Introduction

Bystander behaviour in sexual violence has received a significant amount of research interest, mainly in countries in the Global North (Labhardt, et al., 2017; Mainwaring, et al., 2022). Bystanders are often present before, during, and after sexual violence takes place, and have the potential to intervene to help. However, bystanders frequently fail to act due to various individual, contextual, and situational factors (Haikalis, et al., 2018; Mainwaring, et al., 2022; McMahon, 2015). According to a model by Burn (2009), there could be five main reasons that prevent bystanders from intervening: (i) failure to notice the event (ii) failure to identify the event as risky, (iii) failure to take responsibility, (iv) failure to intervene because they lack skills, and (v) failure to intervene because various factors related to audience inhibition. One potential factor influencing these five barriers is a combination of individual and contextual influences - rape culture (i.e., cultural norms around traditional gender roles, sexism, adversarial sexual beliefs, hostility towards women, and acceptance of violence) and individual’s acceptance of myths around rape (e.g., Johnson & Johnson, 2021; Mainwaring, et al., 2022). When sexual violence is normalised and accepted as part of inherent differences between men and women, bystanders may fail to act (Mainwaring, et al., 2022). In this study, our aim is to explore the relationships between rape myths and bystander barriers in sexual violence in the unique cultural context of Guatemala, a country in Central America.

Guatemala is an interesting country to study due to high acceptance of rape culture where violence against girls and women is normalised as part of the social fabric of “machista” culture (see Gibbons & Luna, 2015; Menjivar & Drysdal Walsh, 2017). Although there are a few reports of sexual violence in the university context in Guatemala (Saénz de Tejada & Buitrago Novoa, 2019; Véliz & Valenzuela, 2020), detailed statistics of occurrences, and the role of bystanders are missing. High levels of violence, coupled with gender inequality (Romero, 2021; World Economic Forum, 2021) may result in fewer opportunities for girls and women in the educational sector (Tarallo, 2019). Understanding the associations between rape culture and bystander behaviour in universities in Guatemala has the potential to contribute to gender equality, one of the priorities for the United Nations development goals.

Acceptance of myths around rape can contribute to hostility against women both at a cultural and an individual level. At a cultural level, rape myths can be understood as widespread sexist beliefs based on gender norms in patriarchal societies, justifying male sexual violence towards women (Lonsway & Fitzgerald, 1994). At an individual level, rape myths relate to blaming the victims (e.g., the way they dress/behave; think that the victims are lying), and excusing the perpetrators and situations (e.g., use of alcohol; male sex drive; lack of clarity in communicating consent; e.g., McMahon & Farmer, 2011). For example, a victim may be considered less worthy of help if they are dressed in a sexualised manner (Gramazio, et al., 2021) or are under the influence of alcohol (Pugh, et al., 2016; Rizzo, et al., 2022). This could be partially due to perceptions that that the victim is purposefully indicating sexual intentions with their own behaviour and are therefore “asking for it”. Interestingly, rape myth acceptance has also been associated with higher perception of consent when the victim is dressed in a provocative manner (Lofgreen, et al., 2021). These kinds of perceptions could prevent a bystander from intervening in a situation of sexual violence.

There has been some research linking rape myths to bystander behaviour in sexual violence. The findings suggest that higher endorsement of rape myths relates to more perceived difficulties and lower intentions/willingness to help (Brown & Messman-Moore, 2010; Diener O’Leary, et al., 2022; Kania & Cale, 2021; Leone, et al., 2021; Lyons, et al., 2022a; McMahon, 2010; Yule, et al., 2022). More specifically, rape myth acceptance has been linked with a failure to take responsibility (Jozkowski, et al., 2021; Lyons, et al., 2022a; Martini & De Piccoli, 2020), a failure to identify risk, as well as skills deficits and audience inhibition (Lyons, et al., 2022a). Despite the recent advances in understanding the associations between rape myths and bystander behaviour, there are some significant gaps in the literature that our study aims to address.

Most of the extant studies have not considered the multifaceted nature of rape myths in the five barriers to bystander help identified by Burn (2009; failure to notice, take responsibility, identify risk, skills deficits, audience inhibition). Rape myths can be understood in terms of four factors: “She asked for it” (i.e., rape happened because the victim dressed/behaved/communicated in a certain way); “He did not mean to “ (i.e., uncontrollable sex drive, alcohol consumption); “It wasn’t really a rape “ (i.e., no physical marks, no resistance from the victim); “She lied” (i.e., victim manipulating the situation to get back to a guy, claiming rape when regretting sex, or cheating on a boyfriend; McMahon and Farmer, 2011). One study on incoming U.S. college students found that all the above myths related to lower intentions to intervene as a bystander (McMahon, 2010), and another demonstrated that in an Italian sample, “She asked for it” related to lower intervention intentions in both sexes, and “It wasn’t really a rape” related to lower intentions only in men (Martini & De Piccoli, 2020). However, these two studies looked at only generic intervention intentions, rather than barriers at different stages of intervention. Based on the knowledge of victim blame and ambiguity of the situation as barriers (e.g., Pugh, et al., 2016), we would expect that “She asked for it” and “It wasn’t really a rape” will have an association at least with the bystander steps of “Failure to take responsibility” (due to thinking the victim is not worthy of help) and “Failure to notice” (due to uncertainty around what sexual violence is), respectively. In this study, we aim to increase the knowledge of rape myths as a bystander barrier by taking into consideration the multi-faceted nature of both rape-myths, and bystander barriers.

In addition to investigating the role of rape myths in bystander behaviours, we will add gender and age as control variables. Generically, women have lower rape myths and higher intentions to intervene than men do (Kania & Cale, 2021; Labhardt, et al., 2017). However, there are variations in different countries. For instance, in a study in Ecuador, women reported a lower likelihood to intervene in sexual violence (Lyons, et al., 2022b), and in another study, Nigerian women held higher rape myth attitudes than their male counterparts (Fakunmoja, et al., 2021). Thus, gender is an important variable to consider because it could influence the results differently depending on the country. In addition, age is another factor that could potentially influence the results. Although most studies have not found an age effect (Mainwaring, et al., 2022), some studies have found that younger (Kania & Cale, 2021; Moschella-Smith, et al., 2022) or older (see Mainwaring, et al., 2022) age facilitates bystander behaviour in sexual violence. Because there are no previous investigations of rape myths and bystander barriers in Guatemalan university context, we will control for the gender and age in our analyses.

In summary, this study adds to the research on multi-factorial rape myths and bystander barriers by investigating these associations in Guatemala, controlling for age and gender. Based on the studies discussed above, we expect that rape myths relate to multiple barriers in both males and females. However, due to the scarce literature on multifaceted rape myths and the five bystander barriers, we will not make specific predictions for the relationship between each factor.

Method

Participants and Procedure

The survey, titled ¿Nos damos cuenta del acoso? Experiencias de espectadores frente al acoso sexual en la USAC” (“Do we notice harassment? Experiences of bystanders in the face of sexual harassment in the USAC”) was advertised via snowball sampling on social media platforms (i.e., Instagram, Twitter, Facebook) of the researchers and the accounts that were created for the project. The only inclusion criteria was that the participants had to be currently working or studying at University of San Carlos de Guatemala (USAC), which is the only public university in Guatemala with around 13 000 staff and 225 000 students. Because sexual violence is widely spread around different campus spaces of USAC, all University community members are potential bystanders (Saénz de Tejada & Buitrago Novoa, 2019). Hence, we decided to include both staff and students as participants in the study. The survey was completed by 1009 participants from diverse disciplines (Mean age 25.37, SD = 10.03; 650 identified as female, 351 as male, 8 as other; 109 staff, 898 students, 2 missing information). Participants first read the participant information sheet and provided online consent. Following completion of the questionnaire, participants were directed to a debriefing page. The study received ethical approval by the Institutional Review Board of Liverpool John Moores University (UREC reference: 22/PSY/017).

Materials

All the materials were translated from English to Spanish. Rather than using direct translations of the English scales, the Guatemalan team went through each item in each questionnaire carefully and worded them so that they make sense in the university context in Guatemala. The items were then reviewed by individuals who are involved in sexual violence prevention work. This type of collaborative translation process is considered as a golden standard approach (e.g., Valdez, et al., 2021), and can bypass some of the limitations that back translation processes have (Ozolins, et al., 2020).

For investigating bystander barriers, we used the 16-item Bystander Barrier Scale (Burn, 2009). The questionnaire has five subscales, which are calculated by averaging the items on each subscale. In order to make the questions relevant to the university campus, we changed them slightly. For example, questions starting “At a party or bar.” were changed to “in the classroom or other university spaces…”. Participants rated all the items on a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree). The subscales are (i) Failure to Notice (one item; “If I am in a classroom or other space on campus, I would probably be too busy to notice when someone is at risk for sexual assault”), (ii) Failure to Identify Situation as High Risk (three items; e.g., “If I am in a classroom or other space on campus, I think I might be uncertain as to whether someone is at-risk for being sexually assaulted, α = .59; the alpha was just below the threshold of being acceptable, so the results should be treated with caution), (iii) Failure to Take Intervention Responsibility (eight items; e.g., “If I saw someone I didn’t know was at risk for being sexually assaulted, I would leave it up to his/her friends to intervene”, α = .79), (iv) Failure to Intervene Due to a Skills Deficit (two items; e.g., “Although I would like to intervene when a guy’s sexual conduct is questionable, I am not sure I would know what to say or do”, Pearson’s r = .66), and (v) Failure to Intervene Due to Audience Inhibition (two items; e.g., “I am hesitant to intervene when a man’s sexual conduct is questionable because I am not sure other people would support me”, α = .78).

For investigating rape myth acceptance, we used the 22-item updated version of the Illinois Rape Myth Acceptance scale (McMahon & Farmer, 2011). We changed the questions so that they reflect the university environment, and adult individuals rather than “boys and girls” (e.g., “If a girl goes to a room alone with a guy at a party, it is her own fault if she is raped” was changed to “If a woman goes to isolated places in the university, it is her own fault if she is raped”). The Likert-options varied from 1 = totally disagree to 5 = totally agree. We analysed the scale in terms of four subscales where the questions for each subscale were summed and averaged. We created variables for “She asked for it” (six items; e.g.,“When women are raped, its often because the way they said “no” wasn’t clear’, α = .76); “He didn’t mean to” (six items; e.g., “When men rape, it is usually because of their strong desire for sex”, α = .83); “It wasn’t rape” (five items; e.g., “A rape probably didn’t happen if the woman has no bruises or marks”, α = .85); and “She Lied” (five items; e.g., “A lot of times, women who say they were raped agreed to have sex and then regret it”, α = .88). All the Spanish versions of the translated questionnaires are available from the first author upon request.

Results

The data met the assumptions for parametric statistics in terms of normality and skewness. Initial t-tests indicate that men were more accepting of all rape myths compared to women (She asked for it: t(994) = 5.87, p < .001, d = 0.39; male M = 2.41, SD = 0.91, female M = 2.10, SD = 0.85; He didn’t mean to: t(994) = 6.18, p < .001, d = 0.41, male M = 2.32, SD = 0.99; female M = 1.92, SD = 0.95; It wasn’t rape: t(994) = 6.43, p < .001, d = 0.43, male M = 1.82, SD = 0.91; female M = 1.49, SD = 0.72; She lied: t(994) = 8.24, p < .001, d = 0.55, male M = 2.89, SD = 1.05; female M = 2.33, SD = 1.02). For the perceived barriers to bystander intervention gender differences were significant only for Failure to take intervention responsibility, t(997) = 5.41, p < .001, d = 0.36, with men more likely to fail to take responsibility for intervention action compared to women male (male M = 2.51, SD = 0.76; female M = 2.26, SD = 0.70). The gender differences for the other four bystander barriers were not statistically significant (all p’s >. 15).

Table 1 shows that all the rape myths had significant, positive correlations with all the five bystander barriers. The correlations were similar for both men and women and are not reported separately here.

Table 1 Zero-order Correlations and Descriptive Statistics for Age, Barriers to Bystander Intervention, and Rape Myths
Full size table

We conducted hierarchical linear regressions for each bystander barrier, where gender and age were added as predictors at first stage, and the four rape myths were predictors at the second stage. For Failure to notice, gender and age did not contribute significantly to the variance at stage 1, F(2,1000) = 2.41, p = .09. The stage 2 model was significant, F(6,996) = 12.36, p < .001, explaining 6.90% of the variance (R2 = .07, Adj R2 = .06). When all variables were entered into the model, ‘She asked for it’ (B = .13, t = 3.11, p = .002) and ‘It wasn’t rape’ (B = .10, t = 2.28, p = .023) were significant individual predictors, indicating that that those who accepted these rape myths were more likely to fail to notice the harassment.

For Failure to identify a situation as high risk, the stage one model containing gender and age only was not significant, F(2,1000) = 1.21, p = .299. The stage 2 model was significant, F(6,996) = 9.24, p < .001, explaining 5.30% of the variance (R2 = .05, Adj R2 = .05). When all variables were entered into the model, ‘She asked for it’ (B = .12, t = 2.82, p = .005) and ‘It wasn’t rape’ (B = .14, t = 3.29, p = .001) were significant individual predictors, such that those accepting these rape myths were more likely to fail to identify a situation as high risk.

For Failure to take responsibility, the model containing gender and age was a significant predictor, F(2,1000) = 18.62, p < .001. The stage 2 model including the four rape myth subscales was also significant, F(6,996) = 46.06, p < .001, explaining 21.7% of the failure to notice variance (R2 = .22, Adj R2 = .21). When all variables were entered into the model, gender (B = − .07, t = -2.41, p = .016), age (B = .07, t = 2.36, p = .019), acceptance of the ‘She asked for it’ (B = .26, t = 6.51, p < .001), ‘He didn’t mean to’ (B = .11, t 2.54, p = .011), and ‘It wasn’t rape’ (B = .12, t = 2.93, p = .003) were significant individual predictors. Men, older participants, and those accepting these rape myths were more likely to fail to take responsibility for action.

For Failure to intervene due to a skills deficit, the stage one model containing gender and age was significant, F(2,1000) = 5.46, p = .004. The stage 2 model including the four rape myth subscales was significant, F(6,996) = 8.87, p < .001, explaining 5.10% of the failure to notice variance (R2 = .05, Adj R2 = .05). When all variables were entered into the model, age (B = − .11, t = -3.48, p < .001), acceptance of the ‘She asked for it’ (B = .18, t = 4.07, p < .001) and ‘It wasn’t rape’ (B = .11, t = 2.47, p = .014) were significant individual predictors. Younger participants, and those accepting these rape myths were more likely to fail to intervene due to a perceived skills deficit.

At stage one, gender and age did not contribute significantly to Failure to intervene due to audience inhibition, F(2,1000) = 2.36, p = .10. The stage 2 model including the four rape myth subscales was significant, F(6,996) = 12.00, p < .001, explaining 6.70% of the failure to notice variance (R2 = .07, Adj R2 = .06). When all variables were entered into the model, age (B = − .08, t = -2.47, p = .014), ‘She asked for it’ (B = .15, t = 3.41, p < .001) and ‘It wasn’t rape’ (B = .17, t = 3.96, p < .001) were significant individual predictors, such that younger participants and those accepting these rape myths were more likely to fail to intervene due to audience inhibition.

Discussion

The main aim of the study was to explore the role of four rape myths in five bystander barriers outside the US/Global North context. Our results echo the findings from other countries (e.g., Lyons, et al., 2022a; 2022b; Martini & De Piccoli, 2020; Yule, et al., 2022), demonstrating that rape myth acceptance is an important barrier for bystanders in a public university in Guatemala, explaining significant amount of variance of in most of the five bystander barriers identified by Burn (2009). For the first two bystander barriers (Failure to notice and Failure to identify a situation as high risk), rape myths “She asked for it” and “It wasn’t a rape” were significant positive predictors. For the Failure to take responsibility, “She asked for it”, “It wasn’t a rape”, “He did not mean to”, as well as male gender and older age were significant predictors. For the barriers of Skills deficits and Audience inhibition, younger age and “She asked for it” and “It wasn’t a rape” were significant.

Of all the rape myths considered, acceptance of ‘She asked for it’ and ‘It wasn’t rape’ were especially important, predicting all the five barriers to bystander intervention. A previous study with Italian participants found that victim blame and situational ambiguity prevented bystander intentions (Martini & De Piccoli, 2020). Our research took into consideration specific barriers rather than generic intentions, demonstrating how these same myths are related to multiple barriers in Guatemala.

The theme “She asked for it” makes sense in the patriarchal social fabric of Guatemala where sexually active women are viewed negatively (Singleton, et al., 2016), and sexual coercion is a form of gendered violence that is normalised (Duffy, 2018). In cultural scripts where traditional gender roles are organised around machismo (i.e., men who are domineering/controlling, and aggressive) and marianismo (i.e., women who serve their families, and exhibit moral and spiritual superiority to men), it is not uncommon to blame the female victims for any aggression directed towards them (Amaya & Gray, 2021). In our study, victim blaming was related to all the barriers to intervention, but the mechanisms of the relationships are not clear.

There could be multiple context-specific reasons why victim blame was an important barrier in Guatemala. For example, Failure to take responsibility could be related to victim blame due to perceptions that the victim lacks morality and is culpable for inviting the abuse with their own behaviour (e.g., Gramazio, et al., 2021; Pugh, et al., 2016; Rizzo, et al., 2022). Alternatively, individuals may fail to act because of their faulty beliefs that the victim is giving consent to be abused (Lofgreen, et al., 2021). Failure to identify the situation as high risk could be related to victim blame because the individual lacks knowledge of what sexual violence is. Audience inhibition could be due to the fear of “losing face” in defending someone, especially if local peer norms are misogynistic and supportive of victim blame (see Leone, et al., 2017). It would be beneficial to deepen the understanding of these relationships in the university setting in Guatemala by (i) conducting qualitative studies (see, for example, Hackman, et al., 2017), (ii) designing questionnaire studies with moderator variables (e.g., perceptions of morality; perceptions of peer-norms), or (iii) conducting experiments (e.g., manipulating victim blame, and measuring the bystander barriers).

With regards to the importance of “It wasn’t a rape” as a barrier to bystander behaviour, it seems like the lack of knowledge around consent in sexual coercion is something that prevents bystanders from helping. Bystander interventions can be designed to simultaneously address rape myths and knowledge around sexual consent (Salazar, et al., 2014), and based on our findings, educating participants about consent would be a beneficial addition to interventions designed in Guatemala. The country is lacking a formal, mandatory, school-based sex education, which could be useful in terms of teaching about sexual consent (Monzón, et al., 2017), facilitating bystander behaviour as a by-product.

Excusing the male perpetrator (i.e., He did not mean to) related to one important barrier- Failure to take responsibility. Excusing the perpetrator for their actions has been related to benevolent sexism (i.e., support for traditional gender roles, requiring men to fill the role of protector and provider; Rollero & Tartaglia, 2019), which has also been linked to higher perceived failure to take responsibility (Yule, et al., 2022). In future studies, it would be interesting to investigate whether excusing the perpetrator is a potential moderator between benevolent sexism and failure to take responsibility to intervene in sexual violence.

Interestingly, the rape myth “She lied” was not related to any of the bystander barriers in our sample. The idea that women lie about rape is a common myth that contributes negatively to credibility and trustworthiness of rape victims (e.g., Stabile, et al., 2019), and has been related to lower willingness to intervene in a US sample (McMahon, 2010). Our findings are similar to Martini and De Piccoli (2020), who did not find correlations between bystander intention to intervene, and the rape myth “She lied” in their sample of Italian participants. The authors speculated that this rape myth could be more important in justifying violence after it has taken place, but less important in bystander behaviour before, and during an incident. Future studies could investigate whether “She lied” is a bystander barrier when providing support for a victim after the event has taken place.

With regards to age, one of our control variables, younger participants were more likely to fail to intervene due to a perceived skills deficit or audience inhibition, but less likely to have the barrier of failing to take responsibility. It is possible that getting involved in a situation with other people comes with a high risk in Guatemala, especially as the country has high levels of homicide and gun violence. Uncertainty about what to do when observing harassment could affect younger people even more than older individuals, as the risk of being a victim of violence is higher in this population (Pineda & Bolaños, 2009). In addition, younger students may seek social acceptance from others, and bystander behaviour comes with potential social costs which could relate to the audience inhibition (Bennett, et al., 2014). It may be necessary to think about slightly different bystander approaches to different demographic groups in Guatemala. This could facilitate the removal of barriers that are more typical to people of different ages.

Unlike in many other studies, we found very few gender differences in the barriers. Men were more likely to fail to take responsibility, but there were no statistically significant gender differences in the other barriers. Previous studies have demonstrated that women have generally higher willingness and efficacy to intervene in sexual violence, although these findings have not been replicated consistently from one sample to another (Labhardt, et al., 2017; Mainwaring, et al., 2022). However, in Ecuador, women reported more intervention difficulties (Lyons, et al., 2022b), and a study with participants from Indonesia, Philippines, and the United Kingdom did not find any gender differences in the five barriers (Lyons, et al., 2022a). This underscores the importance of avoiding generalisations of findings from the U.S. samples to other contexts and highlights the importance of investigating bystander barriers in more diverse regions of the world (see also Labhardt, et al., 2017). Gibbons and Luna (2015) discussed a “rising tide” of growing gender equality in the Central American region, and how especially in Guatemala, male university students are increasingly rejecting gendered ideas relating to traditional machismo. It is possible that men and women in the public university in Guatemala are more similar than different to each other, and this is reflected in the mostly insignificant results regarding gender in bystander barriers. However, the rape myths still had significant gender differences, warranting more investigations into why the differences did not persist in bystander barriers.

Although our study provides interesting preliminary results for rape myths and bystander behaviours in Guatemala, there are several limitations that warrant discussion. First, we utilised etic instruments with Likert-style scales. Using questionnaire responses in Likert-format might not be the most efficient way of collecting accurate data in contexts outside the Global North (e.g., Hruschka, et al., 2018). Second, it is possible that the bystander barriers in our questionnaire are not truly reflecting the kind of situations that University students in Guatemala find themselves in when witnessing violence. We recommend a qualitative approach for investigating the circumstances of sexual violence, and barriers for bystanders. The qualitative responses could lead to the development of emic instruments that would be more appropriate for investigating bystander barriers in Guatemala. Third, we only had a limited number of variables in the study, leaving us to speculate about some of the findings (e.g., relationship between gender roles, victim blame, and bystander barriers). Future studies could add a scale on machismo/marianismo (e.g., Terrazas-Carrillo & Sabina, 2019) to see how conformity with traditional gender roles relates to both rape myths and bystander barriers.

To summarise, our study provided initial evidence for the relationships between rape myth acceptance and bystander barriers in sexual violence in a public university in Guatemala. The results suggest that addressing the myths around victim blame and knowledge of consent would be especially important when designing bystander interventions suitable for this context. We urge researchers from diverse backgrounds to pursue the investigation of barriers and facilitators in order to aid in the development of context-specific interventions.