2010), advocacy efforts (RAINN, n.d.), and in the media more broadly (Deutsch, 2018; Heldman & Dirks, 2013). For example, researchers have identified a plethora of contextual factors related to the survivor (e.g., alcohol consumption, gender) and the perpetrator (e.g., socioeconomic status, attractiveness), each of which impact perceptions of blame following sexual assault (Grubb & Turner, 2012; Penone & Spaccatini, 2019; Van Der Bruggen & Grubb, 2014). Likewise, rape myths, which are “false cultural beliefs that mainly serve the purpose of shifting the blame from perpetrators to victims” have also been studied extensively (see meta-analysis Suarez & Gadalla, 2010). This focus on victim blaming in the context of sexual violence may signal an underlying assumption that sexual assault survivors suffer more victim-blaming attitudes than victims of other crimes. However, empirical comparisons of negative social reactions across different types of victimization experiences have been rare (Felson & Palmore, 2018). Understanding how pervasive victim blaming is following sexual violence relative to other crimes could have important implications for how service training and prevention efforts are targeted. If there was a crime-specific bias to victim blame sexual assault survivors, this might be further investigated in psychopathology research. Specifically, it could be an explanatory pathway for the poorer negative mental health outcomes experienced by sexual violence survivors relative to survivors of other traumatic events (Kessler et al., 1995), and might signal targeting of survivors’ internalized messages from others as especially important in treatment following sexual violence.
The purpose of the current study was to examine whether negative social reactions, and especially victim blaming reactions, are more common following sexual trauma relative to other types of crimes. Study 1 and Study 2 of this research present blaming reactions from observers of vignettes across different crime types. Study 3 compares survivors’ experiences of victim blame, and other negative social reactions, when disclosing sexual and nonsexual crime experiences to others.
Study 1
Although vignette-based research of victim blaming is common, few studies have compared victim blaming across different crime types. An exception is the work of Bieneck and Krahe (2011). They observed greater blame attributed to rape survivors than robbery victims, especially when the victim was too intoxicated to resist or the victim knew the perpetrator. Their within subjects design, unfortunately, confounded the crime outcome with the scenario. Specifically, they presented three rape scenarios and three completely different theft scenarios, meaning other discrepancies in the rape or theft scenarios might have accounted for the differences in blame attributions besides the crime outcome. For example, some differences between their rape and theft vignettes include whether there was mention of camera evidence, the perpetrator’s denial of the crime, and the length of the vignettes.
All other studies on this topic appear to employ between-subjects designs in which participants only view one vignette with a randomized outcome. These studies have all failed to find a bias toward blaming rape survivors more than victims of theft (Brems & Wagner, 1994; Felson & Palmore, 2018; Kanekar et al., 1985) or other nonsexual violence (Felson & Palmore, 2018; Koepke et al., 2014). In fact, these studies sometimes indicate a bias in the opposite direction, such that sexual violence survivors appear to receive less blame than other crime victims. Notably, between-subjects designs indicate group-level bias. However, the question of whether there is a bias toward blaming one victim versus another might be better understood as an individual-level bias in the form of intrapersonal discrepancy best observed with a within-subjects design (Crawford & Popp, 2003).
Previous findings may also reflect other methodological decisions. For example, the existing literature has utilized college student samples. This sampling method may be concerning given that rape myths are less accepted among individuals with higher education levels (Cohen’s d effect size = –0.57; Suarez & Gadalla, 2010) and college samples may be different from community samples in other attitudes toward sexual assault (Keller & Wiener, 2011). For example, Felson and Palmore (2018) utilized a sample of students enrolled in criminology courses, courses that might attract students with less victim-blaming attitudes to start with. Criminology course content might also be expected to impact victim blaming attitudes (Fox & Cook, 2011). Furthermore, undergraduate institutions often host sexual assault awareness trainings and programming (Amar et al., 2014; see also U.S. Violence Against Women Act).
Beyond these sampling and design questions, a number of relevant participant factors are unknown from these vignette-based studies. For example, past research suggests rape survivors are more likely to be blamed when the perceiver also endorses more rape myth acceptance (Hayes et al., 2013 also see meta-analysis Suarez & Gadalla, 2010), benevolent sexism (e.g., the idea that women are purer and need to be protected by men; Abrams et al., 2003), and have not experienced a trauma themselves (Grubb & Harrower, 2008; Hayes et al., 2013). Finally, victim blame research could be susceptible to socially desirable responding. Therefore, victim-blaming attitudes toward sexual assault survivors relative to other crime victims might be obscured by perceiver factors unaccounted for in the previous research.
The primary aim of Study 1 was to test whether the Bieneck and Krahe (2011) finding of crime bias in blame following sexual assault would replicate when the scenario was held constant. To compare intraindividual differences in blaming attitudes toward survivors of different crime outcomes, all participants read the same three vignette scenarios and all three crime outcomes—sexual crime (i.e., rape), a violent nonsexual crime (i.e., physical assault), or a nonviolent crime (i.e., theft)—but the pairing of outcome and scenarios was randomized. We additionally explored whether findings varied as a function of participant factors including rape myth acceptance, sexism, socially desirable responding, and trauma/crime exposure.
Method
Participants
Participants were recruited and paid a small amount for participation in a study on crime perceptions through TurkPrime, a crowdsourcing platform variant of Amazon’s Mechanical Turk for academic research recruitment, commonly used for social science research (Buhrmester et al., 2018). Four participants failed the attention check and were excluded from the sample leaving a final sample of 97 participants. All participants were over the age of 18 and resided in the United States. The sample was predominantly male (70.1%) and most of the participants were White or European-American (82.5%), followed by Black or African American (9.3%), Asian or Pacific Islander (8.2%), Hispanic and/or Latino (6.2%), and Native American or Alaska Native (1.0%). The mean age of the sample was 37.90 years (SD = 12.01) and participants had an average of 15.55 years of education (SD = 2.37), with a range of 12-25 years of education. Personal exposure to criminal victimization was as follows: rape (16.5%), physical assault (54.6%), and theft (59.8%).
Procedure
All procedures were approved by the university’s institutional review board. Following informed consent, participants were presented with three vignettes with a randomized outcome: rape, physical assault, or theft. The design was such that each participant viewed all three vignette scenarios and all three criminal outcomes, but the scenario-outcome pairings were randomized. After reading each vignette, participants completed a blame measure. After all three sets of vignettes and measures were completed, participants then completed measures on socially desirable responding, benevolent sexism, rape myth acceptance, and interpersonal violence and theft exposure.
Materials
Vignettes. Because most vignette research utilizes only one vignette (Kanekar et al., 1985) or the vignettes are designed for a specific crime outcome (Bieneck & Krahe, 2011), new vignettes were created for this within-subjects design study. The vignettes were intentionally similar in several ways including that they each involved a single perpetrator who was a male acquaintance of the female victim without mention of alcohol. These factors (gender, number of perpetrators, degree of acquaintance, and absence of alcohol) were held constant across the three vignettes due to past findings that these impact the degree of blame ascribed to victims (Ayala et al., 2018; Bieneck & Krahe, 2011; Sleath & Bull, 2010; Stepanova & Brown, 2017). The vignettes were also similar in length (between 185 and 187 words) and the victim and perpetrator names for each vignette were based on equivalently popular American names (Social Security, n.d.). To allow for victim blaming judgments, each vignette also included six details about the victim that might in hindsight be attended to for each of the three outcomes: conspicuous possessions (e.g., she asked him to hold her purse), flirtation (e.g., moved closer to the perpetrator while dancing), and verbal conflict (e.g., arguing about local politics). Neutral details about the perpetrator (e.g., he played baseball in the summer) were included so that the number of details about the victim and the perpetrator were equivalent. In all vignettes, the perpetrator asks to socially engage with the victim in a neutral way (i.e., studying together), the victim agrees to the neutral social situation, and the victimization occurs when they are alone together. Similar to other victim-blaming studies (Maurer & Robinson, 2008), we described the crime outcomes in behavioral terms since the failure to label nonconsensual sex as rape in and of itself is considered a form of victim blame (Lonsway & Fitzgerald, 1994). Pilot data revealed that when reading the vignette stories alone, participants predicted all three crime outcomes as a plausible next event. A sample vignette is available in the Appendix.
Measures
Blame attribution questions. Blame attributions toward the victim and perpetrator were measured via an eight-item questionnaire (Adolfsson et al., 2017). An example of one of the questions that assessed blame attributed to the victim is “To what extent is she to blame?” Participants responded to each item on a 7-point Likert scale ranging from 1 (not at all) to 7 (very much). Blame attributed to the victim and perpetrator was assessed by averaging the score of the items the victim blame attribution items and the perpetrator-blame attribution items separately such that a higher score indicated a higher level of blame. The reliability for the Victim Blame and Perpetrator Blame subscales was good for this sample, Cronbach’s α = .82 and α = .86 respectively.
Social desirability. Socially desirable responding was measured via 33 true/false items (Crowne & Marlowe, 1960). An example item is “Before voting I thoroughly investigate the qualifications of all the candidates.” Social desirability was assessed by totaling how many responses they respond to in a “socially desirable” way. For each socially desirable response, the participant was given one point. The total points were added up and a higher score indicated a higher level of social desirability. Scores above 21 suggest socially desirable responding (Lambert et al., 2016), the mean for this sample was 14 (SD = 7.14). The reliability of this measure was good for this sample, Cronbach’s α = .89.
Benevolent sexism. The Ambivalent Sexism Inventory was used to measure benevolent sexism (Glick & Fiske, 1996). This inventory includes 22-items that are rated on a 6-point Likert scale ranging from 0 (disagree strongly) to 5 (agree strongly). An example of one statement is, “Many women have a quality of purity that few men possess.” Benevolent sexism was calculated by averaging the scores of these items. The reliability of this measure was good for this sample, Cronbach’s α = .92.
Rape myth acceptance. Rape myth acceptance was measured via 30-item Acceptance of Modern Myths About Sexual Aggression Scale (Gerger et al., 2013). Each item was rated on a 5-point Likert scale ranging from 1 (completely disagree) to 5 (completely agree). An example of one statement is, “When a man urges his female partner to have sex, this cannot be called rape.” To calculate rape myth acceptance, an average was taken of the scores for all items. The reliability of this measure was good for this sample, Cronbach’s α = .95.
Interpersonal violence and crime exposure. Participants were asked to self-report their exposure to interpersonal violence using two items taken from the Life Events Checklist for DSM-5 (LEC; Weathers et al., 2013) regarding exposure to physical assault and sexual violence. Another item was created for this study to assess exposure to property crime, “property crime (theft, robbery, mugging, stolen money, or other items).” Participants responded to these items on a 4-point nominal scale with the options of “happened to me,” “witnessed it,” “someone close to me experienced this,” or “doesn’t apply.” Responses were coded as exposure if the participant indicated the event happened to them or they witnessed it firsthand.
Results
A repeated measures ANOVA comparing each participants blame ratings for each crime outcome revealed that participants did not blame the rape survivor (M = 1.40, SD = 0.74) significantly more than the survivor of physical assault (M = 1.47, SD = 0.82) nor the theft victim (M = 1.48, SD = 0.75), F(2, 192) = 0.39, p = .67, ηp2 = .004. 1 Likewise, another repeated measures ANOVA revealed that participants did not blame the rapist (M = 6.84, SD = 0.49) significantly less than the physical aggressor (M = 6.74, SD = 0.59) nor the thief (M = 6.81, SD = 0.52), F(2, 192) = 1.47, p = .23, ηp2 = .02.
The pattern of findings for victim blaming after different crime types remained nonsignificant after controlling for the socially desirable responding scale with each of the crime types, F(2, 190) = 0.69, p = .50, ηp2 = .007. In addition, the findings were not significantly moderated by benevolent sexism nor history of crime exposure, all interaction terms ps > .05.
However, differences in victim blame ascribed to different crimes was moderated by rape myth acceptance, crime × rape myth acceptance interaction, F(2, 190) = 4.60, p = .01, ηp2 = .05. Follow-up analyses revealed a group difference such that individuals high in rape myth acceptance blamed the victim more across all three outcomes relative to individuals low in rape myth acceptance. In addition, those low in rape myth acceptance had statistically significant differences in ratings for crime outcomes, F(2, 96) = 3.46, p = .04, ηp2 = .07, such that they blamed the theft victim (M = 1.32 SE = .08) significantly more than the physical assault survivor (M = 1.15, SE = .06), p = .02 and showed a trend toward blaming the theft survivor (M = 1.32, SE = .08) more than the rape survivor (M = 1.22, SE = .07), p = .07. In contrast, individuals high rape myth acceptance did not have statistically significantly different ratings of blame across outcomes, F(2, 94) = 0.89, p = .42, ηp2 = .02.
In addition, the differences in victim blame for crime type was moderated by the education level of the participants, Crime × Education level interaction, F(2, 190) = 4.93, p = .008, ηp2 = .05. A plot of this interaction revealed a similar pattern to rape myth acceptance such that those with a high educational level (college degree or higher) may have blamed the theft victim more; however, none of the simple effects for this moderation effect were statistically significant.
Summary
Across analyses, there was no evidence of a crime-specific bias toward victim blaming sexual assault survivors more than survivors of other crimes. In addition, the rates of victim blame were generally low: less than two on a scale from one to seven. It is worth noting that within-subjects designs strengthen statistical power and a power analysis suggested the sample of 97 participants exceeded the 43 needed to detect a medium effect at the p = .05 level.
Study 2
Past vignette-based studies comparing victim-blaming attitudes following different types of crimes (Felson & Palmore, 2018), including Study 1 of this research, have relied on self-report scales. However, self-report scales are highly susceptible to impression management as it is typically discernable what the scale is attempting to measure (Niemi et al., 2016). This methodological limitation may be particularly problematic for a construct like blame. Blaming can in some circumstances be a social norm violation and third-party blaming is a social act, which can have negative consequences for the accuser (Malle et al., 2014). Although the Crowne and Marlowe (1960) measure used in study 1 suggested that participants in the sample were not prone to a stable personality style of socially desirable responding, it cannot be known whether situation-specific socially desirable responding occurred for blame ratings. Therefore, the possibility remains that past research, including Study 1, may have failed to detect differences due to the transparency of the measures used. Furthermore, overreliance on one method limits our knowledge and the strength of our confidence about a construct (Campbell & Fiske, 1959). Niemi and colleagues proposed an alternative method of measuring blame using implicit causal statements, which they utilized in rape victim blaming research, but it has not previously been applied to the question of crime-specific bias. The aim of Study 2 was to test the comparison of blame for crime outcomes via this less reactive measure of blame.
Method
A total of 151 participants were recruited from TurkPrime and completed the study. All participants resided in the United States and denied having had exposure to interpersonal violence. The final sample was 61.59% (N = 93) male with a mean age of 39.49 years (SD = 13.68). Participants reported the following nonmutually exclusive racial and ethnic identities: White (79.5%), Black or African American (10.6%), Asian American or Pacific Islander (7.3%), Hispanic or Latinx (5.3%), and Native American (0.7%).
Following informed consent, participants read Vignette 1 and were randomly presented with either a rape or theft outcome. Participants then completed an implicit causal attribution statement (Niemi et al., 2016) in which they were asked to write why they thought the rape or robbery occurred by completing the sentence “Christopher did this to Amy because …” Then, two blind coders independently rated whether the response blamed the victim or the perpetrator (typically apparent from the pronoun that followed the word “because,”) with substantial agreement, Kappa = .84. Participants also completed the same self-reported blame questionnaire used in Study 1 after the implicit causal attribution statement.
Results
Analysis of the implicit causal statements revealed that participants were significantly more likely to blame the theft victim and less likely to blame the rape victim than expected by chance, χ2 (1, N = 146) = 6.65 p = .01. That is, 5.4% of participants who received the rape outcome, and 19.4% of the participants who received the theft outcome, blamed the victim. When blame was measured with the traditional self-report scale, blame was not statistically significantly higher for the theft victim (M = 2.23, SD = 1.27) than the rape victim (M = 1.89, SD = 1.13), t(149) = 1.73, p = .09, d = 0.28. Likewise, self-reported perpetrator blame for theft (M = 6.64, SD = 0.70) versus rape (M = 6.77, SD = 0.54) was not statistically significantly different, t(149) = –1.28, p = .20, d = –0.21.
Summary
As with Study 1, results of Study 2 did not show a pattern of greater blame bias following sexual assault relative to theft. In fact, the novel use of implicit causal statements suggested less blame toward the rape survivor, a finding that is consistent with some previous research using self-report measures of blame (Felson & Palmore, 2018).
Study 3
The results of Study 1 and Study 2 of this research suggest low overall rates of victim blaming and no evidence of a particular bias toward blaming sexual assault survivors more than survivors of other crimes. However, these findings are based on observer attitudes toward a hypothetical situation; attitudes that are ultimately important for their relationship to real world behaviors. Survivors themselves may be at the best vantage point to report on potential bias. In a sample of female sexual assault survivors, 98% reported experiencing negative social reactions when disclosing their traumatic experience to others, with victim blame highlighted as a particularly unhelpful response (Filipas & Ullman, 2001). There appears to be a dearth in the literature, however, comparing the experience of negative social reactions based on crime type. For example, a recent meta-analysis on social reactions to interpersonal violence revealed that 70% of the studies found on negative social reactions were exclusively focused on sexual violence samples (Dworkin et al., 2019). This review found only a few studies of social reactions with mixed samples of victimization types (Andrews et al., 2003; Dunmore et al., 2001; Hassija & Gray, 2012) and these studies did not report a comparison of negative social reactions by crime types. This gap in survivors’ accounting of reactions to different types of crime is unfortunate given that the actual reactions sexual assault survivors receive in the real world have implications for a variety of outcomes such as survivors’ mental health (Dworkin et al., 2019) and decisions about reporting the crime to police (Reich et al., 2021).
Furthermore, understanding whether these reactions are more common following one crime type or another could have implications for trauma-informed services and outreach.
Therefore, the aim of Study 3 was to examine survivors’ experiences with social reactions following sexual violence versus the same survivors’ experience disclosing another form of victimization (i.e., physical assault and theft) via a within-subjects design. Survivors who had experienced at least two of these crimes completed social reaction measures to allow for comparisons within the survivors’ life experience. Other forms of negative social reactions were also explored.
Method
Participants
A total of 228 interpersonal trauma survivors were recruited from a larger study on blame perceptions on TurkPrime. Participants were excluded when they: only experienced one of the three crime types (n = 74), failed the attention checks (n = 50), or had incomplete data (n = 10). The final sample (N = 94) reported experiencing and disclosing to others at least two of the following forms of victimization: sexual assault (n = 75), physical assault (n = 65), and theft (n = 76). Participants were mostly female (71%). Racial and ethnic identities were as follows: White (88.3%), Black and/or African American (8.5%), Asian or Pacific Islander (4.3%), Hispanic or Latinx (2.1%), and Native American or Alaskan Native (1.1%). The mean age of participants was 43.04 years (SD = 12.09) with a mean of 15.48 years of education (SD = 2.50).
Timing of First Disclosure and Recipients of Any Disclosures Following Different Victimization Event.
Note. Sexual n = 75, 66 of whom disclosed; physical n = 65, 57 of whom disclosed; theft n = 76, 69 of whom disclosed.
Acquaintance = neighbors, classmates, coworkers.
Procedure
Following informed consent, participants completed a battery of measures that included the same measure of crime exposure described in Study 1. The survey platform, Qualtrics, was programed so that participants were then asked to complete the same set of questions for each of the forms of victimization types they indicated they had experienced: sexual, physical, and/or theft.
Negative social reactions measure
Negative social reactions to each form of victimization were assessed via the repeat administration of the 28 negative items from the Social Reactions Questionnaire (Ullman, 2000). This measure includes five subscales for different types of negative reactions including: victim-blaming (e.g., told victim it was their fault), stigma (e.g., pulled away or treated them differently), control (e.g., told the victim what to do), silencing/distraction (e.g., told victim to stop talking about it), and egocentric responses (e.g., the recipient of the disclosure was so upset that the survivor had to comfort or calm them down). An example item is, “Told that you were to blame or shameful because of this experience.” Participants rated the frequency in which they experienced each negative response on a 5-point Likert scale of 0 (never) to 4 (always). The measure was scored by taking the mean of all items as well as the mean for items within each of the subscales. Higher numbers indicate more negative reactions. Although originally created and commonly used within the context of sexual assault, the measure has previously been used with other forms of victimization (Dworkin et al., 2019), including theft (Andrews et al., 2003). The reliability of this measure was good for the overall negative score for each crime type: sexual assault Cronbach’s α =.91, physical assault α =.88, and theft α = .92. The reliability was also good for the subscales with alphas ranging from .71 to .90.
Results
Paired Sample T-Tests Comparing Survivors’ Experiences of Negative Social Reactions Following One Crime vs. Another.
*p < .05, **p < .01.
Summary
Consistent with the findings in Study 1 and Study 2 from the observer perspective, the results of Study 3 did not demonstrate greater victim blaming following sexual assault relative to other victimization when assessed from the survivor’s perspective. However, sexual assault survivors do appear to receive more stigmatizing and silencing reactions than after other forms of victimization and interpersonal trauma survivors seem to receive more egocentric responses than theft victims.
General Discussion
Across three separate studies, victim blaming was observed following different forms of victimization, however, there does not appear to be a particular bias to victim blame survivors of sexual assault more so than other crimes. This finding was consistent across alternative methods of assessing blame among perceivers and according to survivors’ firsthand experiences of victim blaming following disclosures of different types of victimization. However, the current research did demonstrate a crime-specific bias for other forms of negative social reactions. Specifically, survivors reported experiencing more silencing or stigmatizing reactions following disclosure of sexual assault relative to their experiences with other forms of victimization (i.e., physical assault and theft). Additionally, interpersonal trauma disclosures (i.e., sexual or physical assault) resulted in more egocentric responses from others relative to disclosures of theft. These findings might suggest cultural bias toward sexual crimes (Suarez & Gadalla, 2010) more often manifests in the form of stigmatizing remarks, whereas egocentric distress might arise more readily among disclosure recipients when the victim was subjected to bodily harm rather than nonviolent crimes.
Several limitations for the current research should be noted. This research was conducted in the United States; however, blame is moral judgment (Malle et al., 2014), which may vary depending on cultural context (Yamawaki & Tschanz, 2005). Study 1 and Study 2 utilized vignettes that held constant characteristics of the victim (i.e., female) and the perpetrator (i.e., a single male). A review of past research suggests that blaming attitudes toward sexual assault survivors vary depending on (a) victim characteristics including gender, gender norm conformity, sexual orientation, sexual objectification, attire, weight, degree of acquaintance with the perpetrator, alcohol consumption, and resistance, and (b) perpetrator characteristics including alcohol consumption, socioeconomic status, attractiveness, desirability, prior criminal history, perpetrator motives (violence vs. sexual), and the number of other perpetrators (Grubb & Turner, 2012; Penone & Spaccatini, 2019; Van Der Bruggen & Grubb, 2014). Therefore, it is unknown whether any of these factors could interact with the type of crime to influence judgments of blame. In addition, Study 1 and Study 2 of this research focused on victim and perpetrator blame, which likely does not represent the continuum of supportive to negative reactions survivors might encounter (e.g., denigration, denouncing the crime, helping intentions and behaviors toward the victim, and juror decision making; see for example McGuire et al., 2012). Finally, Study 3 examined naturalistic differences between survivors’ experiences disclosing victimization by one crime versus their experience disclosing another. However, besides the crime type, the survivors may have differed by other unknown factors like severity or contextual factors like the victim and perpetrator characteristics. As such it is unknown how these factors may have influenced the findings.
Conclusions and Implications
Despite media and research attention specific to victim blaming in the context of sexual violence, a parsimonious explanation for the current and past findings (Felson & Palmore, 2018) is that victim blaming has a common general psychological process (e.g., hindsight bias and just world beliefs; Furnham, 2003; Roese & Vohs, 2012), regardless of victimization type, rather than a basis in crime-specific stereotypes or assumptions. Therefore, research aimed at understanding common mechanisms of victim blame appears warranted. For example, Malle et al. (2014) offer a general theory of blame that suggests that perceivers may consider counterfactual preventative actions when making blame judgments, on the basis of perceived negligence to prevent the undesired outcome. This theory predicts blaming will be diminished, or avoided, if the perceiver is convinced of two contextual factors. One factor is convincing the perceiver that the victim lacked the capacity to prevent the crime, meaning they could not have prevented it (e.g., the victim is unconscious or in altered state and unable to prevent the crime). Consistent with this theory, a study found that victims were blamed less when an attorney presented “even-if” counterfactual conditions regarding the victim’s behavior that nonetheless resulted in the same victimization fate (Branscombe et al., 1996). This is similar to the “consider-the-opposite” strategy for combating hindsight bias (Roese & Vohs, 2012). However, Malle argues that the perceiver’s consideration of counterfactuals (favorable or otherwise) is dependent on the perceiver’s judgment of another factor first. Specifically, the perceiver must first decide whether the victim had an obligation to prevent the crime, meaning whether someone should or should not take particular steps to prevent victimization. Future research might, therefore, develop and test strategies to reduce perceptions that victims have an obligation to prevent crimes to in turn mitigate the risk of victim blaming.
However, in some cases, the continued research focus on victim blame within the context of sexual assault remains important. For example, this focus may still be necessary when the research question is specific to unique factors for this form of victimization (e.g., the influence of penetration on attitudes). In addition, although the current findings suggest that victim blaming may not be particularly amplified following sexual assault, these results do not preclude the possibility that victim blaming is particularly pernicious in its consequences for survivors of sexual assault. For example, future research might examine whether victim blaming following sexual assault has a stronger impact for survivors’ decisions to disclose again in the future, whether to seek help or report the crime, and psychological outcomes compared to victim blaming following other forms of victimization.
Although past research suggests negative social reactions can deter sexual assault survivors from disclosing their experience further (Ahrens, 2006), including not reporting the crime (Reich et al., 2021), these studies typically focus on negative social reactions in general but specifically among sexual assault survivors. The current study highlights the specific negative social reactions of silencing and stigmatizing may be more common toward sexual assault survivors. As such, future research might examine whether the higher rates of silencing and stigmatizing reactions following sexual violence relative to other crimes helps explain the lower rate of reporting the crime to police (Reich et al., 2020) or the higher rates of psychopathology following sexual assault (Kessler et al., 1995).
Findings may have implications for trauma-informed services and outreach. In research primarily conducted with sexual assault survivors, victim blame is associated with a myriad of negative outcomes for the survivors including an increased risk of developing psychopathology (Dworkin et al., 2019), poorer recovery (Ullman, 1996), increased self-blame (Bonnan-White et al., 2015; Dworkin et al., 2019), decreased disclosure in the future (Ahrens, 2006), underreporting to police (Reich et al., 2020), and increased risk of revictimization (Ullman & Najdowski, 2011). The current findings suggest trainings for bystanders and service providers should continue to promote avoiding victim blaming following sexual assault but that such efforts may be warranted following any form of victimization also. Service providers and advocates should also be vigilant about recognizing and preventing silencing and stigmatizing reactions following disclosure, particularly when serving sexual assault survivors.
Appendix
Vignette Example
One Friday night, Amy threw a football viewing party at her house. Amy greeted her guests dressed in her team jersey and short shorts. She was sure to show each guest her new, luxury car in the driveway as they arrived. One of her friends, Christopher, was wearing the jersey for the other team and athletic shorts. Amy did not know him very well, but he lived in her neighborhood and worked at the local coffee shop. They all sat together watching the game and everyone was cheering for their favorite team. The game was close, but Amy’s team won. When everyone left, Christopher asked if he could stay a while longer and Amy said yes. While they were cleaning up, Amy bragged about her team winning. Christopher said the game was not fair and their conversation turned into an argument. Amy’s purse was laying on the table and she asked Christopher to move it for her, so he picked it up and moved it to the hall table. Amy turned on music and started dancing, moving closer to Christopher as she swung her hips. Then, [Christopher grabbed and held down Amy with his hand over her mouth as he forcefully had sex with her. OR Christopher hit and shoved Amy, knocking her to the ground hard. OR Christopher took a wad of cash from Amy’s purse without her noticing.]
The full text of all scenarios is available from the first author on request.
Footnotes
Acknowledgments
We would like to thank Naseem Farahid, Jill Jenson and Jeremy Jamieson for feedback on this article.
Declaration of Conflicting Interests
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was made possible due to support from the University of Minnesota Duluth, Psychology Department Internal Project Funding.
