Rape Myth Acceptance is Lower and Predicts Harsher Evaluations of Rape Among Impacted People

Public reactions to rape are often distorted by the acceptance of so-called rape myths. The goal of our research was to examine how rape myth acceptance (RMA) is connected to the evaluation of rape cases among survivors, unimpacted people, and those impacted by rape through a close relation, who can potentially be important allies of survivors in bringing about social change. We tested these connections in three online survey studies. In Study 1 (N = 758) we found that those impacted by rape personally or through a close relation accepted rape myths less. In Study 2, using a nationally representative sample in Hungary (N = 1007), we tested whether RMA predicted uncertain rape cases more strongly than certain (i.e., stereotypical) ones, considering that a stereotypical rape scenario is condemned by most members of society, but not all rape is labeled as such. We found that RMA predicted the evaluation of both rape scenarios, but the prediction was stronger when rape was uncertain. In Study 3 (N = 384), in a pre-registered study we examined how RMA predicted the evaluation of rape cases amongst people with different previous experiences (impacted/unimpacted). We found that unimpacted people accepted rape myths more, blamed the victim more and labeled the case less as rape when the case was uncertain. These findings suggest that rape myth acceptance functions as cognitive schema and that rape impacted people could have a key role not only in the life of survivors but as allies for social change as well. Research Article This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License, CC BY 4.0, which permits unrestricted use, distribution, and reproduction, provided the original work is properly cited.

Keywords rape myth acceptance, rape, rape survivor, rape evaluation, victim-blaming Highlights • Unsupportive reactions to rape survivors are not only a global public health problem but a human rights violation. • People with personal experience with rape endorse rape myths less than unimpacted people. • People with higher rape myth acceptance labeled both uncertain and certain cases less as rape and blamed the victim more, and these people evaluated uncertain rape cases less harshly than certain ones. • The different evaluations of rape cases were predicted by rape myth acceptance, which shows that rape myth acceptance is particularly important when a rape case is not stereotypical which is in accordance with previous research.
One in 20 women have experienced rape according to the estimations of the European Union Agency for Fundamental Rights (FRA, 2014), and only 11 out of 100 000 people report rape to the police on average in the United States in the collected sample, and this number greatly varies between countries (FRA, 2014). People who report rape to the police usually experience guilt, shame, embarrassment, fear of retaliation, and lack of trust in the police (Sable et al., 2006), all of which are connected to rape myths. Rape myths are descriptive and prescriptive beliefs about rape that serve to deny and justify men's aggression against women (Bohner et al., 1998) and trivialize its effects on the survivor (Brownmiller, 1975). They constitute a specific domain of sexism that con tributes to sexual aggression and coercion (Brownmiller, 1975). It is separate from general rape attitudes because the main function of rape myths is to deny its pervasiveness and structural causes (Forbes et al., 2004;Lonsway & Fitzgerald, 1994). Rape myths either put the blame on the victim (e.g., "if a girl acts like a slut, eventually she is going to get into trouble") or excuse the perpetrator (e.g., "rape happens when a guy's sex drive goes out of control") by rationalizing rape (Payne et al., 1999). Rape myths encourage victim-blaming and provide a feeling that the world is predictable and fundamentally just, and only those people get raped who somehow deserve it (Lonsway & Fitzgerald, 1994).
Rape myths function as cognitive schemas and therefore influence how people in terpret social information (Gerger et al., 2007). Eyssel and Bohner (2011) found that the more information participants received, the stronger the effect of rape myth accept ance was on blaming judgements, irrespective of whether the additional information pertained to the survivor or the perpetrator. Furthermore, they found that participants who thought they received additional information about the rape case subliminally felt more entitled to judge -in line with their previous rape myths. Süssenbach and Bohner (2011) and Süssenbach et al. (2017) found that rape myth acceptance even affected visual attention and information processing: those with higher rape myth acceptance were more sensitive or processed clues that were consistent with rape myths more easily, shif ted their visual interest and preferred information about the victim over the perpetrator. Therefore, those who endorse rape myths are increasingly likely to identify women's friendly behavior as sexually teasing (Willan & Pollard, 2003), less likely to suggest rape survivors report the rape (Frese et al., 2004), and less likely to label forced sex as rape (e.g., Peterson & Muehlenhard, 2004).
Furthermore, rape myths provide an explanation and justification to survivors and their environment about why rape happens and why specifically to them. Although accepting rape myths can help regain a sense of control and reassure survivors that rape cannot happen to them again (Hayes et al., 2013), there is no evidence that rape survivors would accept rape myths more than unimpacted people. In fact, Egan and Wilson (2012) found that rape survivors who reported the crime to the police accepted rape myths less than those who did not report the case. Other studies have found either that rape survivors and unimpacted people do not differ in their level of rape myth acceptance (Carmody & Washington, 2001) or that rape survivors accept rape myths less than unimpacted people. This can be explained by the fact, on the one hand, that many survivors could experience rape as uncontrollable (Egan & Wilson, 2012) and on the other, that they may have a better understanding that rape cases are not always in line with the "real rape" scenario. The "real rape" scenario entails the idea that rape is committed by a stranger, the intercourse is physically forced (Bell et al., 1994), women are physically injured, and they report the assault to the police immediately (Hohl & Stanko, 2015). Although rape does not always happen in a stereotypical way, however, because of the lack of contradicting experience or knowledge, unimpacted people can endorse broadly accepted preconceptions about rape (Baugher et al., 2010;Vonderhaar & Carmody, 2015). These results imply that despite the general function of rape myths (Gerger et al., 2007) for survivors of rape, rape myth acceptance does not necessarily serve as a justification and explanation to rape to reduce anxiety, and they do not see it as some kind of guide for women on how to behave to avoid being raped.
Close friends and relatives of survivors of rape could have a key role in helping survivors to process their trauma. However, previous studies compared only survivors and unimpacted people, whereas people may be personally affected by rape not only as survivors but also through the experience of a close relative or friend. As far as we know, no studies have previously investigated rape myth acceptance specifically among people impacted by rape through a close friend or relative. There is some evidence from two studies that college students who knew survivors had a lower level of rape myth acceptance (Ellis et al., 1992;Gilmartin-Zena, 1987) whereas other studies did not find any difference between people who knew rape survivors and the general population (Borden et al., 1988). Because of the contact with survivors, it is possible that rape impacted people accept stereotypical beliefs about rape less. Furthermore, people who think that rape is a sexual act rather than a violent one, are more likely to downplay the seriousness of rape and think that it can be enjoyable for the victim (Emmers-Sommer et al., 2006), which is less likely the case among people impacted by the rape of a close relation.
Close friends and relatives of rape survivors can offer the most direct social support for survivors and play an important role in interventions aiming to reduce rape and sexual assaults and in social change movements. Based on previous studies they may be more effective in confronting others' rape myths than survivors, because they do not directly benefit from the change, just like men are sometimes more effective in confronting sexism than women (Drury & Kaiser, 2014).
Rape myths create a normative environment in which labeling a case not as rape, blaming a victim for it, and excusing the perpetrator is acceptable (Bohner et al., 2006). Based on previous results (FRA, 2014), it is reasonable to conclude that low gender equality and the associated high level of rape myth acceptance creates an environment in which survivors do not trust the police and other authorities, and do not think they will be treated fairly (Sable et al., 2006;Wirth & Winkler, 2015). Therefore, they do not report the rape to the police, which explains why latency is higher in less gender-equal countries that are also likely to endorse rape myths more widely (FRA, 2014). According to the global gender gap index, Hungary -where the current research was conductedholds the 101st position in equality of the positions of men and women in society (World Economic Forum, 2016), suggesting that gender equality is lower than in most of the western world. Estimations suggest that unreported rape cases are 415 times higher than those reported in Hungary (Wirth & Winkler, 2015). Therefore, understanding the phenomenon of rape myth acceptance has great relevance in the context of Hungary from a practical perspective, for example, to understand how people react to rape cases and survivors.

Research Aims and Hypotheses
The main purpose of our research was to examine rape myths as cognitive schemas. In Study 1 we examined the differences between people impacted and unimpacted by rape. In Study 2 we were interested in the consequences of rape myth acceptance, that is, whether they are connected to how people evaluate rape cases. In Study 3 we examined these variables together to understand how prior experience with rape influences the connection between rape myth acceptance and the evaluation of rape cases in a pre-reg istered study. A better understanding of differences between people with and without the experience of rape and the consequences of rape myth acceptance seems crucial because most rape cases are not stereotypical, and rape myth acceptance can be a tool to explain Rape Myth Acceptance, Rape Evaluation, and Experience away these experiences and create obstacles to tackle the widespread problem of rape in society.
We conducted the research in Hungary, where no similar research had been conduc ted before and which can be characterized as a gender unequal social context. We argue that this context can create norms that sanction rape myth acceptance and can therefore affect the degree to which people label cases as rape. As previous research has shown, such labeling can fundamentally affect the overall evaluation of the case and the treatment of perpetrators and survivors (Nyúl et al., 2018).

Study 1
In Study 1, we hypothesized that rape survivors would endorse rape myths less than unimpacted people (H1) and we hypothesized that those people impacted through a close relative or friend accept rape myths less than unimpacted people (H2).

Participants
Participants were recruited in two different ways. We collected data amongst under graduate students and recruited participants online from a community sample using convenience sampling. The final sample size was N = 758 (25.6% men, 74.4% women, age M = 27.91, SD = 10.37, for more see Table S1 in the Supplementary Materials). No participants indicated "other" for gender.
Because of the lack of previous studies to rely on, we did not calculate an adequate sample size in advance, but using G*Power software version 3.1.9.4, a post hoc sensitivity analysis was performed to detect the minimum effect size that could be achieved with 758 participants. The results showed that with a statistical power of 80% and a signifi cance level of α = .05, the minimum effect size to be detected was .11 which we reached in the current study.
We categorized participants into three groups based on their self-report of having personally experienced rape, having a friend/family member/loved one who experienced rape, or not knowing anyone who had been raped. We labeled participants as "rape survivors"; "rape impacted" people who knew a close person (close friend/relative/loved one) who was a survivor of rape, and we labeled participants who had neither experi enced rape, nor knew of any survivors as "unimpacted" people. Fifty-five participants reported that they were rape survivors (1.3% of all men and 11.4% of all women partici pants), 74 participants were impacted by rape by knowing someone close to them who had experienced rape (13.9% of all men and 11.2% of all women), and 496 participants were categorized as unimpacted (69.4% of all men and 64% of all women). Because the aim of Study 1 was to understand the differences between survivors, rape impacted (through a close friend/relative/loved one) and unimpacted people, we excluded people from the analysis who reported to know a rape survivor, but the person was not a close relation (n = 94, 13.4% of all men and 11.9% of all women) or chose not to answer this question (n = 39, 4.6% of all men and 5.3% of all women).

Measures and Procedure
We conducted the first study in 2014. We used a paper-and-pencil questionnaire for the student sample and an online questionnaire for the community sample. We conducted the research following the IRB approval of the Research Ethics Committee (REC) of United Ethical Review Committee for Research in Psychology (EPKEB).
After giving informed consent, participants completed the Updated Illinois Rape Myth Acceptance Scale (18 items, current study's Cronbach α = .91; McMahon & Farmer, 2011) on a 7-point scale (from 1 = completely disagree to 7 = completely agree; Nyúl & Kende, 2021) and different scales for validation purposes 1 , which we do not report in the current analysis because it falls outside the focus of this paper. We excluded those participants who did not finish the questionnaire, but because we supposed that the tests were robust, we did not plan to exclude statistical outliers.

Results
To test our hypothesis regarding prior experience with rape, we conducted a one-way ANCOVA to determine the difference between survivors, rape impacted people, and unimpacted people in rape myth acceptance, controlling for the effect of gender. There was a significant difference in the level of rape myth acceptance among people with different experiences with rape F(2, 621) = 10.77, p < .001, R 2 = .06, η p 2 = .03. Post-hoc comparison showed that unimpacted people's rape myth acceptance (M = 2.81, SD = 1.02) was higher than the RMA of survivors (M = 2.17, SD = 1.01) and rape impacted people (M = 2.44, SD = 1.02), but there was no difference between the latter two groups.

Discussion
Our hypothesis was that rape survivors would accept rape myths less than unimpacted people (H1) and that rape impacted people would accept rape myths less than unimpac ted people (H2). In contrast to some earlier studies (Vonderhaar & Carmody, 2015), but in line with others (Baugher et al., 2010), we found that survivors of rape and rape impacted people endorsed rape myths less than unimpacted people. This result suggests either that surviving or knowing someone who was raped decreases rape myth acceptance or 1) Validated Hungarian version of the Ambivalent Sexism Inventory (10 items hostile sexism scale, current study's Cronbach α = .89; 11 items, benevolent sexism scale, current study's Cronbach α = .86; Glick & Fiske, 1996;Hungarian validation: Szabó, 2008), the validated Hungarian version of Belief in a Just World Scale (8 items, current study's Cronbach α = .84; Dalbert, 1999;Hungarian validation: Berkics, 2008). that those who accept rape myths less, are more likely to label their own or others' experience as rape. Furthermore, these results also imply that rape survivors may be more likely to share their trauma with people who endorse rape myths less.
Following the results of Study 1, we were interested to reveal whether rape myth acceptance predicted uncertain rape cases more strongly than certain ones, considering that rape in its stereotypical form is condemned by all members of society, but cases are not always labeled as rape when they are less stereotypical, which we tested in Study 2.

Study 2
In Study 2 we were interested in the different outcomes connected to rape myth accept ance. The importance of examining the consequences of rape myth acceptance is that they have an important role in how people evaluate rape cases (Nyúl et al., 2018). The novelty of the study is that to the best of our knowledge there have been no other studies that examined the role of rape myth acceptance in the evaluation of certain and uncertain rape cases with an experimental design.
The importance of examining reactions to various rape scenarios is twofold: on the one hand, people's reactions reflect the normative context of rape in society, therefore, it affects whether perpetrators do or do not think that rape is a serious crime and what constitutes rape, and on the other hand, it affects whether survivors report the case to the police and seek help at all (Bohner et al., 2006). As we discussed in the introduction, evaluation of rape cases is affected by the stereotypicality of the rape, which affects the perceived certainty of rape. Previous studies found that a rape scenario is perceived as stereotypical if the perpetrator is a stranger to the victim and a deviant person, he is armed or uses physical force during the rape (McGregor et al., 2000), and the survivor immediately reports the case to the police and cooperates with them (Bongiorno et al., 2016). The gender of the survivor and the perpetrator and their prior relationship also affect whether a case fits into a stereotypical rape scenario, which in turn affects evalua tions of the rape. However, previous research found that stereotypical rape scenarios are rare (Bongiorno et al., 2016); yet people tend to label cases as rape and see them as more certain if they fit these stereotypes. Only in these stereotypical cases are perpetrators evaluated harshly, and victims are less likely to be blamed (Bell et al., 1994). If rape is not stereotypical, people are more likely to blame the victim and not label the case as rape (Bongiorno et al., 2016). Rape myth acceptance affects the evaluation of uncertain and counter-stereotypical rape cases more than it affects stereotypical cases (Nyúl et al., 2018).
Based on previous research (e.g., Chapleau et al., 2007;Eyssel & Bohner, 2011;Nyúl et al., 2018), we expected that people with higher rape myth acceptance would blame the victim more and label the case less as rape (H1) especially when the case is uncertain (H2).
The study was conducted in Hungary and we recruited participants with the help of an opinion poll company (SoliData) who relied on an online pool of respondents that were demographically similar to Hungarian society in terms of gender, age, and type of settlement, but participants had a higher-than-average education (N = 1007; men: n = 495, 49.2%; women: n = 512, 50.8%). There was no option to select "other" for gender. We did not calculate sample size based on the effect sizes of Study 1 but targeted N = 1000 which is typically used in representative opinion poll surveys in Hungary (see Poll of Polls, 2018). Mean age was 41.52 years (SD = 13.05) ranging from 18 to 64 years, level of education and type of settlement are demographically similar to the Hungarian population (see Table S2 in the Supplementary Materials). We excluded those participants who did not finish the questionnaire, but because we supposed that the tests were robust, we did not plan to exclude statistical outliers.

Measures and Procedure
Data was collected in 2016. The language of the questionnaire was Hungarian. After giving their informed consent similarly to Study 1, participants were presented with a certain and uncertain rape scenario in this order (for the exact wording of the cases, see Text S1 in the Supplementary Materials). For technical reasons, randomization was not possible; therefore, we conducted a complementary analysis in order to check whether the responses given to the certain rape vignette influenced the responses given to the uncertain rape vignette. We ran a moderation model (Model 1) in Process (Hayes, 2017) and found that rape myth acceptance influenced the uncertain rape labeling and victim-blaming beyond and above the labeling and victim-blaming in the uncertain rape scenario (see the results of the moderation analysis in Text S2 in the Supplementary Materials) which suggests that the results were not a consequence of the order effect only. We established the level of certainty based on Bongiorno et al. 's (2016) research; in the certain rape scenario the survivor physically resisted the perpetrator, and she fully cooperated with the police (e.g., "Éva [the survivor] said that she screamed and tried to escape but she couldn't. On the same night Éva went to the police and reported the case"), whereas in the uncertain one, she did not (e.g., "She said many times that she did not want to have sex with him, but physically she did not resist" "Szilvia [the survivor] went to the police and reported the case but it was really hard for her to work with them"). We measured victim-blaming with one item ("I think Éva/Szilvia [the survivor] is responsible for what happened") and participants labeled the case as whether they considered it rape or not ("To what extent do you think what happened was rape?"). Both items were measured on a 7-point scale from 1 = it was totally her fault/it was certainly not rape, to 7 = it was not her fault at all/it was certainly rape.
Participants completed the Hungarian version of the Updated Illinois Rape Myth Acceptance Scale (18 items, current study's Cronbach α = .90; McMahon & Farmer, 2011). The data collection was part of an omnibus survey which means that additional scales were administered in the study, independent of the current research 2 .

Results
We tested whether people evaluated the uncertain and certain scenarios differently with a paired sample t-test where we found that participants blamed the survivor more, t(1006) = 28,04, p < .001, and labeled the case as rape less, t(1006) = -33,94, p < .001, in the uncertain rape scenario, victim-blaming M = 3.36, SD = 1.68; rape labeling M = 4.96, SD = 1.92, than in the certain case, victim-blaming M = 1.42, SD = 1.02; rape labeling M = 6.70, SD = 0.97. We ran a hierarchical linear regression analysis to test the role of rape myth acceptance in the evaluation of an uncertain and a certain rape case (see Table  1). The effect of participants' gender was controlled in the model by adding the variable in the first step, whereas RMA was entered in the second step. We found that RMA was a significant positive predictor of rape labeling both in the case of an uncertain rape scenario and in the case of a certain rape scenario. Rape myth acceptance was also a significant predictor of victim-blaming in connection with both the uncertain and certain rape scenarios. The unstandardized regression coefficients were bootstrapped with 1000 draws to examine if rape myth acceptance was a stronger predictor of victim-blaming as well as rape labeling in the corresponding uncertain scenarios. The results confirmed that this was the case since no overlaps between 95% confidence intervals (in certain vs. uncertain cases) were present (Victim-blaming: CI certain [.11, .24 [.48, .66]). Similar to Study 1, G*Power software version 3.1.9.4 was used to perform a post hoc sensitivity analysis to detect the minimum effect size that could be achieved with 1007 participants. With a statistical power of 80% and a significance level of α = .05, the minimum effect size to be detected was .006, which we reached in the current study.
2) Participants also completed the short form of the Ambivalent Sexism Inventory (5 items hostile sexism scale, current study's Cronbach α = .84; 5 items benevolent sexism scale current study's Cronbach α = .79; Glick & Fiske, 1996) and other scales for other studies which were not connected to the current paper and conducted by researchers other than the authors of this paper. Rape Myth Acceptance, Rape Evaluation, and Experience Supporting our hypothesis and in line with previous findings (Eyssel & Bohner, 2011), we found that participants with higher rape myth acceptance blamed the survivor more and labeled the case less as rape (H1). Supporting our second hypothesis, rape myth acceptance explained greater variance and was a stronger predictor of rape labeling and victim-blaming in the evaluation of uncertain rape cases compared to the certain case (H2). These findings are in line with previous research (Nyúl et al., 2018), suggesting that when the case is uncertain (i.e., they do not fit the stereotypical rape scenario which is in fact the case most of the time) rape myths are connected to people's way of thinking about the case resulting in stronger victim-blaming. These results support the hypothesis that rape myths function as cognitive schemas (Gerger et al., 2007) and influence how people interpret information about rape cases. Previous research found that people rely on their rape myths more when they think they received additional information about the rape case and evaluate the case more in line with their rape myths (Süssenbach & Bohner, 2011;Süssenbach et al., 2017). Based on our results, it seems that when the case is uncertain and there is more room for imagination, people also use their rape myths to complement the missing information and rely more on their rape myths when they evaluate the rape case.

Study 3
Besides examining the replicability of our results, in Study 3 we wanted to understand how rape myth acceptance adds to the understanding of the connection between prior experience with rape and the evaluation of rape cases. In a preregistered study (https:// aspredicted.org/SRY_JGM) we tested the following hypotheses: survivors and impacted people accept rape myths less than unimpacted people (H1), people label uncertain rape scenarios less as rape and blame the victim more than in case of certain rape scenarios (H2). We hypothesized a main effect of rape myth acceptance on rape labeling and victim-blaming, specifically that rape myth acceptance would predict rape labeling negatively and victim-blaming positively (H3), that rape myth acceptance would be a stronger predictor of rape labeling and victim-blaming when the case is uncertain than when it is certain (H4), and that survivors and impacted people would label scenarios more likely as rape and blame the victim less than unimpacted people (H5). Furthermore, we hypothesized that rape myth acceptance would have the strongest effect on rape labeling (negative) and victim-blaming (positive), in the case of unimpacted people (in comparison to rape impacted and survivors), when the rape scenario is uncertain (H6).
We collected data amongst university students and recruited additional participants online using convenience sampling. We used G*Power (Faul et al., 2009) to determine the sample sizes necessary to detect small-sized effects with a statistical power of 80% and a significance level of α = .05; therefore our target number was 357. The final sample was N = 384 (19.6% men, 80.4% women, age M = 21.64, SD = 3.58 years). No participants indicated "other" for gender. We categorized participants based on their prior experience with rape similarly to Study 1. Twenty-four participants reported that they were rape survivors (4.9% of all men and 7.2% of all women participants), 50 participants were impacted by rape by knowing someone close to them who had experienced rape (16.4% of all men and 14.5% of all women), and 265 participants were categorized as unimpacted people (78.7% of all men and 78.3% of all women). As the aim was to understand the differences between survivors, rape impacted (through a close friend/relative/loved one) and unimpacted people, we excluded people from the analysis who knew somebody impacted by rape, but this person was not a close relative or friend (n = 50, 22.7% of all men and 10.7% of all women), and those who chose not to answer this question (n = 7, 1.3% of all men and 2% of all women).

Measures and Procedure
We collected data for Study 3 in 2021 using an online questionnaire. We conducted the research following the IRB approval of the Research Ethics Committee (REC) of ELTE Eötvös Loránd University.
After giving informed consent, participants received either a certain or an uncertain rape scenario using the same vignettes as in Study 2, in a randomized order. We meas ured victim-blaming and rape labeling identically to Study 2. After that, participants completed the Updated Illinois Rape Myth Acceptance Scale (18 items, current study's Cronbach α = .89; McMahon & Farmer, 2011;Hungarian version Nyúl & Kende, 2021), then participants indicated their prior experience with rape, as in Study 1. Participants also completed the validated Hungarian version of the Ambivalent Sexism Inventory (10 items hostile sexism scale, current study's Cronbach α = .90; 11 items, benevolent sexism scale, current study's Cronbach α = .86; Glick & Fiske, 1996;Hungarian validation: Szabó, 2008) but we did not use this variable in the analysis. We excluded those participants who did not finish the questionnaire, but because we supposed that the tests were robust, we did not plan to exclude statistical outliers.

Results
Because we did not have enough survivors in the sample, we created two groups based on prior experience with rape: (1) a group of impacted people in which participants were either survivors of rape or had a close friend or relative who experienced rape, and (2) unimpacted people in which participants neither experienced rape, nor knew any survi vors. We chose to do this to be able to run our pre-registered statistical analyses which we would have been unable to do using three groups, given the low number of survivors in the sample. We acknowledge that this simplification conceals important differences between survivors and impacted people that we could not present and analyze in the current analysis.
To test Hypothesis 1, we conducted an ANCOVA to compare rape myth acceptance among impacted and unimpacted participants, controlling for the effect of gender. The results revealed a significant difference between the groups: F(1, 334) = 7.68, p = .006, η p 2 = .02. That is, rape myth acceptance among impacted participants (M = 2.51, SD = 0.86) was lower than among unimpacted participants (M = 2.81, SD = 0.86).
To test Hypothesis 2, two paired sample t-tests were conducted to compare partici pants' victim-blaming and rape labeling scores in certain vs. uncertain rape scenarios. In the certain scenario To test Hypotheses 3-6, two separate mixed ANOVA models were performed for the two dependent variables: rape labeling and victim-blaming, controlling for the effect of gender. Respondents' prior experience (rape-impacted vs. unimpacted) was entered as the between-subject variable to compare the two groups in victim-blaming and rape labeling. Rape scenario (certain vs. uncertain rape cases) was entered as the within-sub ject variable and rape myth acceptance was entered as the moderator in both models. Although, we supposed that the tests would be robust enough to handle outliers, because Levene's test for equality of variances was violated for the certain rape labeling scenario, possibly resulting from the close to ceiling effect on rape labeling, we treated the extreme outliers as missing values for this one variable. The main effect of rape scenario was found to be statistically significant on rape labeling, F(1, 310) = 4.65, p = .032, η p 2 = .01, Observed-Power = .58, while regarding victim-blaming, F(1, 332) = 2.23, p = .136, η p 2 = .007, Observed-Power = .32, the main effect was not significant. That is, participants labeled the scenario more as rape in the certain scenario, but victim-blaming was not significantly different across the two scenarios. No statistically significant between-subject main effects of prior experience were found for rape labeling, F(1, 312) = 2.43, p = .120, η p 2 = .008, Observed-Power = .34, or for victim blaming, F(1, 335) = .004, p = .952, η p 2 = 0, Observed-Power = .05. In other words, impacted and unimpacted people did not differ in victim-blaming and rape labeling based on whether the scenario was either certain or uncertain. The two-way interactions between prior experience and victim-blaming, F(1, 335) = 3.21, p = .074, η p 2 = .009, Observed-Power = .43, as well as between prior experi ence and rape labeling, F(1, 312) = 3.31, p = .070, η p 2 = .01, Observed-Power = .44, were not statistically significant either. Although the two-way interactions were not significant, to understand our results better, we ran four one-way ANCOVA models (controlling for gender) which were not part of the pre-registration. Specifically, we investigated the differences between rape labeling and victim-blaming in certain vs. uncertain scenarios for impacted vs. unimpacted respondents. The results showed that impacted participants labeled the case as rape significantly more than the unimpacted participants regarding uncertain rape scenarios, while no statistically significant difference was found in the case of certain rape scenarios. The pattern was similar with regards to victim-blaming, meaning that in the uncertain scenario impacted individuals blamed the victim to a significantly lower degree as compared to unimpacted individuals, while no difference was found in victim-blaming in the case of certain rape scenarios. Further, rape myth acceptance significantly interacted with both victim-blaming, F(1,335) = 53.806, p < .001, η p 2 = .138, Observed-Power = 1, and rape labeling, F(1, 312) = 41.460, p < .001, η p 2 = .117, Observed-Power = 1, which means that those who accepted rape myths less were more likely to label the case as rape and blame the victim less than those who accepted rape myths more. Finally, the 3-way interactions between rape myth acceptance, prior experience, and victim-blaming, F(1, 335) = .29, p = .588, η p 2 = .001, Observed-Power = .08, as well as between rape myth acceptance, prior experience, and rape labeling, F(1, 312) = .19, p = .661, η p 2 = .001, Observed-Power = .07, were not significant. Descriptive statistics for each group at different rape scenarios are presented in Table 2. The main purpose of our research was to examine rape myths as cognitive schemas, to understand how impacted and unimpacted people accept rape myths and how they evaluate rape cases. Importantly, we could not run the analysis for the originally planned three groups to test our first hypothesis regarding differences between survivor, impac ted, and unimpacted groups, because we did not have enough survivors in our sample. Therefore, we merged the groups of survivors and those who knew someone close to them who had experienced rape, because we wanted to examine whether people with prior experience with rape accepted rape myths differently than those who did not. This allowed us to run the pre-registered analyses, however, admittedly, it reduced the complexity that we were originally striving for.
Replicating the results of Study 1, we found that impacted people accepted rape myths less than unimpacted people.
In support of Hypothesis 2 and 3, and similarly to Study 2, we found that people labeled uncertain rape cases less as rape and blamed the victim more compared to certain rape cases and that people who accepted rape myths more, were more likely to blame the victim and label the case less as rape. Furthermore, in connection with Hypothesis 4, similarly to Study 2, we found that rape myth acceptance is a stronger predictor of rape labeling and victim-blaming of uncertain cases compared to certain ones. These results are in line with previous findings (Eyssel & Bohner, 2011;Nyúl et al., 2018) and with our hypothesis that rape myths are cognitive schemas that relate to that people evaluate rape cases in line with their general attitudes toward rape, that they rely on their a priori attitudes toward rape (e.g. Eyssel & Bohner, 2011) and that rape myths have a more powerful role in uncertain rape situations than in certain ones.
In connection with Hypotheses 5 and 6, we did not find a difference in the effect of rape myth acceptance on rape labeling or on victim-blaming between impacted and unimpacted people, which means that none of the groups relied more on their rape myth acceptance when evaluating the case. However, we found that impacted people labeled the uncertain case more as rape and blamed the victim less, compared to unimpacted people, while there was no such difference between the groups in connection with the certain rape scenario. This means that impacted people interpreted those uncertain situations as rape similarly to the certain one, possibly because of their personal experi ence with rape and the knowledge that rape is often not stereotypical. However, these differences between the groups were not present when the case was certain which both impacted and unimpacted people interpreted similarly.

General Discussion
The aim of our research was, on the one hand, to examine the relationship between rape myth acceptance and prior experience with rape, and on the other, to show that rape myth acceptance predicts different evaluations of different rape scenarios. Unsupportive reactions to rape survivors are not only a global public health problem but a human rights violation too. The cause of the high perpetration and latency rates is that sexual violence is a systematic problem connected to society's views about rape. Therefore, individual attitudinal aspects of rape myth acceptance can only be considered by taking the social system and cultural norms regarding rape into account. Examining rape myth acceptance of rape impacted people was a first step toward understanding the problem in a more complex way.
In Study 1 and 3 we found that people with personal experience with rape endorse rape myths less than unimpacted people. In Study 1 survivors and impacted people separately and in Study 3 impacted people together with survivors accepted rape myths less than unimpacted people. However, our cross-sectional method did not give us information about causality. On the one hand, it is possible that rape survivors share their trauma with people who endorse rape myths less, knowing that they will be more understanding and offer better help. On the other hand, if people learn that a friend or close relative of theirs became a victim of rape it may decrease their rape myth accept ance. This is because rape is more likely to be an event that is counter-stereotypical or uncertain, contradicts the idea of "real rape" and strengthens that rape is more of a violent crime than a sexual one and it is not enjoyable to the victim. Although it is important to mention that people who said that they had known someone who had been raped already accepted the fact that it was rape, while people high in rape myth acceptance may report that they do not know anyone who was raped because they label incidents less as rape. It is for this reason that we wanted to test whether the level of rape myth acceptance would predict rape labeling. Rape does not usually happen in dark alleys and by deviant perpetrators, therefore we examined whether the evaluation of counter-stereotypical and stereotypical (uncertain vs. certain) rape cases were related to the participant's rape myth acceptance. We found, in line with previous research (Nyúl et al., 2018), that people with higher rape myth acceptance labeled both cases less as rape and blamed the victim more, and these people evaluated uncertain rape cases less harshly than certain ones. Furthermore, our results offer an empirical explanation for the contradiction that although people think that rape is unacceptable and suggest that they would punish the perpetrator severely (Virág & Kó, 1998), they also often blame the victim and refuse to label sexual assaults as rape. Our results clearly showed the different evaluations of certain and uncertain rape cases which offers an explanation for these contradicting opinions. Importantly, the different evaluations were predicted by rape myth acceptance, which shows that rape myth acceptance is particularly important when a rape case is not stereotypical which is in accordance with previous research (Nyúl et al., 2018). This connection is extremely important because in real life uncertain or non-stereotypical rape cases are much more common than certain ones (e.g., reported to the police, the perpetrator admits the rape), therefore, supposedly people can often rely on these rape myths.

Limitations and Future Directions
Although not a limitation of the study, it is important to note that we collected the data for Study 1 and Study 2 before the #MeToo campaign which received widespread attention worldwide (Kovács & Szémann, 2018). While rape cases were often reported relying on rape myths previously, the social context has undoubtedly changed after the #MeToo campaign in Hungary as well, which may translate to general changes in attitudes toward sexual assault and rape that we could not account for in this study.
Furthermore, in Study 2 we were not able to randomize rape scenarios for technical reasons, therefore there is a chance that this influenced our results. It is possible that when participants read the certain rape case, they evaluated the uncertain in the light of the previous one and this is the reason that they evaluated it less harshly than the certain one. However, as we mentioned earlier, we ran an additional analysis which suggested that the results were not a consequence of the order effect only (see S2 Text in the Supplementary Materials). Another limitation of Study 2 is that we did not pretest the scenarios for whether they evoked the same emotions in participants. However, we think that reading about an uncertain vs. a certain rape case might evoke different emotional reactions, but we suppose that this difference would also be true in real life situations. Therefore, we think that although these different emotions could influence our results, this would be in accordance with the different emotions in real life. Furthermore, in the uncertain rape vignette, we did not use the term rape, but we wrote that "[the perpetra tor] had sex with [the victim]". We used this term intentionally because we wanted to imitate the unfortunate descriptions of real-life rape cases and leave the interpretation to participants. Admittedly, this terminology may have influenced responses and have been upsetting for some to read which we aimed to temper in the debriefing.
An important additional asset of these studies is that we tested rape myth acceptance and evaluation of rape cases in an underrepresented region of social psychological research, and especially of research on rape and rape myths. This region is not only underrepresented in these research areas, but the level of sexism is higher and gender equality is lower in Hungary than in the US or in Western Europe (World Economic Forum, 2016) where most studies related to rape myths had been carried out. This is important because there is no social psychological study preceding ours in this social context that examined rape survivors and impacted people, or that examined their rape myth acceptance and the evaluation of rape cases. Understanding how impacted people think about rape in general and in specific situations is important for helping survivors to process their trauma, and to use this knowledge in rape education programs.
We found that rape survivors, in general, accepted rape myths less than unimpacted people. These results suggest that research should further examine rape impacted people because they can function as advocates for rape survivors. Furthermore, this may be a fruitful area for future research to investigate whether survivors share their trauma more frequently with people with lower rape myth acceptance or the knowledge that a close person became the survivor of rape decreases rape myth acceptance. The relevance of understanding the reaction of rape impacted people was that they can offer the most direct social support for survivors and can engage in collective action even more effectively than survivors (see Drury & Kaiser, 2014).
Our research supported the hypothesis that rape myths function as cognitive schemas and are connected to how people evaluate rape cases. However, we did not find an interaction between rape myth acceptance, the certainty of the case, and prior experience with rape which means that the connection between evaluating rape cases by impacted and unimpacted people was not dependent on their level of rape myth acceptance. Future research should investigate the possibly more complex connection between these variables than this study could account for. Nevertheless, our results give an explanation for the phenomenon of why people judge rape perpetrators harshly and at the same time blame victims and explain away rape and also highlight the important role of rape impacted people in bringing about change in connection with rape in society.

Funding:
The authors have no funding to report.