Sexual Assault in a ‘Just World’: An Immersive VR Study of Bystanders

Sexual Assault in a ‘Just World’: An Immersive VR Study of Bystanders | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Sexual Assault in a ‘Just World’: An Immersive VR Study of Bystanders Sarah Warbis, Danaë Stanton Fraser, Mitchell Callan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6008933/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This paper explores how the need to believe in a just world (NBJW) affects bystanders’ intentions to intervene when witnessing a sexual assault. More specifically, we explored how non-rational strategies used to resolve a threat to the belief in a just world (BJW) are related to Latané and Darley’s (1970) bystander decision-making process, and whether empathic variables mediate this. In Study 1, the threat to the BJW was indirectly manipulated by varying the sexual assault severity described in a text vignette. Measures assessing victim derogation, blame, psychological distancing from the victim, empathy, empathic concern and the bystander process were completed by 294 participants. The assault severity had no significant effect on BJW strategies, but increased intervention intentions. There was a lack of evidence for the role of victim derogation and blaming, but greater psychological distancing indirectly predicted lower general intervention intentions via reduced empathic concern, and a reduced ability to succeed in several of the bystander stages, especially accepting responsibility to intervene. We replicated this design in Study 2 ( N = 117), but using filmed virtual reality (VR) stimuli. We again found a lack of effect of the severity manipulation beyond increasing the likelihood of intending to intervene. In Study 2, we also found that victim blaming predicted general intentions, but with no clear mediators. Implications are discussed, including how factors like the emotional impact of stimuli and the empathic nature of VR could help to explain some of the differing results found across our studies. Psychology sexual assault bystander intervention virtual reality gender-based violence victim blaming. Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Latané and Darley (1970) suggest that a bystander witnessing a sexual assault goes through a sequential process when deciding whether to intervene. When investigating how to utilise this process to encourage bystanders to intervene, research has largely ignored the need to believe in a just world (NBJW) and its strategies, a concept which explores how individuals restore a sense of justice following an injustice (Lerner, 1980). Across two studies this paper provides an initial exploration of the relationship between the bystander model and NBJW, aiming to explore if strategies used to defend a threat to the BJW affect this bystander process, and through that, the likelihood of the bystander intervening. The study also explores the use of virtual reality (VR) as stimuli compared to more traditionally-used text vignettes in this area of research. Sexual Assault Sexual assault is defined by the Office for National Statistics (ONS; 2023b) as any of the following: rape, assault by penetration using an object other than a penis, unwanted sexual touching or indecent exposure, including attempts of the first two, with no mention of force or threat being necessary. Using this definition, the ONS (2023a) found that in England and Wales, 27% women, 6% men have been the victim of sexual assault since the age of 16 (the age of consent in the UK), with other UK-based research finding a staggering 65% of LGBT+ people (including non-binary and intersex individuals) had been sexually assaulted during adulthood (Galop, 2022). Those at particular risk are adults under 24 years old and full-time university students (ONS, 2023a), with locations off-campus being the most common places for someone of this cohort to be sexually assaulted (O'Neal et al., 2021). Being a survivor of a sexual assault poses both short and long-term risks to physical health, for example, injuries, sexually transmitted diseases and chronic pain, as well as longer-term effects on mental health, including an increased risk of post-traumatic stress disorder (PTSD) and depression (Luce et al., 2010; McFarlane et al., 2005; Zilkens et al., 2017). This is in addition to the social consequences of the victimisation, with victims often being blamed for the assault (Finch & Munro, 2007; Lerner, 1980). Moreover, should the victim wish to proceed with criminal proceedings they stand a 3.7% chance of the perpetrator being charged for a sexual offense in England and Wales (Home Office, 2023). These stark statistics make this a key subject for academic, third sector and policy preventative research (Fenton et al., 2015; Revolt Sexual Assault & The Student Room, 2018; Ullman, 2007). The Role of Bystanders Historically, programmes have put the burden for prevention on potential victims (Katz et al., 2013; Ullman, 2007), whereas more recent efforts have aimed to spread this responsibility to the entire community (Banyard et al., 2004). Research suggests that a bystander is present during up to 29% of sexual assaults, including those taking place at house parties (Haikalis et al., 2018). Yet, the most common response from a bystander is to do nothing (Casper et al., 2022; Haikalis et al., 2018; Hamby et al., 2016; Hart & Miethe, 2008; Orchowski et al., 2022; Planty, 2002; Revolt Sexual Assault & The Student Room, 2018). In order to enable behaviour change, research has been carried out to understand the process bystanders experience when witnessing an emergency event, such as a sexual assault. Latané and Darley (1970) proposed a model which outlined five key steps a bystander goes through in a progressive manner (although cycling back to previous stages is possible). Firstly, a bystander must notice the event, which can be impaired due to distractions or noise (Burn, 2009). Secondly, a bystander must decide that the event is an emergency whereby intervention is needed, which can be facilitated by knowledge of nonconsensual behaviour (Moschella-Smith et al., 2022), and impeded by ambiguity in the situation, such as being unsure if the victim wants help (Shotland & Straw, 1976). Thirdly, a bystander must accept personal responsibility to intervene, which is most notably prohibited by the presence of other inactive, unrelated bystanders (Darley & Latané, 1968), as well as being heightened by a preexisting relationship or shared group membership with the victim, perpetrator and other bystanders (Bennett et al., 2014; Burn, 2009; Fischer et al., 2011; Katz et al., 2015; Levine, 2005). Next, a bystander must decide how to intervene most effectively, which takes into account their level of available skills (Bennett et al., 2014), and their belief in their capability to successfully intervene with the desired outcome, known as self-efficacy (Bandura, 1977; Banyard & Moynihan, 2011; Thornberg, 2013). And lastly, a bystander must implement their intended intervention, and at this stage can be impacted by group norms and concern of how others may judge them (Banyard & Moynihan, 2011; Burn, 2009; Mainwaring et al., 2022). There is an abundance of evidence providing support for the importance of each stage of Latané and Darley’s (1970) model. There is also evidence for the model as a whole, for example Burn (2009) demonstrated correlations between barriers to each stage and real life intervention in a sexual assault context, and Nickerson (2014) demonstrated how each stage was able to predict the next in sequence in a bullying and sexual harassment context in schools. However, no research as of yet has explored the mediated, indirect pathways between the stages in this model in the context of sexual assault. As well as research exploring the importance of the stages of Latané and Darley’s (1970) model themselves in predicting bystander intervention (e.g. Burn, 2009; Nickerson, 2014), separate factors affecting the model have been studied, such as the presence of others and social norms (Banyard & Moynihan, 2011; Darley & Latané, 1968; Fischer et al., 2011; Mainwaring et al., 2022). However, one area which has been largely neglected in this exploration is the NBJW. The Need to Believe in a Just World The just world hypothesis states that there is a universal need to believe that we live in a generally fair world, where people get what they deserve; a belief which is essential for the pursuit of long-term goals (Hafer, 2000; Hafer & Rubel, 2015; Lerner, 1980; Lerner & Miller, 1978). Given that the world around us can be unfair, this need can be easily threatened by seeing something contradicting this belief, for example, by seeing an innocent person being sexually assaulted. Consequently, given its functional importance, individuals are motivated to defend this belief in the face of threat (Lerner, 1980). According to Lerner (1980), individuals can resolve threat to their belief in a just world (BJW) in several ways. One way, is through using a rational strategy in which the individual tries to rebalance the scales by restoring justice to the situation, for example, through compensating a victim or, most crucially to this article, by intervening and stopping the injustice (Harvey et al., 2014; Lerner, 1980). In contrast, people can also use non-rational strategies to resolve the threat to their BJW which involve cognitively distorting the event to deny an injustice occurred to begin with (Hafer & Bègue, 2005). These strategies include, among others, victim derogation, victim blaming and psychological distancing (Hafer & Rubel, 2015; Lerner, 1980). Victim derogation involves belittling the character of a victim in a way that they seem deserving of the bad thing that is happening to them (Lerner, 1980), and has been demonstrated to be enhanced in response to a threat to the BJW (Correia & Vala, 2003; Dawtry et al., 2020). Similarly, victim blaming works by criticising a victim’s behaviour, so that their actions appear to be to blame for the injustice (Lerner, 1980), and indeed personality literature has shown a link between a stronger BJW trait and greater victim blaming of sexual violence victims (Linhares & Torres, 2021; Strömwall et al., 2014). Psychological distancing works in a different way however, being a more subtle method of restoring BJW (Hafer & Rubel, 2015). By highlighting how different you are from the victim through disassociating yourself from them, this removes the concern that if this bad thing can happen to them, it could happen to me, and again has been shown to be higher when the threat to the BJW is greatest (Hafer, 2000). Thus, research supports Lerner’s (1980) theory that individuals will use rational and non-rational strategies to restore their BJW. Despite the clear threat to the BJW that witnessing an injustice like a sexual assault could elicit and the high prevalence of victim blaming experienced by sexual assault survivors (Embury-Dennis, 2018, March 30; Finch & Munro, 2007; Oppenheim, 2022 August 18), very little research has explored how these strategies under the just world hypothesis could be related to bystander intervention when witnessing a sexual assault. Research to date suggests that greater victim blaming is related to lower intentions to intervene in rape, online harassment and injury contexts (Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022). However, there are several pivotal limitations with how confidently one can apply existing findings to understand sexual assault bystanders. Firstly, research has often looked at the prevalence of victim blaming after a crime has been committed, for example as a court case. In such cases, rational strategies like intervention, which research suggests would be the preferred option (Lerner & Simmons, 1966), are no longer possible, meaning victim blaming rates could be inflated (Finch & Munro, 2007; Spaccatini et al., 2022). Secondly, a large body of literature in this field conceptualises the BJW as an individual differences trait upon which individuals vary, rather than as the universal NBJW as Lerner (1980) intended (Hafer & Bègue, 2005), questioning whether this personality literature can be generalised. And most crucially, research to date has not fully explored how multiple BJW strategies are related to Latané and Darley’s (1970) entire model, instead cherry picking individual concepts (Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022). All of which combined highlights the need for a comprehensive exploration of how multiple strategies which restore the BJW could be related to all stages of Latané and Darley’s (1970) model within the context of a bystander witnessing a sexual assault occurring. While there is limited research to date, it is possible to make some informed predictions of what this relationship might look like. When considering which strategies were preferred, research suggests that people are inclined to use rational strategies over non-rational strategies to resolve a threat to their BJW (Lerner & Simmons, 1966), but that non-rational strategies are used when the individual does not believe they can fully restore justice to the situation (Montada & Lerner, 1998; Walster et al., 1973; White et al., 2012). When a bystander is witnessing a sexual assault already occurring, a bystander may be aware of the negative impacts the assault has likely already had on the victim’s mental and physical health (Carey et al., 2018; Luce et al., 2010; Mason & Lodrick, 2013; McFarlane et al., 2005), and as such it could be suggested that they would view the victim’s suffering as irreparable. Based on this, it could be argued that they may decide that intervention is futile and resort to non-rational strategies (Montada & Lerner, 1998; Walster et al., 1973; White et al., 2012). If proceeding with non-rational strategies, it could be hypothesised that victim derogation and blaming could negatively impact a bystander’s ability to see an assault as an emergency in which intervention is needed. If the victim is viewed as either a bad person or as having carried out a bad action which enabled something to happen to them, then the bystander may view the event as just, and therefore not requiring external intervention. These strategies could also impair the bystander’s ability to accept responsibility to intervene. After all, if the victim’s character or behaviour is responsible for what is happening to them, then the bystander may decide that they themselves are not responsible and therefore decide not to intervene. Psychological distancing could work in a similar way by hindering a bystander’s ability to accept responsibility to act. Wang (2020) suggested that sharing similarities with someone fosters a sense of responsibility for them, so feeling dissimilar to the victim through psychological distancing may result in the reverse effect and reduce their sense of responsibility to intervene. Therefore, we propose that derogating, blaming and psychologically distancing oneself from a victim, could collectively affect two stages in Latané and Darley’s (1970) model, namely, deciding that intervention is needed and accepting personal responsibility to do so. However, this needs to be tested empirically, so the influence on all stages of the bystander model should be considered. The Role of Empathy and Empathic Concern Two related variables of interest to develop our understanding of the possible mechanisms behind how the NBJW and bystander intervention are related are empathy and empathic concern. Understood here as two distinct constructs (Jordan et al., 2016), empathy is the cognitive ability to take on another person’s perspective (Davis, 1980), whilst empathic concern goes beyond this to be an emotional concern for the welfare of another based on their circumstances (Davis, 1980; Jordan et al., 2016). Research has found links between empathy and perceptions of fairness and lower victim derogation (Aderman et al., 1974; Singer et al., 2006), as well as a link between greater empathic concern and lower victim blaming (Katz et al., 2015), all of which suggests that these empathic variables could be related to the NBJW. In addition to this, research has also shown links between increased empathy and increased helping (Ahn et al., 2013), as well as positive relationships between empathic concern and all five stages in Latané and Darley’s (1970) model (Katz, 2015; Nickerson, 2014; Patil et al., 2018). With empirical evidence suggesting that empathy and empathic concern could play a role within both the NBJW and bystander intervention but little indication of directionality, there is grounds for a novel exploration of whether they are related to both areas in the same sexual assault bystander context, and whether they could act as potential mediators between the influence of the NBJW strategies on Latané and Darley’s (1970) model. Moreover, given that past research has on occasions conflated empathy and empathic concern into one variable (Davis, 1980), or indeed used the terms interchangeably (Nickerson, 2014), despite evidence showing them to be distinct constructs (Jordan et al., 2016), it is important to understand which empathic variable could play a mediating role here. Based on our predictions related to the NBJW, we explored whether empathy and empathic concern mediated the relationship between BJW non-rational strategies and the stages of Latané and Darley’s (1970) model, and in particular, a bystander’s belief that intervention is needed and their ability to accept responsibility to act. Research Methodology for Studying Bystanders Having established a theoretical approach for this research, another decision relates to the research methodology. Dawtry et al. (2020) found that inconsistencies in findings of experimental research on victim derogation could be explained by differences in the emotional impact (i.e. emotional engagement or threat) of the context for observers. Studies using methodologies that were more vivid, severe, felt more real, and were physically and chronologically closer, reported larger derogation effects. Thus, it is important to explore our hypothesised relationships using emotionally impactful stimuli. Past BJW research has utilised emotionally impactful methodologies like witnessing in-person suffering (Fine, 1983; Latta, 1976), watching CCTV footage (Lerner & Simmons, 1966), and watching recorded interviews with victims (Warner et al., 2012), but there are key limitations to each of these. Firstly, conducting a sexual assault bystander experiment featuring an in-person experience would present serious ethical concerns, especially as participants would likely believe this event to be real and would not have an opportunity to withdraw from the study if they found it distressing. Secondly, while CCTV footage and recorded interviews have high realism, both feel like a distant event physically and/or chronologically. However, there is another methodology gaining popularity in social psychological research which would score highly on all four of Dawtry et al.’s (2020) emotional impact criteria, namely, virtual reality (VR). VR was not included in Dawtry et al.’s (2020) meta-analysis. VR involves artificial sensory information being used to perceptually surround a user, often using a head-mounted display (Blascovich, 2002). VR has already been used effectively in both bystander and sexual offense research (Jicol et al., 2022; Slater et al., 2013 ). For example, Slater et al. (2013) exposed participants in VR to a violent attack between two men to investigate if bystanders were more likely to intervene when the victim of the attack was a fan of the participant’s own football team (i.e. their ingroup). Using VR enabled the researchers to test a scenario that may not have been ethical to stage in real life nor easy to control, while also allowing for greater mundane realism than a vignette in a traditional experiment. Research has demonstrated the strong emotional impact of VR stimuli. Jicol et al. (2022) exposed participants to a sexual harassment victimisation simulation either on a flat-screen monitor or in VR, finding that participants had a strong negative emotional response in VR, as would be expected if being harassed in real life, and that this response was greater than the response to the flat-screen simulation. Moreover, because of the VR setting, they were able to measure this emotional impact both through self-reported measures and more objectively through physiological measures using heart rate (HR) as an indicator. Patil et al. (2014) similarly found that a moral dilemma was more emotionally arousing, measured through skin conductance, when presented in virtual reality, compared to when presented in a text format. Other research has shown VR to be capable of eliciting feelings of emotions like panic and positive affect in general, showing its versatility as an instrument (Baños et al., 2008; Pan & Slater, 2011). Research has also found this to be the case when using a specific type of VR, namely, filmed virtual reality (Kahn & Cargile, 2021). Filmed VR involves recording a 2D video using a 360-degree camera and displaying this footage on a head-mounted display. Although animated VR is more widely used in research, filmed VR holds the benefits of time and cost efficiency, requiring much less training and specialised knowledge to develop than animated VR (Barreda-Ángeles et al., 2021). Kahn and Cargile (2021) found that filmed virtual reality led to greater feelings of awe than watching the same content on a flat-screen. Moreover, Kahn and Cargile (2021) found that this effect was fully mediated by an increased sense of presence when watching the scene in VR. Presence refers to the participant’s illusionary sense of being there in the virtual environment (Slater, 2018; Slater et al., 1994). Indeed, Jicol et al. (2023) and Baños et al. (2008) found similar positive relationships between emotional intensity and presence in animated VR. A benefit of this relationship is that, in a similar way to emotional intensity (Dawtry et al., 2020), greater presence has been linked to behavioural responses we would expect to see in real life. For example, a replication of Milgram’s infamous obedience shock experiment, but this time conducted in immersive VR, found the same pattern of results as the original study (Slater et al., 2006). Van Gelder et al. (2017) also found that in a VR burglary simulation, users reported high levels of presence, and factors which would be expected to be related to stealing behaviour in real life, like the level of deterrence, consciousness and sensation seeking, were related to stealing behaviours in VR too. Thus, feelings of presence in VR, as well as its greater emotional intensity, appears to be beneficial for eliciting behaviours we would expect to see in real life. Therefore, VR could be a highly suitable method for studying the effect of the NBJW on bystander intervention due to its high emotional impact and presence, while maintaining a controlled and safe setting for participants. Current Research We tested the model outlined in Fig. 1. Within this model, a sexual assault in the presence of a bystander causes a threat to the BJW, which in turn results in the use of non-rational strategies of derogation, blame and/or psychological distancing from the victim. In turn, these strategies may result in a reduction in empathy or empathic concern. Meanwhile, the sexual assault means a bystander should be engaged in the five stages outlined by Latané and Darley (1970). The use of non-rational strategies and corresponding reduction in empathy and empathic concern may in turn result in the individual being less likely to believe that intervention is needed and a reduced sense of responsibility to act, in turn, reducing their likelihood of intervening. As outlined above, research provides some evidence linking the aspects of this model together. However, no research to date has explored it as a whole, and through the lenses of both Latané and Darley’s (1970) bystander model and multiple BJW non-rational strategies from Lerner’s (1980) theory. Due to the lack of NBJW research using VR, another aim of this research was to compare the effects of a threat to the BJW on bystander intervention intentions across two studies using different types of research stimuli, comparing a traditional text vignette in Study 1 to a filmed VR scene in Study 2. Taken together, these two studies allow us to test if the model shown in Fig. 1 holds across stimuli varying in emotional impact and presence. Study 1 Introduction In this first experiment, the severity of the sexual assault, which participants read about in a text vignette, was manipulated. This previously established manipulation was chosen as research suggests that the threat to the BJW is greater when the injustice is more severe (Dawtry et al., 2020). After reading the vignette, participants completed scales on all variables of interest discussed above. Hypotheses Based on the literature, several hypotheses were formed. Firstly, we proposed that in the more severe condition, victim derogation, blame and psychological distancing would be greater, empathy and empathic concern would be lower, and there would be a reduced ability to complete each of Latané and Darley’s (1970) stages (e.g. lower acceptance of responsibility to intervene). In-line with Latané and Darley’s (1970) bystander model, we predicted that each stage in the model would predict those following it (including intervention intentions), and that earlier stages would indirectly predict intervention intentions via other stages in the model’s order. For example, we predicted that a bystander’s ability to notice the sexual assault (stage one) would be indirectly predict intervention intentions (both at a general and behaviour-specific level) via an increased perception that the event was an emergency in which intervention was needed (stage two), there would be increased acceptance of responsibility to intervene (stage three) and increased belief in their ability to intervene (stage four). We also predicted the mediation model outlined in Fig. 1. More specifically, we hypothesised that the severity manipulation would have a negative indirect effect on intervention intentions (both at a general and behaviour-specific level) through serial mediation via increased use of BJW strategies, reduced empathy or empathic concern, reduced success in viewing the event as an emergency in which intervention was needed (stage two) and reduced acceptance of responsibility to intervene (stage three). Each element of this hypothesis would be tested in turn. However, pathways via other bystander stages would be explored. And lastly, to confirm whether our two measures of intervention intentions reflected actual intervention behaviour, we predicted that intervention intentions (both at a general and behaviour-specific level) would be correlated with past intervention in real life. Study 1 Method Participants The sample consisted of full-time UK university students aged between 18 and 24, as research suggests that this occupation and age group are the adults most vulnerable to sexual assault (ONS, 2023b). Sensitivity power analysis conducted using GPower software indicated that a sample size of 300 would provide 80% power to detect at least small-to-medium effect sizes ( d = .32, α = .05). In total, the study was completed by 327 students. Thirty-three participants were removed from the dataset for the following reasons: four did not meet the screening criteria (i.e. age and occupation), 25 did not give consent to take part after reading the initial information sheet, and a further four participants failed to strongly agree or agree with an attention check item (“I am paying full attention to this survey”). This left 294 participants in the final sample (141 men, 145 women, 8 non-binary individuals). In line with the screening criteria, participants were aged between 18 and 24 ( M = 20.87, SD = 1.64). 38.44% of the sample reported having previously been sexually assaulted, and 60.88% had been previously sexually harassed. Design A between-participants design was used whereby the severity of the sexual assault was manipulated, as an indirect manipulation of the threat to a BJW. Participants read a text description about a woman, Emily, being sexually assaulted. In the less severe condition ( n = 146), the man in the description grabbed the woman by the waist, whereas in the more severe condition ( n = 148), the man grabbed the woman by the waist and then forced his hands up her dress (for the full text vignette, see the Procedure section below). As outlined in hypotheses one to three, this study examined the effect of this severity manipulation on the stages of Latané and Darley’s (1970) bystander intervention model, BJW strategies, and empathic variables. All other analysis used a correlational design. Measures Descriptive statistics for each measure are provided in Table 1, along with Cronbach’s alphas for multi-item measures. Manipulation Checks. Taken from Dawtry et al. (2018), participants were asked to rate how unfair what happened to Emily was on a scale from one ( slightly unfair ) to seven ( a great deal unfair ). Bystander Stages of Intervention Stages One to Four of Latané and Darley (1970). The four precursor stages of Latané and Darley’s (1970) model were each measured with a single new item, as no suitable existing scale could be identified. Participants were asked to “rate the extent to which you agree with the following statements” on a scale from one ( strongly disagree ) to seven ( strongly agree ), with a “don’t know” option. For stage one, noticing the emergency, the item read “At the party I would have noticed what Jack was doing to Emily”. For stage two, recognising the event as an emergency, the item read “Emily needed help in this situation”. For stage three, accepting responsibility to intervene, the item was “As someone at the party, I would feel personally responsible for helping Emily”. And lastly, for stage four, belief in your ability to intervene, the item was “I believe I would have the ability to successfully intervene and help Emily”. Intervention Intentions.It was not possible in this study to measure actual bystander intervention behaviour (stage five of Latané and Darley’s (1970) model) in response to a sexual assault, and indeed is very difficult to measure in an experimental setting for ethical reasons. This study measured intervention intentions insteadin several ways. Firstly, on an identical scale to stages one to four, a single item was used to measure overall intervention intentions in general: “In this situation, I would intervene in some way to help Emily”. Secondly, participants were asked to “rate the likelihood of you carrying out the following actions in response to Jack’s actions at the party” on a scale from one ( not at all likely ) to five ( extremely likely ; a “don’t know” option was included, but excluded from analysis). The 12 items included “Covertly trying to keep Jack away from Emily” and “Confronting Jack after the incident”. The question phrasing and scale was taken from Banyard and Moynihan (2011), and the items were adapted from Bowes-Sperry and O'Leary-Kelly’s (2005) theoretical work. Item scores were averaged to create an overall intervention intention scale whereby higher scores indicated increased likelihood of intervening. Empathy. This seven-item scale was adapted from Smith and Frieze’s (2003) Rape-Victim Empathy Scale’s subscale specific to empathy during a rape. This scale was adapted by making it specific to the victim and more broadly to sexual assault instead of rape, and by removing one item which was irrelevant to this context: “I get really involved with the feelings of a rape victim in a movie”. Participants were asked to “Please rate the extent to which you agree with the following statements”. For example, the statements included “I can imagine how Emily feels during the incident” and “It’s not hard to understand the feelings of Emily”. The scales ranged from one ( strongly disagree ) to five ( strongly agree ), with a “don’t know” option. After reverse coding one negative item, these scores were summed and divided by seven to create an empathy score. Higher scores indicated higher levels of empathy, and more specifically perspective-taking, with the victim. Empathic Concern. This eight-item scale was used to measure the participant’s concern for the victim and was adapted from Davis’s (1980) Interpersonal Reactivity Index (IRI)’s Empathetic Concern Subscale. Participants were asked to indicate how well each item described them on a scale from one ( does not describe me well) to five ( describes me very well ) with a “don’t know” option. Items included “I have tender, concerned feelings for Emily” and “When I think of Emily being taken advantage of, I feel kind of protective towards her”. Three negative items were included, such as “Emily’s misfortunes do not disturb me a great deal”, which were reverse coded before analysis. An attention check item was also included here “I am paying full attention to this survey”, which was not included in analysis for this measure. Item scores were averaged, creating an overall empathic concern score, whereby higher scores indicated a higher level of concern for the victim. Victim Derogation. This relative scale was adapted from Dawtry et al.’s (2018) Relative Derogation items and asked participants to rate the victim (Emily) on certain traits (carelessness, irresponsibility, & foolishness) in comparison to the average university student. Participants were asked to rate the victim on a scale from zero ( not at all careless / irresponsible / foolish compared to the average university student ) to 100 ( a great deal careless / irresponsible / foolish compared to the average university student ). A relative scale was used to measure derogation because relative (vs. absolute) scales permit more negative, counter-normative responses to victimisation like derogation (see Dawtry et al., 2018). Item scores were averaged to form an overall victim derogation score, with higher scores indicating more derogation of Emily’s character. Victim Blame. This six-point scale measured the level of blame put on Emily for what happened at the party in the scenario and was adapted from Gerber et al.’s (2004) Questions Involving Victim Blame. Item phrasing varies but was always measured on a seven-point scale from one ( not at all ) to seven ( completely ). Items included “To what extent was Emily’s behaviour responsible for what happened at the party?” and “How much do you consider the incident at the party to be Emily’s fault?”. Item scores were averaged, with higher overall scores showing more blame placed on the victim. Psychological Distancing from the Victim. Hafer’s (2000) Disassociation Scale was adapted to measure psychological distancing from the victim (Emily). The three items were “To what extent do you think Emily is like you?”, “In general, how similar would you say you are to Emily?” and “Overall, how much do you identify with Emily?”. Scale wording varied but all items were measured on a seven-point scale, with higher scores indicating that Emily was more similar to the participant. In order for the scale to reflect disassociation, the scores were reverse coded and then averaged to create an overall psychological distancing score, whereby higher scores indicated greater psychological distancing from the victim. Past Bystander Intervention. As behavioural intentions do not always equate to actual behaviour (Ajzen, 1985), our questionnaire included a measure of past bystander behaviour in real life to compare intentions to. This 12-item scale asked participants to indicate if they had ever carried out any of the bystander intervention behaviours listed. All items were specific to sexual assault and taken from Bowes-Sperry and O'Leary-Kelly (2005), mirroring the behaviour-specific intervention intention items mentioned previously. Items included “Covertly trying to keep a sexual assault perpetrator / potential perpetrator away from the sexual assault victim / potential victim” and “Confronting a sexual assault perpetrator after the assault”. When answering if they had previously carried out these behaviours, participants could select from the following options: “Yes”, “No”, “Don’t know / Can’t remember” or “Prefer not to say”. “Yes” answers selected in response to each item were summed to create an overall past bystander intervention score, with higher scores indicating that the participant has intervened in more ways in the past. Measures of several extraneous variables were included to control for: past sexual assault victimisation, past sexual harassment victimisation and social desirability bias. Research suggests prior assault victimisation and social desirability bias as factors that could increase intervention intentions in questionnaires (Banyard et al., 2007; Labhardt et al., 2017; Mauer et al., 2022; Moschella-Smith et al., 2022; Vonderhaar, 2015). No literature could be identified suggesting the role of past sexual harassment victimisation, but given the potential role of past assault victimisation, past harassment was included to err on the side of caution. Past Sexual Assault. Participants were asked if they had ever been sexually assaulted (following a detailed definition shown in Supplementary Materials; coded in data as 1 = assault experienced, 0 = no assault experienced; 3=prefer not to say). Past Sexual Harassment. An extensive definition of sexual harassment (see Supplementary Materials), was provided for participants, who were asked to indicate if they had ever been sexually harassed (coded in data as 1 = harassment experienced, 0 = no harassment experienced; 3= prefer not to say). Social Desirability Bias. This 16-point scale was directly taken from Stöber (2001). Participants were asked to “read each statement carefully and decide if that statement describes you or not. If it describes you, check the word ‘true’; if not, check the word ‘false’”. Statements included “In traffic I am always polite and considerate of others” and “During arguments I always stay objective and matter-of-fact”. Six reverse coded questions were also included, such as “I take out my bad moods on others now and then”. After these six items were transformed into positive scores, all “True” item answers were summed to create an overall social desirability bias score. Higher scores implied a higher tendency to answer questions in ways deemed more socially desirable. Procedure The study took place during March 2022. When taking part in the study, participants were taken from the Prolific website via a hyperlink to the online questionnaire, hosted by Qualtrics (https://www.qualtrics.com/). The questionnaire could only be completed on a desktop computer or laptop, to ensure the desired format was presented. In the online questionnaire, participants were presented with an information sheet about the study, and on the following pages were asked to give consent to take part and to confirm that they were aged between 18-24 years old and were a full-time student at a UK university. Following this screening process, basic age and gender demographics were collected. It was at this point that participants were asked if they had ever been sexually assaulted. If they answered yes, they were shown more information about the focus of the study, specifically that it related to sexual assault (previously only “distressing sexual experiences” had been mentioned) and they were asked to reconfirm if they would like to take part. Following this, participants were asked to read the text vignette. The less severe vignette was as follows: “Imagine you are at a party at a new friend’s student house. There’s music playing and lots of people. A while into the party, you notice a man, Jack, who keeps staring at a woman, Emily, who is standing alone across the room. Emily looks visibly uncomfortable when she notices Jack staring and quickly looks down at her phone. After a few minutes, Jack walks over to Emily, standing very closely next to her. You can overhear some of their conversation: “I like your dress” Jack says, “It’d look better on my bedroom floor”. He laughs. “No thanks” Emily replies uneasily and looks down. “What? Can’t take a joke?” Jack says, moving even closer. She turns to walk away. Jack quickly wraps his arm around her waist, pulling her closer despite her efforts to lean away from him. "I'm not interested" Emily says. Ignoring her, he continues “Hey all I mean is you’ve got a really hot body”.” The more severe vignette was identical except for a final additional line “He pauses and then moves a hand up her dress and between her legs, still holding her waist so tightly that she cannot move.” On the next page, participants were asked to describe what happened in the scenario, to check comprehension. After this, participants completed the measures described above, with an emphasis included on answering honestly, as if they had been present at the party in the scenario. On completion of these measures, participants were shown a debrief form, providing information regarding the nature of the study, the researchers’ contact information, and advice to contact their university’s wellbeing team or relevant charities (listed in the form) if they required additional support. At this point participants were asked to give final consent for their data from this questionnaire to be kept and processed by the [Redacted Name of Institution]. Study 1 Results and Discussion Data Preparation Quality checks were completed to check if the same response on a scale was selected regardless of reverse-coded questions, and checking for gibberish answers in the open end questions. No participants were removed as a result of these checks. Comprehension checks were also conducted on the open-end question of what happened at the party, but no participants were removed as a result. No participants were removed for completing the study in under five minutes. All “don’t know” and “prefer not to say” response options were coded as missing data to allow the bystander stages, intervention intentions, empathy and empathic concern data to be treated as continuous, and so that past bystander intervention, sexual assault victimisation and sexual harassment victimisation items could be treated as “Yes or No” dichotomous data, all of which would then be suitable for multiple regression analysis. The Effect of the Severity Manipulation Manipulation Checks. To check if the severity manipulation had successfully threatened the BJW, Welch’s t-test was used to test for a significant effect of the condition on the perceived unfairness of what happened to the victim (see Table 2). The condition had a significant medium-sized effect on perceived unfairness. Participants in the more severe condition rated what happened to the victim as significantly more unfair ( M = 6.91, SD = 0.28) than those in the less severe condition ( M = 6.65, SD = 0.63). The Effect of the Manipulation on BJW Strategies, Empathic Variables and the Bystander Stages. In line with our hypotheses, Welch’s t-test was used to test the effect of the severity of the scenario on the use of BJW strategies, each of the five bystander stages, and the empathic variables. As shown in Table 2, Welch’s t-tests found that the severity manipulation had no significant effect on any of the three non-rational BJW-threat-reduction strategies: victim derogation, victim blaming and psychological distancing. The severity manipulation also had no significant effect on empathy or empathic concern, thus not supporting our hypotheses. However, the condition did have a significant effect on stages two and five of Latané and Darley’s (1970) model. Although the average level of agreement was high to begin with, participants in the more severe condition had significantly higher levels of agreement with the statement “Emily needed help in this situation” (stage two; M = 6.81, SD = 0.47) compared to those in the less severe condition ( M = 6.34, SD = 0.94). This suggests that the severity of the situation affected whether it was recognised by bystanders as an emergency in which intervention was needed. Again, although agreement with the stage five general statement, “In this situation, I would intervene in some way to help Emily” was high overall, participants in the more severe condition gave significantly higher agreement scores ( M = 6.39, SD = 0.90) than those in the less severe condition ( M = 6.13, SD = 1.14). The condition also had a significant effect on intervention intentions relating to specific behaviours, whereby participants in the more severe condition stated that they were significantly more likely to intervene in a number of different ways ( M = 3.85, SD = 0.53) than those in the less severe condition ( M = 3.64, SD = 0.59). The effect of the condition on all other stages of the model were nonsignificant, therefore providing mixed support for the predicted effect on the bystander decision-making process. Moreover, we had hypothesised that agreement that intervention was needed (stage two) and intervention intentions would be lower in the more severe condition due to the negative mediating influence of BJW non-rational strategies (e.g. Li et al., 2018) but in fact found the effect to be in the opposite direction. Exploring the Predictive Model To explore the proposed relationships outlined in the hypotheses, hierarchical multiple regression was conducted to test pathways between variables, allowing us to examine the predictive abilities of variables without taking account of the influence of potential mediators, thus measuring the total effects. Mediation analysis was then conducted using Models 4 and 6 of PROCESS (Hayes, 2022; 1,000 bootstrapped samples; percentile bootstrapping used) to test the significance of indirect relationships. The analysis in this section adopted the more cautious approach of Rohrer (2019), only performing mediation analysis where there was a significant total path from the predictor variable to dependent variable, to minimise the risk of spurious correlations. In the case of serial mediation, Rohrer’s (2019) approach was inferred to mean that there should be significant total effects of earlier mediators on later mediators and dependent variables, which is the approach taken here. Thus, mediation was only deemed appropriate to explore where all pathways were significant. To reduce this risk further, the potential extraneous variables of past experiences of sexual assault and sexual harassment, and social desirability bias, were entered into regression and mediation analysis as covariates to control for their influence, along with all other variables of interest (excluding potential mediators, e.g. excluding stages two to four from the covariates when examining how stage one predicts stage five). The results of our hierarchical regression analysis are shown in Table 3 and our mediation pathways are shown in Table 4. The Bystander Model. Firstly, we examined our prediction that each bystander stage would predict the stages that followed it in the sequence outlined in Latané and Darley’s (1970) model, and could even mediate the relationship between earlier and later stages. In line with this, noticing the event (stage one), positively, significantly predicted recognising the event as an emergency where intervention was needed (stage two). This meant that those who gave higher ratings of agreement that they would have noticed the sexual assault also had higher levels of agreement that intervention was needed to help the victim. Stage three, accepting responsibility to intervene, was in turn significantly predicted by the previous two stages. Higher levels of agreement that the participant would have noticed the event and that intervention was needed both predicted higher levels of agreement that they were personally responsible to intervene. Stage four, belief in one’s ability to intervene, was predicted by stages one and three, whereby those who gave higher scores to agreeing that they would notice the assault and that they were personally responsible to intervene both predicted stronger belief in their ability to intervene successfully. However, stage two, recognising the event as an emergency, did not significantly predict stage four. As discussed in the methods section, stage five, intervention intentions, was measured in two ways: firstly, as a measure of general intentions, and secondly as intentions to perform specific bystander intervention behaviours. When looking at the general measure, this was positively predicted by all previous stages. This meant that those with higher levels of agreement that they would have noticed the event, that intervention was needed to help the victim, that they were personally responsible for helping, and higher levels of belief in their ability to intervene, had higher levels of intervention intentions. However, for the behaviour-specific stage five measure, this was only significantly predicted by stages three and four, whereby those with higher levels of responsibility acceptance and belief in their ability to successfully intervene were more likely to intend to intervene in a number of ways. However, stages one and two did not significantly predict the stage five behaviour-specific measure. In summary, looking at these total effects in isolation provides mixed support for our hypotheses, largely supporting the findings of Nickerson (2014) and Burn (2009). Moreover, because of the significant direct pathways identified, this met the criteria for mediation analysis to explore whether the relationship between earlier stages in the model and intervention intentions is mediated by later stages in the model. As shown in Table 4, stage one only had a significant indirect pathway on the general measure of stage five via either stage three, stage four, or stages three and four in sequence. Moreover, the remaining direct pathway indicated that stage one’s ability to predict bystander interventions in general was fully mediated by stages three and four. There were also several cases of partial mediation, whereby the mediation pathway was significant, but could not fully explain the relationship between the earlier bystander stage and intervention intentions (see Table 4). Therefore, through examining both direct and indirect pathways, we can see that the data provides mixed support for our predictions. All five stages predicted general bystander intentions. However, this was not the case when looking at the behaviour-specific measure of intentions. Moreover, there were mixed results for each stage predicting and mediating each other in the sequence outlined in Latané and Darley’s (1970) bystander intervention model. The Role of the Severity Manipulation in our Predictive Model. Our hypotheses included the effect of the severity manipulation on intervention intentions being mediated by other variables in our model, specifically, BJW non-rational strategies, empathic variables and the bystander stages. However, due to the lack of significant effects of the severity manipulation, it was not possible to test for indirect effects in all but one case: the indirect effect of the severity manipulation on general intervention intentions via stage two. However, this pathway was not significant (see Table 4). This suggests that our earlier point that the lessened ambiguity in the more severe condition, and therefore greater recognition of the event as an emergency (stage two), could explain greater intervention intentions is indeed not the case. Thus, the evidence here would suggest that our hypothesis was not supported, that none of these variables mediate or explain the effect of the severity manipulation and intervention intentions. The Predictive Role of BJW Non-Rational Strategies. In order to assess the predictive abilities of the BJW non-rational strategies, we explored the relationship these strategies had with the bystander stages, which in turn we know significantly predict general intervention intentions. As shown in Table 3, victim derogation did not significantly predict any of the bystander stages. Victim blaming however can be seen to predict stage three negatively and significantly, whereby those who blamed Emily (the victim) more for her actions in the scenario were less likely to accept responsibility to intervene. Victim blaming did not significantly predict any other stages. Psychological distancing, however, negatively predicted stages one, three, four and the general measure of stage five. This means that those who psychologically distanced themselves from Emily (the victim) were less likely to think they would notice the event, accept responsibility to intervene, think they would be able to intervene successfully and intend to intervene in some way. However, it is worth noting that this significant relationship between distancing and intervention intentions was not found when talking about the specific intervention strategies measure. Thus, there is very mixed evidence in relation to our hypothesis that BJW non-rational strategies would predict stages of the bystander theory. More specifically, the BJW strategies did not predict ability to recognise the event as an emergency as hypothesised, but victim blaming and psychological distancing did negatively predict sense of responsibility to intervene, as we had predicted. Due to psychological distancing significantly predicting multiple stages, there were grounds to explore indirect pathways. As shown in Table 4, psychological distancing had several significant indirect pathways to general intervention intentions, almost all of which involved mediation via stage three, which collectively fully mediated this relationship. This provides some support for our hypotheses, although the involvement of stages one and four were not predicted in this hypothesis, and that these patterns did not occur for the behaviour-specific measure of intervention intentions. In summary, results on victim derogation do not support our hypotheses related to the role of BJW non-rational strategies, and victim blaming providing limited support, while the patterns surrounding psychological distancing provide some supporting evidence. Several key mediators of psychological distancing predicting intervention intentions became evident: noticing the event (stage one), accepting responsibility to act (stage three) and belief in one’s ability to act (stage four; in that order). Although going against our original predictions that stages two and three would be most crucial, our findings still highlight the importance of stage three, and substantiate past research already linking these variables together (Burn, 2009; Katz et al., 2015; Nickerson, 2014; Wang, 2020). The Mediating Role of Empathy and Empathic Concern. This hypothesis predicted that the relationship between BJW non-rational strategies and the bystander stages could be mediated by empathy or empathic concern. As shown in Table 3, psychological distancing negatively predicted empathy, whereby those distancing themselves from the victim to a greater extent showed less empathy towards the victim. However, in turn, empathy did not significantly predict any of the bystander stages, suggesting that it is not a mediator here. When empathic concern was examined, the analysis showed that it was negatively predicted by both victim blaming and psychological distancing. This means that those who blamed the victim more, and psychologically distanced themselves more so from the victim, displayed lower levels of empathic concern for the victim. And unlike empathy, empathic concern in turn predicted all the bystander stages (including both the general and specific behaviour measures of intervention intentions). This means that those showing greater empathic concern for the victim were more likely to agree that they would notice the event, that intervention was needed, that they were personally responsible for intervening, and that they would be able to successfully intervene, as well as being more likely to intervene in general and using specific behavioural strategies. This would suggest that empathic concern could mediate the relationship between BJW non-rational strategies and the bystander stages. As victim blaming did not predict intervention intentions, it was only possible to examine whether the pathway from victim blaming to stage three was mediated by empathic concern. We found this relationship to be fully mediated by empathic concern, thus providing evidence in support of our hypothesis that a BJW strategy would predict a bystander stage via empathic concern. Of the indirect pathways from psychological distancing, via empathic concern to the bystander stages prior to intervention, several, shown in Table 4, predicting stages one, three and four were found to be significant. Moreover, empathic concern could fully explain psychological distancing’s ability to predict stages one and four. In addition, empathic concern, followed by stages one, three and four of the bystander model (or combinations of these stages) significantly fully mediated the relationship between psychological distancing and general intervention intentions (see Table 4 for each combination of mediators). However, it is worth highlighting that stage three was the only bystander stage that was capable of significantly mediating this process without any other bystander stages in the mediation chain, suggesting perhaps that it is the most crucial bystander stage. In summary, although not providing full support for the hypothesised mediating role of empathy and empathic concern, the data supports the importance of empathic concern in line with the findings of Katz et al. (2015), Nickerson (2014), and Patil et al. (2018). They suggest that it is the compassion element specific to empathic concern that is key to the bystander process, and not the perspective-taking element of empathy. This supports Jordan et al.’s (2016) argument that empathy and empathic concern are two distinct concepts. The Relationship between Bystander Intentions and Behaviour In line with our hypothesis, past bystander intervention was positively correlated with both the general stage five measure of intervention intentions, and the behaviour-specific stage five measure (see Table 5). This means that those who answered that they had stronger intentions to intervene to help the victim (Emily) in the scenario reported having previously engaged in more real-life bystander intervention methods related to sexual assault. This casts some validity onto our novel measures, meaning we can have some level of confidence regarding predictions of intervention intentions in this study, although more research would be beneficial to support this. Exploratory Analysis Our hierarchical regression analysis discussed above was analysed from the perspective of our theoretical model shown in Fig. 1, whereby variables were entered in order of their position in the sequence so that slopes and p-values would not take account of the shared variance with later potential mediators in the model. However, due to the theoretical nature of the model, as well as a lack of evidence directly testing the sequential order of the stages of Latané and Darley’s (1970) bystander model, and indeed evidence of later stages affecting earlier ones (cycling, Latané and Darley, 1970), we also conducted exploratory regression analysis of all variables of interest predicting intervention intentions, free from our theoretical approach (see Table 6). With all variables entered simultaneously, we were also able to directly compare semi-partial correlations squared in order to compare how much variance in intervention intentions each variable could account for. This analysis was not pre-registered and was included upon reflection during analysis. We found that when all variables were entered together, the only variables which significantly predicted greater general intervention intentions were, a greater belief in your ability to intervene (stage four), greater acceptance of responsibility to intervene (stage three), greater recognition that intervention was needed (stage two) and lower empathic concern (in descending order of variance in intervention intentions explained). This means that participants who had a greater belief that they could successfully intervene, a greater belief that they were personality responsible for intervening in the assault, agreed to a greater extent that intervention was needed or showed lower emotional upset for the victim’s situation, were more likely to intervene in some way. For behaviour-specific intervention intentions, we found that it was positively predicted by responsibility to intervene (stage three), the severity manipulation, and past intervention intentions in real life (again in descending order of variance in intervention intentions explained). This meant that participants who had a greater belief that they were responsible for intervening, who read the more severe vignette, or had intervened in more ways in real life were more likely to intervene in the party scene in a variety of ways. This analysis again highlights the importance of accepting responsibility to intervene, being the only variable which significantly predicted both intervention intention measures, when controlling for all other variables. This supports a finding from Burn’s (2009) regression analysis that when you control for the influence of all other stages in the bystander model, accepting responsibility to intervene was the only stage that remained significant in predicting intervention in both men and women. In summary, this study set out to explore how BJW non-rational strategies might be linked to the bystander decision-making process, in the context of sexual assault. Our analysis suggests that intervention intentions are greater when the sexual assault being witnessed is more severe, although the mechanisms by which this effect occurs are unclear. Our mediation models suggests that those who psychologically distance themselves from a victim show lower levels of empathic concern to the victim, believe they would be likely to notice the events, have a reduced sense of responsibility to intervene, have less confidence in their ability to intervene, and in turn, have lower intentions to intervene in some way. However we found a lack of significant effects of the severity manipulation on BJW non-rational strategies, despite past research suggesting significant effects (Dawtry et al., 2020). A possible explanation for this could lie in the methodology, namely the use of a text vignette. Although this method helped us to perform an initial exploration of the relationship between these variables, it lacks emotional impact compared to other stimuli formats, such as filmed VR (Dawtry et al., 2020; Patil et al., 2014). Study 2 Introduction Study 2 aimed to replicate Study 1, using VR stimuli to explore whether the use of BJW non-rational strategies which may not have occurred when using the relatively emotionally-unstimulating text vignette in Study 1 (Dawtry et al., 2020; Patil et al., 2014). We also added several new measures. Firstly, we added a measure of the participant’s emotional response to the stimuli to capture the emotional impact Dawtry et al. (2020) discussed. This measure included three dimensions: emotional arousal, which is the intensity of the emotional response (from calm to excited), emotional valence, meaning the pleasantness of the type of emotions experienced (from unpleasant, e.g. angry, to pleasant, e.g. happy), and emotional dominance, which encapsulates the level of control you feel you have over your surroundings (from submissive to in control, Bradley & Lang, 1944; Russell & Mehrabian, 1977; Sutton et al. 2019). Secondly, we added measures of physiological arousal, specifically HR, heart rate variability (HRV) and electrodermal activity (EDA) as indirect indicators of emotional arousal (Fernández et al., 2012; Greco et al., 2016; Kreibig et al., 2007; Navarrete et al., 2012; Schaaff & Adam, 2013 ), allowing us to compare the emotional arousal in several stages of the study. We also added a measure of presence in the VR environment to explore how this may be related to the BJW and bystander variables, as previously unconnected, but theoretically relevant, constructs. And lastly, we added a measure of subtle victim blame, to use in addition to the existing victim blame measure, but this one designed to capture more indirect methods of blame that may have been missed in Study 1 (Hafer et al., 2019). Hypotheses As well as the hypotheses from Study 1, the addition of the new measures mentioned above resulted in several additional hypotheses. Firstly, we predicted that the emotional response (arousal, negative valence and dominance) and the increase in physiological arousal compared to baseline (HR, HRV and EDA) would be higher in the more severe condition. This hypothesis was based on Dawtry et al.’s (2020) suggestion that the emotional impact of stimuli is greater with more severe stimuli. We also predicted a within-participants effect of the measurement period on physiological arousal (HR, HRV and EDA) would be highest when participants are watching the staged sexual assault video in VR, followed by when they were watching a sight test in VR, and lowest during the baseline, following the pattern of the expected emotional impact. Also based on Dawtry et al.’s (2020) research, we predicted that the three dimensions of emotional response to the stimuli, especially emotional arousal would be positively related to the use of BJW non-rational strategies. We examined correlations between the emotional response and presence with all other variables, in an exploratory manner (see Supplementary Materials). Subtle victim blame predictions followed the same hypotheses as the victim blame measure used in both studies (see Hypothesis section for Study 1). Study 2 Method Participants For this study, we aimed for a sample of 120 full-time UK university students aged between 18 and 24, as a feasible compromise between trying to maximise statistical power and being realistic with recruitment for in-person VR studies. Sensitivity power analyses conducted in G*Power indicated that our sample size provided 80% power to detect at least medium effect sizes ( d = .52, α = .05) in our t-test analysis. Due to the potentially higher emotional impact of Study 2, we excluded people who had experienced sexual assault, and people who had experienced six or more symptoms of PTSD in the past week related to past sexual harassment. In line with the [Redacted Name of Institution]’s Standard Operating Procedure for VR headsets, we also excluded individuals with health conditions which would prevent them from using VR, such as those with epilepsy or a heart condition. We also advised potential participants that they should not take part if tired, in need of sleep, under the influence of drugs or alcohol, hungover or stressed. Participants were recruited in two ways. Eighty-one were recruited through the research participation scheme for undergraduate psychology students at the [Redacted Name of Institution] and received course credits as an incentive. Of these 81, 13 had also participated in another VR sexual assault study although the focus of these studies differed enough that we do not believe this would have affected responses. Forty-one were recruited through either the research community panel or flyers (displayed around campus, on university websites and on social media), and were given £10 for completing both the eligibility survey and the 45-minute in-person session,. From our recruited sample of 122, the data from five participants were removed from the dataset for various reasons: one participant completed the 45-minute study in less than 15 minutes (a cut off we deemed ahead of data collection to be too short to give meaningful answers), two participants answered “not sure” to the closed-end comprehension question, one participant failed to accurately describe the VR scene in the open-end comprehension check, and one participant experienced serious technical issues during the VR scene (which prevented them from seeing the stimulus as intended). This left us with a final sample size of 117 (16 men, 96 women, 5 gender unknown), aged between 18 and 24 ( M = 19.38, SD = 1.40; age unknown for 6 but within inclusion age range). 28.21% of the sample reported having previously been sexual harassed. Design A between-participants design very similar to Study 1 was used, in which we manipulated the severity of the sexual assault in the filmed VR scene (see Fig. 2, full scenes can be viewed here: [INSERT LINK: https://youtube.com/playlist?list=PLCfqtCP4BMd2z1ljF7wAQhTXZKKtV4lss&feature=shared] ), as an indirect way of manipulating the threat to the BJW. In the less severe condition ( n = 50), as well as sexually harassing comments and stares, the man in the video also grabbed the woman’s waist. In the more severe condition ( n = 67), he also stroked her upper thigh. In addition, within-participant analysis was used to explore the effect of the task (baseline, sight test in VR, staged sexual assault in VR) on physiological arousal. Measures All measures from Study 1 were used in this study, as well as several additional measures outlined here and included in Table 1. Trauma Related to Past Sexual Harassment. This measure was adapted from the Trauma Screening Questionnaire (Brewin et al., 2002). Participants were asked “Please consider the following reactions which sometimes occur after a traumatic event. This question is concerned with your personal reactions to previous experiences of sexual harassment. Please indicate (Yes/No) whether or not you have experienced any of the following at least twice in the past week, specifically in relation to past experiences of sexual harassment”. Items included “Upsetting dreams about the event” and “Being jumpy or being startled at something”. Participants could respond to the 10 items with “yes”, “no”, or “prefer not to say”. Participants were deemed ineligible to take part in the study if “Yes” or “Prefer not to say” were selected for six or more items. Subtle Victim Blame. Hafer et al.’s (2019) subtle victim blame measure was adapted for this study. This four-item measure is designed to capture a less explicit form of victim blaming, one which is less susceptible to the influence of social norms. More specifically, participants were asked “How likely do you think it is that the incident at the party happened because…” followed by two factors within the victim’s control, and two factors outside of the victim’s control. For this study, the two factors within the victim’s control were that the victim was standing on her own and her choice of clothing. The two factors outside of the victim’s control were the perpetrator’s intoxication, and the victim being in “the wrong place at the wrong time”. Participants were asked to respond on a seven-point scale from one (not at all likely) to seven (very likely). The scores for the two questions with factors within the victim’s control were averaged to create a subtle victim blame measure, with higher scores indicating greater blame on the victim. Presence. This was adapted from Slater and Steed’s (1994) measure of presence in VR. Participants were asked three questions: “Please rate your sense of being there in the party”, “To what extent were there times during the experience when the party became the "reality" for you, and you almost forgot about the "real world" outside?” and “When you think back about your experience, do you think of the party more as something that you saw, or more as somewhere that you visited?”. Scale wording varied but all items were measured on a seven-point scale, with higher scores indicating a greater sense of presence in the VR setting. The scores for each question were averaged to create an overall presence score. Emotional Responses to VR Scene. Bradley and Lang’s (1994) Self-Assessment Manikin (SAM) was used to measure participants’ emotional reactions to the VR scene, more specifically capturing three variables: the emotional valence, arousal and dominance. Participants were asked to “Please rate how you felt while you were watching the scene at the party”, and responded on three 9-point scales. The scales were visual, with images of the SAM, showing varying emotions (see Fig. 3). These SAM images came with descriptive cues at the extremes of the scales, for example, for emotional arousal, the lower end of the scale had cues for “Stimulated, excited, frenzied, jittery, wide awake, aroused” and the other end had cues for “Relaxed, calm, sluggish, dull, sleepy, unaroused”. A higher score on the emotional valence scale indicated greater negative emotions. A higher score on the arousal SAM scale indicated lower emotional arousal, and was therefore be reverse coded so that a greater score in our analysis indicates greater arousal. A higher score on the emotional dominance scale indicates a greater feeling of control. Physiological Arousal. HR (interbeat-interval in raw form; IBI), HRV (IBI in raw form), and EDA were measured using the Empatica E4 wristband. We recorded during a seated 5-minute baseline, when watching the sight test in VR, and when watching the staged sexual assault condition video in VR. Mean HR, HRV and EDA for each task were calculated from individual participant files using R. To compare the effect of condition and correlations, the baseline measurement was subtracted from the measurement when watching the condition stimuli video in VR, to create an increase score for each measure to account for individual baseline differences. Procedure Data collection took place between March and December 2023. Participants completed an eligibility survey (hosted by Qualtrics) before they were able to sign up to take part in the laboratory study. In this survey, participants were given an information sheet, followed by questions on age, gender, student status, past sexual assault and harassment experience, trauma related to past sexual harassment (if selecting that they had experienced it or preferred not to say) and relevant health conditions. In the in-person session, participants were given a detailed verbal walkthrough of what they would be doing in the session and were given the opportunity to ask questions. They then read an information sheet in an online questionnaire and confirmed their consent to take part. Participants then had an Empatica E4 wristband attached to them and a five-minute seated baseline physiological recording was collected, and this wristband remained on for both VR videos. Participants were given instructions for the VR scenes, namely that they would be seeing two videos (a sight test and the main video), in both they would be able to look in any direction and rotate on the spot, but would not be able to move around the room or interact with the environment. As part of this, they were also reminded of their right to withdraw if feeling cybersickness in VR or feeling upset by the video content. Participants then had the Oculus Quest 2 head-mounted display fitted and saw a sight test in VR to confirm that they could see details and to get participants used to the VR setting. This sight test depicted a 360-degree photo of a room containing a Snellen chart on the wall in front of them (see Fig. 4; an optical sight chart traditionally used by opticians). Participants were asked to read the top seven lines of random letters on the chart. If unable to do so, the VR headset was adjusted to improve the focus, and if still unable to read the lines, the participant would have been informed that they were not able to take part in the study (although this last step did not happen with any of the participants). Following the sight test, participants were then instructed that they would be watching a two-minute video next and that again they would be able to look around, but could not interact with the environment. Participants then watched the stimuli video from the condition they were assigned to (see Fig. 2), which followed the same story as in Study 1 (except for the more severe condition’s exact form of sexual assault, see design section for details). Following this, the VR headset and wristband were removed. Participants then completed the main questionnaire. As well as the comprehension question used in Study 1, an additional question was added whereby they were asked to confirm if they saw the man in the white t-shirt (perpetrator) and the woman in the red dress (victim). They then completed the measures described above. After completing the questionnaire, participants were given a verbal debrief, explaining the rationale behind the study, it’s equipment and the conditions, and were actively encouraged to ask questions and talk about how they were feeling after watching the stimuli. They were given a debrief form to take away with them, following the same structure as the form in Study 1, and the contact information of the research team and relevant organisations offering support were verbally highlighted. Participants then returned to the VR headset and watched a filmed VR video of the actors from the stimuli VR scene explaining that they are actors and are safe and well. Following this, participants were given the option of watching one of three happy VR videos to ensure that they left the study on a positive note and did not leave with any negative associations with VR due to the study’s subject matter. These videos were obtained from the DeoVR app, and each showed a 180-degree filmed VR video of a relaxing sunset over a river and trees, cherry blossom trees in bloom, and puppies. Study 2 Results and Discussion Data Preparation Quality checks identical to Study 1 were used, and again, did not result in any participants’ data being removed. We were unable to match the screener data to the rest of the data sources for six participants, meaning there is missing data on screening measures including age, gender, and sexual harassment, for those people. As with Study 1, there were several measures which included a “don’t know” options, and these response options were coded as missing data so that the remaining scale could be treated as continuous or dichotomous. There were errors with incorrect or missing tagging when using the Empatica E4 wristband for 13 participants, so their physiological data was not included in analysis. There were also issues with substantial noise in the IBI raw data for an additional 28 participants, meaning that their HR and HRV data was unusable (while their EDA data was fine). This left a sample 104 participants for the EDA analysis, and 76 participants for the HR and HRV analysis. The Effect of the Severity Manipulation Manipulation Checks. We again used Welch’s t-test to see if the condition had a significant effect on our manipulation check variable (see Table 2). However, in this study, there was no significant effect on perceived unfairness. It is worth noting that the effect size was comparable to that for Study 1 and did not statistically significantly differ (see confidence intervals in Table 2). The Effect of the Manipulation on BJW Strategies, Empathic Variables and the Bystander Stages. In our hypotheses, we predicted that in the more severe condition, if BJW non-rational strategies were used, empathy and empathic concern would be lower, and there would be a reduced ability to complete each of the bystander stages (e.g. lower intention to intervene). As shown in Table 2, Welch’s t-test found significant effects of the severity manipulation on both the general and behaviour-specific measures of intervention intentions. However, we found the opposite to what we had predicted: instead of intervention intentions being higher in the less severe condition, general intervention intentions were significantly greater in the more severe condition ( M = 6.62, SD = 0.67), than in the less severe condition ( M = 6.14, SD = 1.08). Similarly, intentions to intervene in specific ways was higher in the more severe condition ( M = 3.71, SD = 0.55), than in the less severe condition ( M = 3.50, SD = 0.51). Contrary to our hypotheses, no other effects of the severity manipulation were significant. The Effect of the Manipulation on Emotional and Physiological Responses. In Study 2, we also added the hypothesis that the emotional response to the stimuli (arousal, negative valence, and dominance) would be greater in the more severe condition. However, as shown in Table 2, there was no significant effect of the manipulation on any of the emotion response variables. Similarly, we tested whether the increase in physiological arousal, measured as the difference in HR, HRV and EDA from baseline to the stimuli watching period, was greater in the more severe condition than the less severe one. But we found no significant difference in this across conditions. Exploring the Predictive Model An analysis plan identical to Study 1 was used for Study 2, whereby to examine the hypothesised relationships, we used hierarchical multiple regression, conducted in the sequence of steps in our model (Fig. 1), and where appropriate, then conducted mediation analysis. Although we had planned to include it, subtle victim blame was not included in regression and mediation analysis due to its low Cronbach’s alpha (see Table 1) and on reflection to allow a fairer comparison of other BJW regression pathways with those from Study 1 in which this variable was not recorded. The results of our hierarchical regression analysis are shown in Table 3 and all mediation pathways we tested are in Table 4. The Bystander Model. As with Study 1, we tested our prediction that each of the bystander stages would significantly predict all the stages following it in Latané and Darley’s (1970) model, and most importantly, would predict stage five, intervention intentions. In Study 2 we found that stage one, noticing the event, did not predict stage two, deciding that intervention is needed. Noticing the event and deciding that intervention was needed did both positively predict stage three, accepting responsibility to intervene. Those who gave higher agreement scores that they would notice the sexual assault and that help was needed had a greater sense of responsibility for intervening as someone attending the party. We also found that both noticing the event and accepting responsibility to intervene positively predicted stage four, belief in your ability to intervene. Greater agreement that the participant would have noticed the assault and that they were responsible for intervening was linked to a greater belief in their ability to intervene. But, agreement that intervention was needed did not predict belief in your ability to intervene. When looking at general intervention intentions, we found that this was positively predicted by stages one, three and four of the model. Participants with higher agreement that they would notice the event, that they would be responsible to intervene, and that they believed they could successfully intervene answered that they would be more likely to intervene in some way. However, we found that the second stage, deciding that intervention is needed, did not predict general intervention intentions. For the behaviour-specific intervention intentions measure, we found that this was positively predicted by stages one, two, and three, but not stage four. Those who agreed to a greater degree that they would notice the event, that intervention was needed, and that they were responsible for intervening, agreed to a greater extent that they would intervene in a number of suggested ways. Thus, direct regression pathways provide mixed results for our hypothesis that each stage in Latané and Darley’s (1970) model would predict those following it in sequence. Nonetheless, our findings that several of the bystander stages predicted those following it, and that all stages in the model predicted one or more measures of intervention intentions, supports the works of Burn (2009) and Nickerson (2014). Examining all possible indirect pathways, we found several significant indirect pathways in which stage three, accepting responsibility to intervene, mediated it. Firstly, stage three fully mediated the pathway from stage one to the general intervention intentions, Secondly, the pathway from stage two to the behaviour-specific measure of bystander interventions via stage three was also fully mediated. And lastly, stage three mediated the pathway from stage one to behaviour specific intervention intentions, but could not fully explain this relationship. Paired with the direct pathways, these indirect pathways provide only partial support for our hypothesis, but highlight the importance of stage three, accepting responsibility to intervene. The Role of the Severity Manipulation in our Predictive Model. Due to the severity manipulation only having a significant effect on intervention intentions, we were not able to test our hypotheses that this effect would occur via BJW non-rational strategies, empathic variables or the bystander stages (specifically stages two and three). The lack of significant effects does not provide evidence for our hypotheses. The Predictive Role of BJW Non-Rational Strategies. Next examining if the BJW non-rational strategies negatively predicted the stages of Latané and Darley’s (1970) model, we found limited support for this hypothesis. We found that victim derogation positively predicted greater intervention intentions. Those who blamed the woman in the red dress (victim)’s character more for the assault were more likely to intend to intervene in some way. This is in direct contrast to past research which had found a negative relationship between BJW non-rational strategies and bystander intervention (Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022), and the lack of a significant prediction from derogation to intervention intentions in Study 1. However, this pattern in Study 2 was not found in our simultaneous regression or correlational analysis (see Tables 5 & 6), suggesting this pattern could be due to controlling for the variance shared with certain variables within our data in the hierarchical sequence. An alternative suggestion could be that critiquing the victim’s character meant that the bystander viewed her as incompetent and therefore not capable of looking after herself, perhaps as a form of benevolent sexism (Glick & Fiske, 1996), although further research would be needed to confirm if this was the mechanism. Again in contrast to our hypothesis, psychological distancing positively predicted greater agreement that they would notice the event. Those who said they were more different from the woman in the red dress believed they were more likely to notice the harassment and assault taking place. Victim blaming however, negatively predicted the participant’s belief that they would notice the event and their general intervention intentions. Those who blamed the victim’s actions more believed that they would be less likely to have noticed the assault and less likely to intervene in some way. Mediation analysis was run to explore the only possible mediation path from BJW strategies to intervention intentions via the bystander stages, namely whether victim blaming predicted general intervention intentions via reducing noticing ability, but this was not significant (see Table 4). Overall, Study 2 provides no evidence to support derogation and psychological distancing negatively predicting the stages of Latané and Darley’s (1970) model either directly or indirectly, but evidence of victim blaming doing so to a degree directly. Although this last finding provides evidence in line with past research (Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022) linking victim blaming and intervention intentions, our predictions that this relationship would be mediated by the stages of the bystander model were not found to be the case. The Mediating Role of Empathy and Empathic Concern. We also predicted that empathy and empathic concern may act as a mediator between BJW non-rational strategies and the bystander stages (specifically stages two and three), and then intervention intentions. In line with our hypothesis, we found that empathy and empathic concern were both negatively predicted by psychological distancing. This means that participants who believed they were very different from the woman in the red dress had lower levels of empathy and empathic concern for her. We also found that victim blaming negatively predicted empathic concern, meaning that those who blamed the victim more had lower levels of empathic concern for her. These patterns support our hypothesis, and past research (Katz et al., 2015), suggesting that there is a negative relationship between some BJW non-rational strategies and empathic variables, although this was not the case with victim derogation. Examining if our empathic variables could predict the bystander stages, we found limited evidence to support this hypothesis, with most predicted pathways being nonsignificant. We did however find two significant regression pathways. Firstly, empathy positively predicted general intervention intentions. This meant that those who were better at taking the perspective of the woman in the red dress rated themselves as more likely to intervene in some way. And secondly, we found that empathic concern positively predicted stage two of the bystander model, recognising the event as an emergency whereby intervention is needed. Under this regression pathway, those who felt more empathic concern for the woman in the red dress were more likely to agree that intervention was needed. Thus, there is some evidence from this study connecting our empathic variables with the bystander decision-making process, but this is more limited than we had predicted. Due to a lack of significant pathways between the severity manipulation, BJW non-rational strategies, empathy and empathic concern, and the bystander stages, it was not possible to test our empathic variables as mediators in this process. The Relationship between Bystander Intentions and Behaviour We predicted that both our general and behaviour-specific measures of intervention intentions would be correlated with past bystander intervention against sexual assault in real life. However, we found that neither of these correlations were significant (see Table 5). Physiological Arousal Across Tasks New to Study 2, we predicted a within-participants effect on physiological arousal, whereby physiological arousal (HR, HRV & EDA) would be highest when participants were watching the staged sexual assault in VR, followed by when they are watching the VR sight test, and then during the seated baseline. Three repeated-measure ANOVAs revealed that the task had a significant effect on HR, HRV and EDA (see Table 7). Planned comparisons showed that HR during the sight test and sexual assault video were both significantly higher than during the baseline. However, there was no significant difference in HR between during the sight test and sexual assault video. Similarly, HRV was significantly higher in the sight test and sexual assault video than in the baseline. But, like HR, HRV during the sight test and sexual assault video did not significantly differ. For EDA however, we found significant differences between all three tasks. More specifically, EDA during the sight test and sexual assault video was significantly higher than during the baseline, and EDA during the sexual assault video was significantly higher than using the sight test. These findings are largely in support of our hypotheses, showing that the task influenced physiological arousal, although failing to provide support for circulatory indicators having heightened levels during the sexual assault stimuli VR video compared to the VR sight tests. This could suggest either issues with measurement quality, or that HR and HRV were more greatly impacted by standing up and being in VR, rather than by the content within VR, while EDA was able to capture the impact of the emotional nature of the content too. Exploratory Analysis Simultaneous Regression. As with Study 1, we conducted simultaneous regression analysis to compare the relative contribution of our variables of interest in accounting for intervention intentions, free from our theoretical preconceptions. This analysis was not pre-registered and was added as an exploratory addition during analysis. In Study 2, we found that greater general intervention intentions were significantly predicted by a greater sense of responsibility to intervene (stage three), greater belief in their ability to intervene (stage four), lower victim blaming, and higher social desirability bias (listed in descending order of the amount of variance in intentions they could account for). This means that participants who had a greater sense of responsibility, who had a greater belief in their ability to intervene, blamed the victim’s behaviour less or displayed a greater bias towards answering questions in a socially desirable way, were more likely to answer that they would intervene in some way. Greater behaviour-specific intervention intentions however were predicted by greater responsibility to intervene (stage three) and greater belief that they would notice the assault (stage one; in descending order of variance explained). This means participants who agreed to a stronger degree that they were personally responsible for intervening, or had a greater belief that they would notice the harassment and assault take place were more likely to think they would intervene in a variety of ways. Influence of Sample Differences. We also ran unplanned analysis to explore differences in our findings between Studies 1 and 2. There were several crucial differences in the samples of our two studies, namely, the gender split (49.32% female, 47.96% male, 2.72% non-binary vs. 82.05% female, 13.68% male, 4.27% unknown), percentage of the sample who had previously been sexually assaulted (38.44% vs 0%), and percentage who had previously been sexually harassed (60.88% vs 28.21%). Additional analysis was therefore performed to understand if these sample differences could account for differences in effects and relationships. Comparing confidence intervals for our t-tests and hierarchical regression results across both studies revealed that there was no overlap between confidence intervals for, and therefore key differences in, regression pathways of victim blaming predicting the ability to notice the event (stage one), psychological distancing predicting the ability to notice the event, and belief in their ability to intervene (stage four) predicting general intervention intentions (see Table 3). Moderation analysis was conducted using model 1 of PROCESS (Hayes, 2022) on data from Study 1 to examine whether gender, past sexual assault or past sexual harassment moderated these patterns. Firstly, looking at victim blaming predicting their ability to notice the event, we found that there was a significant victim blaming x sexual assault history interaction, b = -2.11, SE = 0.78, t (251) = -2.71, p = .01, 95% CI [-3.64, -0.58]. Among those who had previously been sexually assaulted, victim blaming significantly, negatively predicted belief that they would notice the event b = -1.91, SE = 0.76, t (251) = -2.50, p = .01, 95% CI [-3.40, -0.41]. This meant that sexual assault survivors who blamed the victim to a greater extent were more likely to believe that they would not notice the assault take place. Whereas, among those who had not been previously sexually assaulted, victim blame did not significantly predict agreement that they would notice the event, b = 0.21, SE = 0.16, t (251)= 1.25, p = .21, 95% CI [-0.12, 0.53]. Gender and sexual harassment victimisation did not however significantly moderate the relationship between victim blaming and noticing the event, b = -.20, SE = 0.42, t (253) = -0.48, p = .63, 95% CI [-1.03, 0.63], and b = -.71, SE = 0.47, t (256) = -1.53, p = .13, 95% CI [-1.63, 0.20] , respectively. Thus, the inclusion of sexual assault survivors in Study 1, but not in Study 2, may account for differences between victim blaming’s ability to predict noticing the event between our two studies. Looking at potential moderators of psychological distancing predicting noticing the event, we found that gender, past sexual assault and past sexual harassment all did not interact with this pathway, b = 0.19, SE = 0.12, t (253) = 1.58, p = .11, 95% CI [-0.05, 0.42] , b = 0.01, SE = 0.11 t (251) = 0.07, p = .94, 95% CI [-0.20, 0.22], and b = -.005, SE = 0.11, t (256) = -0.04, p = .97 , 95% CI [-0.22, 0.22], respectively. Similarly, we found that gender, past sexual assault and past sexual harassment all did not moderate the relationship between belief in your ability to intervene and general intervention intentions, b = -.03, SE = 0.07, t (273) = -0.37, p = .71 , 95% CI [-0.17, 0.12], b = -.03, SE = 0.08, t (269) = -.31, p = .76, 95% CI [-0.19, 0.14], and b = .02, SE = 0.07, t (276) = 0.28, p = .78 , 95% CI [-0.13, 0.17], respectively. Thus, these sample differences cannot account for differences in these pathways between Studies 1 and 2. General Discussion We conducted two experiments, exploring the effect of indirectly manipulating the threat to the NBJW on BJW non-rational strategies, empathic variables and Latané and Darley’s (1970) bystander decision-making process when witnessing a sexual assault. We expanded on our initial text vignette exploration in Study 1 by replicating the study using emotionally impactful (Patil et al., 2014), realistic and immersive filmed VR in Study 2. Differences Between our Studies Between the two studies, we found differences in the statistical significance of the success of the manipulation, the predictive ability of the BJW non-rational strategies and the role of empathic variables as mediators between non-rational strategies and bystander stages including intervention, albeit most of these differences between the studies were not themselves statistically significant. The most obvious explanation for these differences is the use of filmed VR instead of a text description, which research suggests would have increased the emotional impact of the stimuli and presence (Baños et al., 2008; Dawtry et al., 2020; Jicol et al., 2023; Jicol et al., 2022; Kahn & Cargile, 2021; Pan & Slater, 2011; Patil et al., 2014). However, future research is needed to confirm this influence in this context, experimentally manipulating the stimuli type using a randomised sample. There are also other explanations for the differences in our findings, which will be discussed here. We found differences in which BJW non-rational strategy was the most important in predicting intervention intentions. An explanation for psychological distancing being the key BJW predictor of intervention in Study 1, while victim blaming was the key predictor in Study 2, may lie in BJW literature. Dawtry et al. (2020) found that victim derogation effects were greater in response to more emotionally impactful, realistic stimuli, where the threat to the need to BJW is likely greater. This may suggest that more extreme counter-normative non-rational strategies like derogation and victim blaming are reserved for these greater belief threats. Comparing our two studies, it may have been the case that where participants were responding to a less emotionally evocative text stimuli (Dawtry et al., 2020; Patil et al., 2014), the simpler strategies of psychological distancing from the victim may have been sufficient in resolving the threat. Whereas, in Study 2, where participants were witnessing the assault in an arguably more highly emotionally arousing immersive VR context, more drastic measures, like victim blaming, may have played a crucial role in determining bystander decision-making. It could also be argued that there was a greater opportunity for victim blaming in the second study. In the first study, there is no mention of Emily’s (the victim’s) appearance. Whereas, in the second study, there are several factors of her appearance which could be used under rape myths as reasons to blame the victim. A commonly held rape myth is the idea that a victim is ‘asking for it’ because of the way they were dressed (Burt, 1980). Under this logic, the victim in Study 2 is wearing a low cut, short dress, and due to the lack of names used in the scene, is only referred to in the questionnaire as ‘the woman in the red dress’, a colour associated with sexual receptivity (Pazda et al., 2012). In this way, it could be suggested that participants adhering to rape myths may have used these components as fuel for victim blaming, which therefore may have become a greater factor in either influencing or justifying their bystander intervention intentions. However, relevant to this argument, is the subtle victim blame measure in Study 2. This measure was not included in our regression analysis, partially due to it’s poor Cronbach’s alpha (see Table 1). The reason for this low alpha was that, on average, participants were willing to engage in subtle victim blaming based on the woman standing on her own ( M = 5.00, SD = 1.44), but reluctant to blame her based on what she was wearing ( M = 2.87, SD = 1.48). This could possibly reflect a rejection of more widely known rape myths among younger adults (Chauhan, 2022), such as those based on clothing, or stronger social norms against this (Dawtry et al., 2018). Regardless, the low agreement with this second item would suggest that the clothing of the victim in our stimuli did not account for the more prominent role of victim blaming in predicting intervention intentions in Study 2. Moreover, as shown in Table 2, the mean victim blaming score did not drastically increase in Study 2, but instead all BJW non-rational strategy use slightly increased, regardless of their predictive strength. This suggests that the differences in psychological distancing and victim blaming predictions between the two studies cannot be explained by an increased use of victim blaming and a resultant reduction in the need for distancing. Moving our focus onto our potential mediators, empathy and empathic concern, Study 2 found a quite stark difference in significant patterns compared to Study 1. In Study 1, we found that greater empathy was only related to reduced psychological distancing, while greater empathic concern was a more crucial variable in our model, being predicted by both lower victim blaming and lower psychological distancing, as well as predicting greater agreement with all stages in the bystander model, including both measures of intervention intentions. In Study 2, the empathic variables were both still predicted by the same BJW strategies as in Study 1. However, their predictive patterns were very different, with greater empathy predicting greater general intervention intentions, and empathic concern only predicting greater agreement that intervention was needed (stage two). Although the reason for the changes in empathic concern’s ability to predict the bystander stages remain unclear, a possible explanation for the increase in empathy’s importance may involve the use of VR in Study 2. The very nature of VR involves the use of perspective taking, putting yourself in another person’s shoes (either literally or figuratively). Due to this, VR has been referred to as the “ultimate empathy machine” (TED, 2015, 3:12), and has indeed been harnessed for this specific purpose of increasing empathy towards particular groups or individuals (De La Peña et al., 2010; Seinfeld et al., 2018; Tassinari et al., 2022; Ventura et al., 2021). And although our average level of empathy did not increase from Study 1 to Study 2 (likely due to already nearing the scale’s ceiling in Study 1), we can see an increase in the importance of empathy in our VR study, through it’s now significant prediction of general intervention intentions. Returning to our hypotheses, we can see that both our empathic variables are related to some BJW strategies and parts of the bystander model, but did not act as mediators in Study 2 in this instance. Consistent Patterns Across our Findings Having considered the different findings between the studies, it is important to also consider the key consistent findings. Most crucially to our hypotheses, we found that manipulating the severity of the sexual assault participants witnessed (as an indirect manipulation of the threat to the BJW) had no significant effect on BJW non-rational strategies, empathy, empathic concern, and stages one, three and four of Latané and Darley’s (1970) bystander model. After Study 2 we can rule out a lack of emotional impactful stimuli as a reason for this lack of effects (Dawtry et al., 2020; Patil et al., 2014). However, another possible explanation of the lack of an effect of the manipulation on BJW non-rational strategies relates to Lerner’s (1980) theory itself. As discussed, a bystander is unlikely to proceed with non-rational strategies to remove a threat to the BJW if it is possible to resolve the threat rationally and they believe they can fully remove the suffering. Thus, the lack of significant effects of condition on non-rational strategies could be due to participants believing they could intervene in this scenario and thus resolve the threat to their BJW rationally. This explanation is further supported by the high means for measures of belief in their ability to successfully intervene (stage four) and general intervention intentions (stage five) in both studies. The use of rational strategies like intervention was ruled out in our predictions because research suggests that an individual would be unlikely to use these strategies if they believed that they could not fully remove the suffering (Montada & Lerner, 1998; Walster et al., 1973; White, 2012). As the victim had already been assaulted, it may not be possible to remove this existing suffering to their physical and mental health (Carey et al., 2018; Luce et al., 2010; Mason & Lodrick, 2013; McFarlane et al., 2005). However, it could be that participants thought that the situation might escalate into a more severe assault, which would involve far greater suffering, which they could prevent by intervening. Future research could test this explanation by manipulating the personal risks of intervening, a factor Lerner (1980) theorised affects this decision to use a rational or non-rational strategy. It would be insightful to see whether a high-cost condition would result in greater derogation, blame and psychological distancing. While it’s possible for this BJW-based explanation to account for the lack of significant effects, another explanation lies in bystander research. Based on our BJW theoretical approach, we proposed that witnessing a more severe sexual assault would elicit reduced intervention intentions due to the use of non-rational BJW strategies. However, we found the exact opposite, that intervention intentions were higher when witnessing a more severe assault. Removing our BJW lens, this finding appears to be consistent with some bystander research. For example, Huang et al. (2023) found greater intervention intentions in response to a more severe incident of cyberbullying than a less severe one. Similarly, Zhao et al. (2023) found that the perceived severity of cyberbullying predicted greater intentions to defend and support the victim. Thus, our findings appear to match existing bystander research. However, the explanation for this effect is still unclear. One would expect that increased severity has an effect on intervention intentions by removing any ambiguity and making it clearer that it is an emergency in which help is needed (stage two). But, the lack of a significant mediation pathway via stage two in Study 1, and lack of any significant effect on stage two in Study 2, casts doubt on this explanation. Huang et al. (2023) found that the effect of increased cyberbullying severity on intervention intentions was mediated by the bystander’s sense of responsibility. However, due to our severity manipulation having no effect on the responsibility to intervene, we can rule out this explanation. Thus, further research is needed to explore the mechanisms behind the effect of severity on intervention intentions in a sexual assault context. Another key finding across our two studies is the relative importance of the bystander stages. We had predicted that recognising the event as an emergency where help is needed (stage two) and accepting responsibility to intervene (stage 3) would be the key mediators, but instead found that stage two was less crucial than predicted, and that noticing the assault (stage one), and belief in your ability to intervene 4) were also important. On reflection it may not be surprising that stages one and four are involved. Firstly, the means for our bystander stage measures (see Table 1) show a greater hesitance in being able to notice the assault (stage one), perhaps being due to the party setting of our stimuli, where alcohol is likely consumed by bystanders (Haikalis et al., 2018), which research has suggested could hinder this stage (Burn, 2009; Steele & Josephs, 1990). Secondly, relating to belief in your ability to intervene, BJW highlights that it’s the belief in your ability to remove the injustice that determines whether you will use a rational or non-rational strategy (Lerner, 1980; Montada & Lerner, 1998; Walters et al., 2007; White et al., 2012), and much bystander research has demonstrated the importance of self-efficacy in determining bystander intervention (Banyard & Moynihan, 2011; Banyard et al., 2007; Bennett et al., 2014; Burn, 2009; Jouriles et al., 2019; Thornberg, 2013). Thus, the unpredicted role of stages one and four as mediators in Study 1 and predictors in Study 2 appear to fit with the literature. But regardless of these other stages, both studies consistently highlighted the importance of a bystander accepting responsibility to intervene, the third stage in Latané and Darley’s (1970) bystander model. In both studies, we found that a greater sense of responsibility for intervening predicted both greater general and behaviour-specific intervention intentions. When all variables of interest were entered simultaneously in our exploratory analysis, we found that acceptance of responsibility was the only variable which consistently predict intervention intentions, regardless of the stimuli type and measure of intervention intentions. Moreover, our mediation analysis showed that it consistently acted as a mediator between noticing the event (stage one) predicting general intervention intentions, and recognising the event as an emergency (stage two) predicting intervention intentions (although which intention measure this predicted differed between studies), further corroborating its importance. Although the role of accepting responsibility to intervene is not a novel finding, it does support the body of empirical research in this area which has found this (Bennett et al., 2014; Burn, 2009; Darley & Latané, 1968; Fischer et al., 2011; Hortensius & de Gelder, 2014; Jouriles et al., 2019; Katz et al., 2015; Levine, 2005). Where our research does add novelty is in highlighting the relative importance responsibility above and beyond other important factors such as severity, BJW strategies, and empathic variables. A further interesting pattern from our studies was that we found more direct and indirect pathways predicting general intervention intentions than behaviour-specific intervention intentions. There could be several reasons for this. Firstly, it could be that individuals were more willing to say that they would intervene “in some way” than to agree to specific actions, in the same way that it’s easier to say you intend to do something but drop out when it gets down to the practicalities of implementing that intention (Ajzen, 1985). Limitations Unlike in Study 1, the severity manipulation within Study 2 did not have a significant effect on our manipulation check, namely, perceptions of unfairness, nor on deciding that intervention is needed (stage two of Latane and Darley’s (1970) model). A possible explanation for this could be related to the distinction between the two conditions. In Study 1, the difference in severity between conditions was quite prominent, being defined by whether the perpetrator put his hand up the victim’s dress. Whereas in Study 2, the difference between conditions was a more subtle stroke of the victim’s thigh. The reason for this was due to the ethical considerations when working with actors. During filming for Study 2, a priority was ensuring that the student actors felt comfortable and were not pushed to do any actions that they were not comfortable performing, resulting in potentially less of a difference between conditions. In future research using filmed 360-degree videos, we would recommend using more experienced actors (although more difficult in this case where the characters were student age) and hiring an intimacy co-ordinator when filming. Another related limitation of our Study 2 design worth highlighting at this stage relates to the VR scene filming. In order to make the participant feel as if they are viewing from the first-person perspective of someone at the party, instead of having a fly on the wall point-of-view, and to ensure that they would witness the assault in full, the 360 camera was positioned in the middle of the room. The woman in the red dress (victim) was in one corner, and diagonally across from here in the other corner, was the man in the white t-shirt (perpetrator). As an unforeseen consequence of this, some participants mentioned that when the man was staring at the woman, it felt as if he was staring at the camera instead. As a result, this may have made some participants view themselves as a potential victim, instead of a bystander, which, for better or worse, could have influenced their decisions. There are several broader limitations worth mentioning. A key difference between our two studies which could explain the different findings is the sample composition. Our two samples differ in proportions representing genders, past sexual assault victimisation and past sexual harassment victimisation. With regards to gender, for Study 1, we were recruiting from a very large online pool, so were able to ensure a roughly equal split between male and female participants. This decision was based on research suggesting that gender could impact factors in the model (Burn, 2009; Katz et al., 2015; Mainwaring et al., 2022; Moschella-Smith et al., 2022). Whereas, in Study 2, we were recruiting from students limited by geographic location and subject, resulting in our final sample that was 82% women. This is likely due to 81% of psychology undergraduate students in the UK (from which we recruited most of our Study 2 sample) being female (Higher Education Statistics Agency, 2023), and women generally being more engaged in gender-based violence research (e.g. Bennett et al., 2014; Katz et al., 2015; Kistler et al., 2021; Pagliaro et al., 2021). Regarding past sexual assault victimisation, 38% of our Study 1 sample had previously been the victim of sexual assault, while we excluded those who had been assaulted from our Study 2 sample due to the emotionally impactful nature of the stimuli (Patil et al., 2014). Moreover, although a similar classification was used, Study 1’s percentage was much higher than what is seen in nationally representative data (16.64%; ONS, 2023a). However, this is likely due to the ONS survey using an age cut off, whereby they only included sexual assaults committed from a victim age of 16 onwards, while the questionnaire used here did not include an age restriction, thus capturing sexual assaults on minors too. We also had a much higher proportion of people who had been sexually harassed in our Study 1 sample (60.88%) compared to our Study 2 sample (28.21%). This difference could be due to the exclusion of people who have experienced sexual assault, which we indeed found in our Study 1 data to be positively correlated with sexual harassment experience (see Table 5), as well as screening out those who have experienced six or more symptoms of trauma related to experiences of sexual harassment in Study 2 only. Taking these differences, as well as findings from Study 1 and past research suggesting that these factors could be related to our variables of interest (Bennett et al., 2017; Brockdorf et al., 2023; Mauer et al., 2022; Moschella-Smith et al., 2022; Vonderhaar, 2015), we ran impromptu exploratory analysis after Study 2 (on Study 1 data) to see if these variables moderated any of the effects and relationships in our predictions for which we found no confidence interval overlap between Studies 1 and 2. The only incidence where we found this to be the case was for sexual assault victimisation moderating the relationship between victim blaming and noticing the event (stage one), so that this negative relationship was only significant for those who had previously been sexually assaulted. Although the logic behind this moderation is unclear at this stage, it is important to take this into account when considering this relationship in our findings. The lack of any other significant moderation pathways helps to strengthen our conclusions when considering explanations of the differences between Study 1’s and Study 2’s findings outside of these sample differences. This project’s specific sample was chosen to reflect the adult age group and occupation most vulnerable to sexual assault, there are of course issues related to generalising these findings beyond university students. Related to this, there are also issues of generalising beyond the stimuli scenario. In both studies, our stimuli described an assault which bystanders directly witnesses, which covers only up to 29% of sexual assaults (Casper et al., 2022; Hamby et al., 2016; Hart & Miethe, 2008; Planty, 2002; Revolt Sexual Assault & The Student Room, 2018). Indeed it is far more common for bystanders to not be present at all, or only be present to witness events leading up to an assault (Haikalis et al., 2018). Moreover, our scenario featured a victim and perpetrator who appeared not to know each other, so there could be issues generalising to a situation where they have a pre-existing relationship, as there could be greater ambiguity for bystanders in this situation (Shotland & Straw, 1976). There also may be issues generalising from a party setting, to other assault settings such as at the victim’s or perpetrator’s home or in a public outdoor space (ONS, 2021). Thus, further research is needed to explore if factors like severity and accepting responsibility to intervene are as key when witnessing sexual assaults in other contexts too, varying the victim, perpetrator and participant age and occupation, at what stage bystanders are present, pre-existing relationships, and the assault setting. Moreover, an additional limitation across both studies relates to measurement. The differing patterns according to which intervention intention measure was used could also be due to a methodological limitation of our behaviour-specific measure. As discussed in the Study 1 methods section, the behaviour-specific measure is calculated by listing 12 different possible intervention strategies and asking participants how likely they would be to use each strategy, and then averaging the likelihood score. It may have been the case that some participants would prefer some intervention strategies over others, for example due to the safety concerns of directly intervening. This would have meant that they might have selected that they would have been unlikely to intervene in direct ways, but very likely to intervene in indirect ways (e.g. talking to the victim afterwards, getting help from someone else). So although the participant may have been adamant that they would intervene in some way, this would not have been reflected in their mean score from the behaviour-specific measure. Thus, although adapted from an existing scale (Banyard & Moynihan, 2011), the behaviour-specific scale may need further development before use in future research. In terms of interpretation of our findings, these considerations suggest that perhaps caution should be given to pathways in these studies which were only found to predict one of our intervention intentions measures and not the other. And lastly, a key limitation for our research is the lack of causality. Beyond the few effects of our severity manipulation on our variables of interest, and our within-participants examination of changes in physiological arousal, it is important to emphasise that the rest of the analysis was purely correlational. Despite efforts to use a more conservative approach to mediation analysis to minimise the risks of spurious correlations, there is no guarantee of direction or causality. Predicted directions in our mediation models are purely hypothetical until further research has manipulated each element. Implications Firstly, these findings help to further our understanding of the NBJW and the bystander decision-making process, providing initial evidence for a link between two largely unconnected areas of social psychology. Secondly, this research helps to bolster Dawtry et al.’s (2020) arguments on the importance of emotionally impactful stimuli, and Study 2 demonstrates how filmed VR can be used effectively in a social psychological setting. However, although we can see clear distinctions in our findings from Studies 1 and 2, a more direct experimental comparison of text vignettes and filmed VR to confirm effects on emotional impact, BJW strategy use, empathic variables and bystander intervention is needed. Most importantly, this research has positive implications beyond academic research in informing the design of bystander intervention programmes. Universities are coming under increasing pressure to tackle sexual assault due to campus locations being a hotspot for assaults (Fenton et al., 2015; Revolt Sexual Assault & The Student Room, 2018). In the US, it is now a legal requirement of colleges to include bystander intervention training in their educational programmes designed to combat sexual violence ("Campus Sexual Violence Elimination Act," 2013). Despite a lack of equivalent legislation in the UK, many universities now have their own bystander intervention training schemes in place. Some training programmes like these include a focus on the role of Latané and Darley’s (1970) model (College of Policing, 2022; Fenton et al., 2015), and our findings demonstrating the importance of a bystander’s responsibility to intervene suggest that this could be an effective approach to take. However, our findings on the importance of psychological distancing in Study 1 and victim blaming in Study 2, suggest that there is need to take a more comprehensive approach to account for these roles. Such programmes could also be adapted for use in other settings, such as workplaces, schools, and community centres. Thus, these studies have positive implications for informing real-world efforts to reduce sexual assault, by encouraging bystanders to intervene. Conclusion To summarise, we conducted two studies to explore how BJW non-rational strategies might be linked to the bystander decision-making process, in the context of sexual assault. We found key differences in our findings across the two studies and offered potential explanations for these. However there are still several important results across both. Firstly, our severity manipulation had limited effects, but most notably higher severity made participants more likely to intervene, but without an obvious mechanism behind this. Mediation analysis suggested that in both studies, a BJW non-rational strategy was negatively related to the bystander’s general intentions to intervene. In study one, this relationship was mediated by reduced empathic concern and acceptance of responsibility to intervene, while in Study 2, there was no mediation among our variables of interest. Although explorative in nature and correlational, this research makes a first attempt at bringing together two previously unconnected areas of social psychology, casting a new perspective on theory and real-world sexual assault prevention. Declarations Ethical Approval: Prior to fieldwork commencing, this study was given ethical approval by the Psychology Research Ethics Committee at the University of Bath (PREC reference numbers: 22-013 & 22-095). CRediT author statement: Sarah Warbis: conceptualisation, methodology, formal analysis, investigation, resources, writing – original draft, visualisation, and project administration. Danaë Stanton Fraser: conceptualisation, methodology, writing – review & editing and supervision. Mitchell Callan: conceptualisation, methodology, writing – review & editing and supervision. Competing interests: There are no relevant financial or non-financial competing interests to report. Ethical Approval: Prior to fieldwork commencing, this study was given ethical approval by the Psychology Research Ethics Committee at the University of Bath (PREC reference numbers: 22-013 & 22-095). Data Availability Statement: Anonymous data created during this research from participants who all consented to data sharing, plus other supporting information, will be made available upon publication from the University of Bath Research Data Archive. Acknowledgements The authors would like to thank Dr Harry Farmer, Jonathan Yip, Niamh Mulhern, Yeeun Kim, Ben Robins, Nathan Taylor, as well as the actors who starred in our filmed scene, and the University of Bath for funding this research. References Aderman, D., Brehm, S. S., & Katz, L. B. (1974). Empathic observation of an innocent victim: The just world revisited. Journal of Personality and Social Psychology , 29 (3), 342-347. https://doi.org/10.1037/h0036001 Ahn, S. J., Le, A. M. T., & Bailenson, J. (2013). The effect of embodied experiences on self-other merging, attitude, and helping behavior. Media Psychology , 16 (1), 7-38. https://doi.org/10.1080/15213269.2012.755877 Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action Control (pp. 11-39). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-69746-3_2 Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review , 84 (2), 191-215. https://doi.org/https://doi.org/10.1037/0033-295X.84.2.191 Baños, R. M., Botella, C., Rubió, I., Quero, S., García-Palacios, A., & Alcañiz, M. (2008). Presence and emotions in virtual environments: The influence of stereoscopy. CyberPsychology & Behavior , 11 (1), 1-8. https://doi.org/10.1089/cpb.2007.9936 Banyard, V. L., & Moynihan, M. M. (2011). Variation in bystander behavior related to sexual and intimate partner violence prevention: Correlates in a sample of college students. Psychology of Violence , 1 (4), 287-301. https://doi.org/10.1037/a0023544 Banyard, V. L., Moynihan, M. M., & Plante, E. G. (2007). Sexual violence prevention through bystander education: An experimental evaluation. Journal of Community Psychology , 35 (4), 463-481. https://doi.org/10.1002/jcop.20159 Banyard, V. L., Plante, E. G., & Moynihan, M. M. (2004). Bystander education: Bringing a broader community perspective to sexual violence prevention. Journal of Community Psychology , 32 (1), 61-79. https://doi.org/10.1002/jcop.10078 Barreda-Ángeles, M., Serra-Blasco, M., Trepat, E., Pereda-Baños, A., Pàmias, M., Palao, D., . . . Cardoner, N. (2021). Development and experimental validation of a dataset of 360°-videos for facilitating school-based bullying prevention programs. Computers & Education , 161 , 104065. https://doi.org/10.1016/j.compedu.2020.104065 Bennett, S., Banyard, V. L., & Edwards, K. M. (2017). The Impact of the Bystander's Relationship With the Victim and the Perpetrator on Intent to Help in Situations Involving Sexual Violence. J Interpers Violence , 32 (5), 682-702. https://doi.org/10.1177/0886260515586373 Bennett, S., Banyard, V. L., & Garnhart, L. (2014). To act or not to act, that is the question? Barriers and facilitators of bystander intervention. Journal of Interpersonal Violence , 29 (3), 476-496. https://doi.org/10.1177/0886260513505210 Blascovich, J., Loomis, J., Beall, A.C., Swinth, K.S., Hoyt, C.L., & Bailenson J.N. (2002). Immersive Virtual Environment Technology as a Methodological Tool for Social Psychology. Psychological Inquiry , 13 (2), 103-125. Boppana, S. (2021). When men are the victims: Factors affecting rape victim blame & bystander aid [Doctoral dissertation, University of Mississippi]. eGrove. https://egrove.olemiss.edu/etd/2149 Bowes-Sperry, L. (2005). To act or not to act: The dilemma faced by sexual harassment observers. Academy of Management Review , 30 (2), 288-307. https://doi.org/10.5465/AMR.2005.16387886 Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The self-assessment manikin and the semantic differential. Journal of Behavior Therapy and Experimental Psychiatry , 25 (1), 49-59. https://doi.org/https://doi.org/10.1016/0005-7916(94)90063-9 Brewin, C. R., Rose, S., Andrews, B., Green, J., Tata, P., McEvedy, C., . . . Foa, E. B. (2002). Brief screening instrument for post-traumatic stress disorder. British Journal of Psychiatry , 181 (02), 158-162. https://doi.org/10.1017/s0007125000161896 Brockdorf, A. N., Eagan, S. R., Zosky, J. E., Grandgenett, H. M., Gervais, S. J., & Dilillo, D. (2023). Associations Between Sexual Objectification and Bystander Efficacy: The Mediating Role of Barriers to Bystander Intervention [Article]. Violence and Victims , 38 (1), 148-164. https://doi.org/10.1891/VV-2021-0168 Burn, S. M. (2009). A situational model of sexual assault prevention through bystander intervention. Sex Roles , 60 (11/12), 779-793. https://doi.org/10.1007/s11199-008-9581-5 Burt, M. R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology , 38 (2), 217-230. https://doi.org/10.1037/0022-3514.38.2.217 Campus Sexual Violence Elimination Act, S.128 (2013). https://www.congress.gov/bill/113th-congress/senate-bill/128/text Carey, K. B., Norris, A. L., Durney, S. E., Shepardson, R. L., & Carey, M. P. (2018). Mental health consequences of sexual assault among first-year college women. Journal of American College Health , 66 (6), 480-486. https://doi.org/10.1080/07448481.2018.1431915 Casper, D. M., Witte, T., Gibson, E., & McCulley, E. (2022). “I Pulled Them Apart and Told Them to Stop”: A Mixed-Methods Examination of Bystander Behavior, Preparedness, and Emotional Reaction. Journal of Interpersonal Violence . https://doi.org/10.1177/08862605221092071 Chauhan, M. (2022). Gender and Age Related Differences in Attitude towards Rape and Acceptance of Rape Myths. The International Journal of Indian Psychology , 10 (3). https://doi.org/10.25215/1003.130 College of Policing. (2022). Bystander Programmes: Evidence briefing . Retrieved from https://assets.college.police.uk/s3fs-public/2022-03/Bystander-programmes-evidence-briefing.pdf Correia, I., & Vala, J. (2003). When will a victim be secondarily victimized? The effect of observer's belief in a just world, victim's innocence and persistence of suffering. Social Justice Research , 16 (4), 379-400. https://doi.org/10.1023/A:1026313716185 Darley, J. M., & Latané, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology , 8 (4), 377-383. https://doi.org/10.1037/h0025589 Davis, M. H. (1980). A multidimensional approach to individual differences in empathy. JSAS catalog of selected documents in psychology , 10 , 85. Dawtry, R. J., Callan, M. J., Harvey, A. J., & Gheorghiu, A. I. (2020). Victims, vignettes, and videos: Meta-analytic and experimental evidence that emotional impact enhances the derogation of innocent victims. Personality and Social Psychology Review , 24 (3), 233-259. https://doi.org/10.1177/1088868320914208 Dawtry, R. J., Callan, M. J., Harvey, A. J., & Olson, J. M. (2018). Derogating innocent victims: The effects of relative versus absolute character judgments. Personality and Social Psychology Bulletin . https://doi.org/10.1177/0146167217733078 De La Peña, N., Weil, P., Llobera, J., Spanlang, B., Friedman, D., Sanchez-Vives, M. V., & Slater, M. (2010). Immersive Journalism: Immersive Virtual Reality for the First-Person Experience of News. Presence: Teleoperators and Virtual Environments , 19 (4), 291-301. https://doi.org/10.1162/pres_a_00005 Embury-Dennis, T. (2018, March 30). Judge who asked alleged rape victim if she tried to ‘close her legs’ charged with misconduct. The Independent . https://www.independent.co.uk/news/world/americas/judge-rape-victim-close-your-legs-john-russo-new-jersey-complaint-a8281356.html Fenton, R. A., Mott, H. L., McCartan, K., & Rumney, P. N. (2015). A review of evidence for bystander intervention to prevent sexual and domestic violence in universities . Public Health England. https://www.gov.uk/government/publications/sexual-and-domestic-violence-prevention-in-universities-evidence-review Fernández, C., Pascual, J. C., Soler, J., Elices, M., Portella, M. J., & Fernández-Abascal, E. (2012). Physiological Responses Induced by Emotion-Eliciting Films. Applied Psychophysiology and Biofeedback , 37 (2), 73-79. https://doi.org/10.1007/s10484-012-9180-7 Finch, E., & Munro, V. E. (2007). The demon drink and the demonized woman: Socio-sexual stereotypes and responsibility attribution in rape trials involving intoxicants. Social & Legal Studies , 16 (4), 591-614. https://doi.org/10.1177/0964663907082737 Fine, M. (1983). The social context and a sense of injustice: The option to challenge. Representative Research in Social Psychology , 13 (1), 15-33. Fischer, P., Krueger, J. I., Greitemeyer, T., Vogrincic, C., Kastenmüller, A., Frey, D., . . . Kainbacher, M. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous and non-dangerous emergencies. Psychological Bulletin , 137 (4), 517-537. https://doi.org/10.1037/a0023304 Galop. (2022). LGBT+ People & Sexual Violence . https://galop.org.uk/resource/lgbt-people-sexual-violence-report/ Gerber, G. L., Cronin, J. M., & Steigman, H. J. (2004). Attributions of blame in sexual assault to perpetrators and victims of both genders. Journal of Applied Social Psychology , 34 (10), 2149-2165. https://doi.org/10.1111/j.1559-1816.2004.tb02694.x Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism. Journal of Personality and Social Psychology , 70 (3), 491-512. https://doi.org/10.1037/0022-3514.70.3.491 Greco, A., Valenza, G., Citi, L., & Scilingo, E. P. (2016). Arousal and valence recognition of affective sounds based on electrodermal activity. IEEE Sensors Journal , 17 (3), 716-725. Hafer, C. L. (2000). Investment in long-term goals and commitment to just means drive the need to believe in a just world. Personality and Social Psychology Bulletin , 26 (9), 1059-1073. https://doi.org/10.1177/01461672002611004 Hafer, C. L., & Bègue, L. (2005). Experimental research on just-world theory: Problems, developments, and future challenges. Psychological Bulletin , 131 (1), 128-167. https://doi.org/10.1037/0033-2909.131.1.128 Hafer, C. L., & Rubel, A. N. (2015). The why and how of defending belief in a just world. Advances in Experimental Social Psychology , 51 (1), 41-96. https://doi.org/10.1016/bs.aesp.2014.09.001 Hafer, C. L., Rubel, A. N., & Drolet, C. E. (2019). Experimental evidence of subtle victim blame in the absence of explicit blame. PLOS ONE , 14 (12), e0227229. https://doi.org/10.1371/journal.pone.0227229 Haikalis, M., Leone, R. M., Parrott, D. J., & Dilillo, D. (2018). Sexual Assault Survivor Reports of Missed Bystander Opportunities: The Role of Alcohol, Sexual Objectification, and Relational Factors. Violence Against Women , 24 (10), 1232-1254. https://doi.org/10.1177/1077801218781941 Hamby, S., Weber, M. C., Grych, J., & Banyard, V. (2016). What difference do bystanders make? The association of bystander involvement with victim outcomes in a community sample. Psychology of Violence , 6 (1), 91-102. https://doi.org/10.1037/a0039073 Hart, T. C., & Miethe, T. D. (2008). Exploring bystander presence and intervention in nonfatal violent victimization: When does helping really help? Violence and Victims , 23 (5), 637-651. https://doi.org/10.1891/0886-6708.23.5.637 Harvey, A. J., Callan, M. J., & Matthews, W. J. (2014). How much does effortful thinking underlie observers’ reactions to victimization? Social Justice Research , 27 (2), 175-208. https://doi.org/10.1007/s11211-014-0209-3 Hayes, A. F. (2022). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (3rd ed.). The Guilford Press. Higher Education Statistics Agency. (2023, January 19). Higher Education Student Statistics: UK, 2021/22 . https://www.hesa.ac.uk/news/19-01-2023/sb265-higher-education-student-statistics/subjects Home Office. (2023). Crime outcomes in England and Wales, year ending March 2023: Data tables [Data tables]. Retrieved December 21, 2023 from https://www.gov.uk/government/statistics/crime-outcomes-in-england-and-wales-2022-to-2023 Hortensius, R., & de Gelder, B. (2014). The neural basis of the bystander effect — The influence of group size on neural activity when witnessing an emergency. NeuroImage (Orlando, Fla.) , 93 (1), 53-58. https://doi.org/10.1016/j.neuroimage.2014.02.025 Huang, L., Li, W., Xu, Z., Sun, H., Ai, D., Hu, Y., . . . Zhou, Y. (2023). The Severity of Cyberbullying Affects Bystander Intervention Among College Students: The Roles of Feelings of Responsibility and Empathy [Article]. Psychology Research and Behavior Management , 16 , 893-903. https://doi.org/10.2147/PRBM.S397770 Jicol, C., Cheng, H. Y., Petrini, K., & O’Neill, E. (2023). A predictive model for understanding the role of emotion for the formation of presence in virtual reality. PLOS ONE , 18 (3), e0280390. https://doi.org/10.1371/journal.pone.0280390 Jicol, C., Feltham, J., Yoon, J., Proulx, M. J., O'Neill, E., & Lutteroth, C. (2022, April 29-May 5, 2022). Designing and Assessing a Virtual Reality Simulation to Build Resilience to Street Harassment. CHI Conference on Human Factors in Computing Systems, New Orleans, LA, USA. Jordan, M. R., Amir, D., & Bloom, P. (2016). Are empathy and concern psychologically distinct? Emotion , 16 (8), 1107-1116. https://doi.org/10.1037/emo0000228 Jouriles, E. N., McDonald, R., Rosenfield, D., & Sargent, K. S. (2019). Increasing bystander behavior to prevent adolescent relationship violence: A randomized controlled trial. Journal of Consulting and Clinical Psychology , 87 (1), 3-15. https://doi.org/10.1037/ccp0000355 Kahn, A. S., & Cargile, A. C. (2021). Immersive and Interactive Awe: Evoking Awe via Presence in Virtual Reality and Online Videos to Prompt Prosocial Behavior. Human Communication Research , 47 (4), 387-417. https://doi.org/10.1093/hcr/hqab007 Katz, J. (2015). Effects of Group Status and Victim Sex on Male Bystanders' Responses to a Potential Party Rape. Journal of Aggression, Maltreatment & Trauma , 24 (5), 588-602. https://doi.org/10.1080/10926771.2015.1029184 Katz, J., Olin, R., Herman, C., & DuBois, M. (2013). Spotting the signs: First‐year college students'responses to bystander‐themed rape prevention posters. Journal of Community Psychology , 41 (4), 523-529. Katz, J., Pazienza, R., Olin, R., & Rich, H. (2015). That’s what friends are for. Journal of Interpersonal Violence , 30 (16), 2775-2792. https://doi.org/10.1177/0886260514554290 Kistler, T., Stuart, G. L., & Cornelius, T. L. (2021). Barriers to Bystander Interventions for Sexual Assault: Impact of Personal History of Sexual Victimization. Journal of Interpersonal Violence , 37 (17-18), NP16727-NP16749. https://doi.org/10.1177/08862605211023491 Kreibig, S. D., Wilhelm, F. H., Roth, W. T., & Gross, J. J. (2007). Cardiovascular, electrodermal, and respiratory response patterns to fear‐ and sadness‐inducing films. Psychophysiology , 44 (5), 787-806. https://doi.org/10.1111/j.1469-8986.2007.00550.x Labhardt, D., Holdsworth, E., Brown, S., & Howat, D. (2017). You see but you do not observe: A review of bystander intervention and sexual assault on university campuses. Aggression and Violent Behavior , 35 , 13-25. https://doi.org/10.1016/j.avb.2017.05.005 Latané, B., & Darley, J. M. (1970). The unresponsive bystander: Why doesn’t he help? Appleton Century Crofts. Latta, R. (1976). An attribution theory interpretation of reactions to victims [Doctoral Dissertation, Iowa State University]. Iowa State University Digital Repository. https://dr.lib.iastate.edu/handle/20.500.12876/78338 Lerner, M. (1980). The Belief in a Just World : A Fundamental Delusion (1st ed.). New York, NY : Springer US : Imprint: Springer. Lerner, M. J., & Miller, D. T. (1978). Just world research and the attribution process: Looking back and ahead. Psychological Bulletin , 85 (5), 1030-1051. https://doi.org/10.1037/0033-2909.85.5.1030 Lerner, M. J., & Simmons, C. H. (1966). Observer's reaction to the "innocent victim": Compassion or rejection? Journal of Personality and Social Psychology , 4 (2), 203-210. https://doi.org/10.1037/h0023562 Levine, M. (2005). Identity and Emergency Intervention: How Social Group Membership and Inclusiveness of Group Boundaries Shape Helping Behavior. Personality & Social Psychology Bulletin , 31 (4), 443-454. https://doi.org/10.1177/0146167204271651 Li, X., Lu, H., Wang, H., Zhu, P., & Zhang, J. (2018). General belief in a just world, moral disengagement, and helping propensity in emergencies. Social Behavior & Personality: An International Journal , 46 (11), 1923-1937. https://doi.org/10.2224/sbp.7407 Linhares, L., & Torres, A. R. R. (2021). She deserved it: Analysis of variables that influence the accountability of victims of sexual violence. Acta Colombiana de Psicología , 25 (1), 218-229. Luce, H., Schrager, S. B., & Gilchrist, V. (2010). Sexual assault of women. American Family Physician , 81 (4), 489-495. Mainwaring, C., Gabbert, F., & Scott, A. J. (2022). A systematic review exploring variables related to bystander intervention in sexual violence contexts. Trauma, Violence, and Abuse , 0 (0), 1-16. https://doi.org/10.1177/15248380221079660 Mason, F., & Lodrick, Z. (2013). Psychological consequences of sexual assault. Best Practice & Research Clinical Obstetrics & Gynaecology , 27 (1), 27-37. https://doi.org/10.1016/j.bpobgyn.2012.08.015 Mauer, V. A., Edwards, K. M., Waterman, E. A., Dardis, C. M., Dworkin, E. R., Rodriguez, L. M., & Ullman, S. E. (2022). Disclosure recipients’ perceptions related to helping victims of dating and sexual violence. Journal of Interpersonal Violence , 0 (0), 1-24. https://doi.org/10.1177/08862605211063004 McFarlane, J., Malecha, A., Watson, K., Gist, J., Batten, E., Hall, I., & Smith, S. (2005). Intimate partner sexual assault against women: Frequency, health consequences, and treatment outcomes. Obstetrics & Gynecology , 105 (1), 99-108. https://doi.org/10.1097/01.AOG.0000146641.98665.b6 Montada, L., & Lerner, M. J. (1998). Responses to Victimizations and Belief in a Just World (1st ed.). Plenum Press. Moschella-Smith, E. A., Potter, S. J., & Moynihan, M. M. (2022). Correlates of bystander intentions and behavior among community college students in situations of sexual violence. Journal of Prevention , 43 , 277-295. https://doi.org/10.1007/s10935-021-00663-y Navarrete, C. D., McDonald, M. M., Mott, M. L., & Asher, B. (2012). Virtual morality: Emotion and action in a simulated three-dimensional “trolley problem”. Emotion , 12 (2), 364-370. https://doi.org/10.1037/a0025561 Nickerson, A. B. (2014). Measurement of the bystander intervention model for bullying and sexual harassment. Journal of Adolescence , 37 (4), 391-401. https://doi.org/10.1016/j.adolescence.2014.03.003 Office for National Statistics. (2021). Nature of sexual assault by rape or penetration, England and Wales: Year ending March 2020 . Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/ natureofsexualassaultbyrapeorpenetrationenglandandwales/yearendingmarch 2020#sources-of-support-sought-by-victims Office for National Statistics. (2023a). Sexual offences prevalence and victim characteristics, England and Wales [Data tables]. Retrieved January 26, 2022, from https://www.ons.gov.uk/peoplepopulationandcommunity/ crimeandjustice/datasets/sexualoffencesprevalenceandvictimcharacteristicsenglandandwales Office for National Statistics. (2023b). Sexual offences victim characteristics, England and Wales: year ending March 2022 . Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/ sexualoffencesvictimcharacteristicsenglandandwales/yearendingmarch2022/pdf Oppenheim, M. (2022 August 18). Judge dismisses woman’s sexual assault complaint because she wore ‘provocative clothes’. The Independent . https://www.independent.co.uk/asia/sexual-assault-judge-india-provocative-clothes-b2147722.html Orchowski, L. M., Oesterle, D. W., & Haikalis, M. (2022). What Stops Unwanted Sexual and Social Advances Made by Heavy Drinking College Men? Journal of Interpersonal Violence . https://doi.org/10.1177/08862605211072157 Pagliaro, S., Paolini, D., & Pacilli, M. G. (2021). Intimate Partner Violence and Same-Sex Couples: Examining the Antecedents of the Helping Intentions of Bystanders. Journal of Interpersonal Violence , 36 (21-22), NP11593-NP11617. https://doi.org/10.1177/0886260519888530 Pan, X., & Slater, M. (2011, 2011). Confronting a Moral Dilemma in Virtual Reality: A Pilot Study. Patil, I., Cogoni, C., Zangrando, N., Chittaro, L., & Silani, G. (2014). Affective basis of judgment-behavior discrepancy in virtual experiences of moral dilemmas. Social Neuroscience , 9 (1), 94-107. https://doi.org/10.1080/17470919.2013.870091 Patil, I., Zanon, M., Novembre, G., Zangrando, N., Chittaro, L., & Silani, G. (2018). Neuroanatomical basis of concern-based altruism in virtual environment. Neuropsychologia , 116 (Pt A), 34-43. https://doi.org/10.1016/j.neuropsychologia.2017.02.015 Pazda, A. D., Elliot, A. J., & Greitemeyer, T. (2012). Sexy red: Perceived sexual receptivity mediates the red-attraction relation in men viewing woman. Journal of Experimental Social Psychology , 48 (3), 787-790. https://doi.org/https://doi.org/10.1016/j.jesp.2011.12.009 Planty, M. (2002). Third-party involvement in violent crime, 1993-99 . U.S. Department of Justice: Office of Justice Programs. https://www.ojp.gov/ncjrs/virtual-library/abstracts/third-party-involvement-violent-crime-1993-99 Revolt Sexual Assault, & The Student Room. (2018). Sexual Violence at Universities Statistical Report . https://revoltsexualassault.com/research/ Russell, J. A., & Mehrabian, A. (1977). Evidence for a three-factor theory of emotions. Journal of Research in Personality , 11 (3), 273-294. https://doi.org/https://doi.org/10.1016/0092-6566(77)90037-X Schaaff, K., & Adam, M. T. (2013). Measuring emotional arousal for online applications: Evaluation of ultra-short term heart rate variability measures. 2013 Humaine association conference on affective computing and intelligent interaction, Seinfeld, S., Arroyo-Palacios, J., Iruretagoyena, G., Hortensius, R., Zapata, L. E., Borland, D., . . . Sanchez-Vives, M. V. (2018). Offenders become the victim in virtual reality: impact of changing perspective in domestic violence. Scientific Reports , 8 (1). https://doi.org/10.1038/s41598-018-19987-7 Shotland, R. L., & Straw, M. K. (1976). Bystander response to an assault: When a man attacks a woman. Journal of Personality and Social Psychology , 34 (5), 990-999. https://doi.org/10.1037/0022-3514.34.5.990 Singer, T., Seymour, B., O'Doherty, J. P., Stephan, K. E., Dolan, R. J., & Frith, C. D. (2006). Empathic neural responses are modulated by the perceived fairness of others. Nature , 439 (7075), 466-469. https://doi.org/10.1038/nature04271 Slater, M. (2018). Immersion and the illusion of presence in virtual reality. British Journal of Psychology , 109 (3), 431-434. https://doi.org/10.1111/bjop.12305 Slater, M., Antley, A., Davison, A., Swapp, D., Guger, C., Barker, C., . . . Sanchez-Vives, M. V. (2006). A virtual reprise of the Stanley Milgram obedience experiments. PLOS ONE , 1 (1), e39-e39. https://doi.org/10.1371/journal.pone.0000039 Slater, M., Rovira, A., Southern, R., Swapp, D., Zhang, J. J., Campbell, C., & Levine, M. (2013). Bystander Responses to a Violent Incident in an Immersive Virtual Environment. PLOS ONE , 8 (1), e52766-e52766. https://doi.org/10.1371/journal.pone.0052766 Slater, M., Usoh, M., & Steed, A. (1994). Depth of Presence in Immersive Virtual Environments. Teleoperators and Virtual Environments - Presence , 3 . Smith, C. A., & Frieze, I. H. (2003). Examining rape empathy from the perspective of the victim and the assailant. Journal of Applied Social Psychology , 33 (3), 476-498. https://doi.org/10.1111/j.1559-1816.2003.tb01907.x Spaccatini, F., Pacilli, M. G., Pagliaro, S., & Giovannelli, I. (2022). Victim blaming 2.0: Blaming sexualized victims of online harassment lowers bystanders’ helping intentions. Current Psychology . https://doi.org/10.1007/s12144-022-02884-8 Steele, C. M., & Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous effects. American Psychologist , 45 (8), 921-933. https://doi.org/10.1037/0003-066X.45.8.921 Stöber, J. (2001). The Social Desirability Scale-17 (SDS-17): Convergent validity, discriminant validity, and relationship with age. European Journal of Psychological Assessment , 17 (3), 222-232. https://doi.org/10.1027/1015-5759.17.3.222 Strömwall, L. A., Landström, S., & Alfredsson, H. (2014). Perpetrator characteristics and blame attributions in a stranger rape situation. The European Journal of Psychology Applied to Legal Context , 6 (2), 63-67. https://doi.org/10.1016/j.ejpal.2014.06.002 Sutton, T. M., Herbert, A. M., & Clark, D. Q. (2019). Valence, arousal, and dominance ratings for facial stimuli. Quarterly Journal of Experimental Psychology , 72 (8), 2046-2055. Tassinari, M., Aulbach, M. B., & Jasinskaja-Lahti, I. (2022). The use of virtual reality in studying prejudice and its reduction: A systematic review [Review]. PLOS ONE , 17 (7 July), Article e0270748. https://doi.org/10.1371/journal.pone.0270748 TED. (2015, April 22). Chris Milk: How virtual reality can create the ultimate empathy machine [Video]. YouTube. https://www.youtube.com/watch?v=iXHil1TPxvA Thornberg, R. (2013). Bystander behavior in bullying situations: Basic moral sensitivity, moral disengagement and defender self-efficacy. Journal of Adolescence , 36 (3), 475-484. https://doi.org/10.1016/j.adolescence.2013.02.003 Ullman, S. E. (2007). A 10-year update of “Review and critique of empirical studies of rape avoidance”. Criminal Justice and Behavior , 34 (3), 411-429. https://doi.org/10.1177/0093854806297117 Van Gelder, J.-L., Nee, C., Otte, M., Demetriou, A., Van Sintemaartensdijk, I., & Van Prooijen, J.-W. (2017). Virtual Burglary: Exploring the Potential of Virtual Reality to Study Burglary in Action. Journal of Research in Crime and Delinquency , 54 (1), 29-62. https://doi.org/10.1177/0022427816663997 Ventura, S., Cardenas, G., Miragall, M., Riva, G., & Baños, R. (2021). How Does It Feel to Be a Woman Victim of Sexual Harassment? The Effect of 360°-Video-Based Virtual Reality on Empathy and Related Variables. Cyberpsychology, Behavior and Social Networking , 24 (4), 258-266. https://doi.org/10.1089/cyber.2020.0209 Vonderhaar, R. L. (2015). There are no “innocent victims”: The influence of just world beliefs and prior victimization on rape myth acceptance. Journal of Interpersonal Violence , 30 (10), 1615-1633. https://doi.org/10.1177/0886260514549196 Walster, E., Berscheid, E., & Walster, G. W. (1973). New directions in equity research. Journal of Personality and Social Psychology , 25 (2), 151-176. https://doi.org/10.1037/h0033967 Walters, J. T. R., Bisson, J. I., & Shepherd, J. P. (2007). Predicting post-traumatic stress disorder: validation of the Trauma Screening Questionnaire in victims of assault. Psychological Medicine , 37 (1), 143-150. https://doi.org/10.1017/s0033291706008658 Wang, S. (2020). Standing up or standing by: Bystander intervention in cyberbullying on social media. New Media & Society , 23 (6), 1379-1397. https://doi.org/10.1177/1461444820902541 Warner, R. H., Vandeursen, M. J., & Pope, A. R. D. (2012). Temporal distance as a determinant of just world strategy: Mental Time Travel: Social Psychological Perspectives on a Fundamental Human Capacity. European Journal of Social Psychology , 42 (3), 276-284. White, K., MacDonnell, R., & Ellard, J. H. (2012). Belief in a just world: Consumer intentions and behaviors toward ethical products. Journal of Marketing , 76 (1), 103-119. https://doi.org/10.1509/jm.09.0581 Zhao, Y., Chu, X., & Rong, K. (2023). Cyberbullying experience and bystander behavior in cyberbullying incidents: The serial mediating roles of perceived incident severity and empathy [Article]. Computers in Human Behavior , 138 , Article 107484. https://doi.org/10.1016/j.chb.2022.107484 Zilkens, R. R., Smith, D. A., Kelly, M. C., Mukhtar, S. A., Semmens, J. B., & Phillips, M. A. (2017). Sexual assault and general body injuries: A detailed cross-sectional Australian study of 1163 women. Forensic Science International , 279 , 112-120. https://doi.org/https://doi.org/10.1016/j.forsciint.2017.08.001 Tables Tables 1 to 7 are available in the Supplementary Files section Additional Declarations The authors declare no competing interests. Supplementary Files SexualAssaultinaJustWorldSup.Materials271124.docx Supplementary Materials SexualAssaultinaJustWorldTable1271124.docx Table 1 SexualAssaultinaJustWorldTable2271124.docx Table 2 SexualAssaultinaJustWorldTable3271124.docx Table 3 SexualAssaultinaJustWorldTable4271124.docx Table 4 SexualAssaultinaJustWorldTable5271124.docx Table 5 SexualAssaultinaJustWorldTable6271124.docx Table 6 SexualAssaultinaJustWorldTable7271124.docx Table 7 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6008933","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":414335458,"identity":"3c403671-ae95-4ac5-9440-95776eb69a2b","order_by":0,"name":"Sarah Warbis","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxUlEQVRIiWNgGAWjYHACxsd/Kg4wMEhAeAbEaGE24DlzgIGHFC1sArxtpGiRb+8xY5Ccd0feXrqB8cMPhsPGBLUYnDlj9sBw2zPDHpkDzJI9DIfNCGuRyN1ukLjtMGOPRAKDNAPDYRvCDpv/dpvEwTmH7YFamH8TpYXhBu82ycaGw4lALWwgW4hw2Jn8z8YMx54l99xIbLPsMUgn7H359mOJjxlq7ti2z0g+fONHhbVhA2GXwQFjA5EROQpGwSgYBaOAIAAA1go8QzuUY3oAAAAASUVORK5CYII=","orcid":"","institution":"University of Bath","correspondingAuthor":true,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Warbis","suffix":""},{"id":414335459,"identity":"c4f30c24-4a7a-4025-aedf-51b258431b6a","order_by":1,"name":"Danaë Stanton Fraser","email":"","orcid":"","institution":"University of Bath","correspondingAuthor":false,"prefix":"","firstName":"Danaë","middleName":"Stanton","lastName":"Fraser","suffix":""},{"id":414335460,"identity":"5712ddfc-1cbb-4857-831d-71779b9123de","order_by":2,"name":"Mitchell Callan","email":"","orcid":"","institution":"University of Bath","correspondingAuthor":false,"prefix":"","firstName":"Mitchell","middleName":"","lastName":"Callan","suffix":""}],"badges":[],"createdAt":"2025-02-11 16:02:03","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6008933/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6008933/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":77037123,"identity":"0c1ae679-e8a9-41d4-abba-23b32e996ff5","added_by":"auto","created_at":"2025-02-24 13:34:20","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":103637,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003ePredicted Paths Between Non-Rational BJW Strategies, Empathic Variables and the Bystander Decision-Making Process, in Response to a Sexual Assault\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/2ca17cb681d1a7f78426e0ad.png"},{"id":77037127,"identity":"42e2ef78-cccf-4150-b31e-80a083009d4d","added_by":"auto","created_at":"2025-02-24 13:34:20","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":610486,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eImages from the 360° Virtual Reality Stimuli Scene in Study Two Showing the Key Events\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNote. \u003c/em\u003ePanel A: Man in the white shirt staring at the woman in the red dress. Panel B: Woman in the red dress looking uncomfortable upon noticing staring. Panel C: Man in the white shirt approaches woman in the red dress and says sexual joke. Panel D: Woman in the red dress says she is not interested and tries to move away, but the man in white shirt puts his hands on her arms to stop her. Panel E: Man in the white shirt pulls the woman in the red dress close to him. Panel F: Man in the white shirt says second sexual comment and (in severe condition only) strokes the woman in the red dress’s upper thigh.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/e7d9368e3a4616c9ce5edbe8.png"},{"id":77039404,"identity":"72d0f419-aab4-4b96-9984-73451be2c2b2","added_by":"auto","created_at":"2025-02-24 13:50:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":164045,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eBradley and Lang’s (1994) Self-Assessment Manikin (SAM) Depicting Scales of Emotional Valence (top row), Arousal (middle row) and Dominance (bottom row).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/110bf890de69ca8094e74f23.png"},{"id":77039406,"identity":"1c1fc47a-deff-473a-8ea1-5f6887a56e4f","added_by":"auto","created_at":"2025-02-24 13:50:20","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":229527,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSnellen Chart Used in the VR Sight Test in Study Two.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/b068a248187fe114881d54c6.png"},{"id":77040733,"identity":"5b6ad88c-3372-4411-9d91-3d0f7baaba66","added_by":"auto","created_at":"2025-02-24 13:58:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2113370,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/afb691a8-d9b0-47d3-843f-44102e408971.pdf"},{"id":77039062,"identity":"a21e7471-2ffc-44a9-bbb3-19502b6df8ce","added_by":"auto","created_at":"2025-02-24 13:42:20","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":42262,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Materials\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldSup.Materials271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/3180ce2a7d2beb1ff2aaa66e.docx"},{"id":77037122,"identity":"6cc3d7e5-6fcc-4f58-ade9-5deb3dc1ce6e","added_by":"auto","created_at":"2025-02-24 13:34:20","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":64889,"visible":true,"origin":"","legend":"\u003cp\u003eTable 1\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable1271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/4cf477a330e78dc1ea6d0744.docx"},{"id":77039405,"identity":"4d16b8f8-29c2-4c90-91ca-fbbd2364ec67","added_by":"auto","created_at":"2025-02-24 13:50:20","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":67508,"visible":true,"origin":"","legend":"\u003cp\u003eTable 2\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable2271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/350d5f20c43cec014420629d.docx"},{"id":77039067,"identity":"76d54c2d-2919-4241-b090-b426a53ec285","added_by":"auto","created_at":"2025-02-24 13:42:20","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":81843,"visible":true,"origin":"","legend":"\u003cp\u003eTable 3\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable3271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/7a90927da7025101229b0be2.docx"},{"id":77037131,"identity":"a90d1bf6-c68d-446a-9dec-d45b2ef3565b","added_by":"auto","created_at":"2025-02-24 13:34:20","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":71735,"visible":true,"origin":"","legend":"\u003cp\u003eTable 4\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable4271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/b3e314523f6f3ea79fa2c046.docx"},{"id":77039066,"identity":"17c2da3b-c6a4-4462-816d-842b3b5e9e8a","added_by":"auto","created_at":"2025-02-24 13:42:20","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":83625,"visible":true,"origin":"","legend":"\u003cp\u003eTable 5\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable5271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/8f7145621f7417d66870d5aa.docx"},{"id":77039069,"identity":"17c1a7e2-52a5-4204-8f44-99d84f1ab033","added_by":"auto","created_at":"2025-02-24 13:42:21","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":69818,"visible":true,"origin":"","legend":"\u003cp\u003eTable 6\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable6271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/dfa702f062d46645b33cff5a.docx"},{"id":77037138,"identity":"2fffafbb-a9a5-41c2-899d-7c95ea676634","added_by":"auto","created_at":"2025-02-24 13:34:21","extension":"docx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":16285,"visible":true,"origin":"","legend":"\u003cp\u003eTable 7\u003c/p\u003e","description":"","filename":"SexualAssaultinaJustWorldTable7271124.docx","url":"https://assets-eu.researchsquare.com/files/rs-6008933/v1/dff511a3542f2808ac6a16ee.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eSexual Assault in a ‘Just World’: An Immersive VR Study of Bystanders\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLatan\u0026eacute; and Darley (1970) suggest that a bystander witnessing a sexual assault goes through a sequential process when deciding whether to intervene. When investigating how to utilise this process to encourage bystanders to intervene, research has largely ignored the need to believe in a just world (NBJW) and its strategies, a concept which explores how individuals restore a sense of justice following an injustice (Lerner, 1980). Across two studies this paper provides an initial exploration of the relationship between the bystander model and NBJW, aiming to explore if strategies used to defend a threat to the BJW affect this bystander process, and through that, the likelihood of the bystander intervening. The study also explores the use of virtual reality (VR) as stimuli compared to more traditionally-used text vignettes in this area of research.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSexual Assault\u003c/p\u003e\n\u003cp\u003eSexual assault is defined by the Office for National Statistics (ONS; 2023b) as any of the following: rape, assault by penetration using an object other than a penis, unwanted sexual touching or indecent exposure, including attempts of the first two, with no mention of force or threat being necessary. Using this definition, the ONS (2023a) found that in England and Wales, 27% women, 6% men have been the victim of sexual assault since the age of 16 (the age of consent in the UK), with other UK-based research finding a staggering 65% of LGBT+ people (including non-binary and intersex individuals) had been sexually assaulted during adulthood (Galop, 2022). Those at particular risk are adults under 24 years old and full-time university students (ONS, 2023a), with locations off-campus being the most common places for someone of this cohort to be sexually assaulted (O\u0026apos;Neal et al., 2021).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBeing a survivor of a sexual assault poses both short and long-term risks to physical health, for example, injuries, sexually transmitted diseases and chronic pain, as well as longer-term effects on mental health, including an increased risk of post-traumatic stress disorder (PTSD) and depression (Luce et al., 2010; McFarlane et al., 2005; Zilkens et al., 2017). This is in addition to the social consequences of the victimisation, with victims often being blamed for the assault (Finch \u0026amp; Munro, 2007; Lerner, 1980). Moreover, should the victim wish to proceed with criminal proceedings they stand a 3.7% chance of the perpetrator being charged for a sexual offense in England and Wales (Home Office, 2023). These stark statistics make this a key subject for academic, third sector and policy preventative research (Fenton et al., 2015; Revolt Sexual Assault \u0026amp; The Student Room, 2018; Ullman, 2007).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe Role of Bystanders\u003c/p\u003e\n\u003cp\u003eHistorically, programmes have put the burden for prevention on potential victims (Katz et al., 2013; Ullman, 2007), whereas more recent efforts have aimed to spread this responsibility to the entire community (Banyard et al., 2004). Research suggests that a bystander is present during up to 29% of sexual assaults, including those taking place at house parties (Haikalis et al., 2018). Yet, the most common response from a bystander is to do nothing\u0026nbsp;(Casper et al., 2022; Haikalis et al., 2018; Hamby et al., 2016; Hart \u0026amp; Miethe, 2008; Orchowski et al., 2022; Planty, 2002; Revolt Sexual Assault \u0026amp; The Student Room, 2018).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn order to enable behaviour change, research has been carried out to understand the process bystanders experience when witnessing an emergency event, such as a sexual assault. Latan\u0026eacute; and Darley (1970) proposed a model which outlined five key steps a bystander goes through in a progressive manner (although cycling back to previous stages is possible). Firstly, a bystander must notice the event, which can be impaired due to distractions or noise (Burn, 2009). Secondly, a bystander must decide that the event is an emergency whereby intervention is needed, which can be facilitated by knowledge of nonconsensual behaviour (Moschella-Smith et al., 2022), and impeded by ambiguity in the situation, such as being unsure if the victim wants help (Shotland \u0026amp; Straw, 1976). Thirdly, a bystander must accept personal responsibility to intervene, which is most notably prohibited by the presence of other inactive, unrelated bystanders (Darley \u0026amp; Latan\u0026eacute;, 1968), as well as being heightened by a preexisting relationship or shared group membership with the victim, perpetrator and other bystanders (Bennett et al., 2014; Burn, 2009; Fischer et al., 2011; Katz et al., 2015; Levine, 2005). Next, a bystander must decide how to intervene most effectively, which takes into account their level of available skills (Bennett et al., 2014), and their belief in their capability to successfully intervene with the desired outcome, known as self-efficacy (Bandura, 1977; Banyard \u0026amp; Moynihan, 2011; Thornberg, 2013). And lastly, a bystander must implement their intended intervention, and at this stage can be impacted by group norms and concern of how others may judge them (Banyard \u0026amp; Moynihan, 2011; Burn, 2009; Mainwaring et al., 2022).\u003c/p\u003e\n\u003cp\u003eThere is an abundance of evidence providing support for the importance of each stage of Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model. There is also evidence for the model as a whole, for example Burn (2009) demonstrated correlations between barriers to each stage and real life intervention in a sexual assault context, and Nickerson (2014) demonstrated how each stage was able to predict the next in sequence in a bullying and sexual harassment context in schools. However, no research as of yet has explored the mediated, indirect pathways between the stages in this model in the context of sexual assault.\u003c/p\u003e\n\u003cp\u003eAs well as research exploring the importance of the stages of Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model themselves in predicting bystander intervention (e.g. Burn, 2009; Nickerson, 2014), separate factors affecting the model have been studied, such as the presence of others and social norms (Banyard \u0026amp; Moynihan, 2011; Darley \u0026amp; Latan\u0026eacute;, 1968; Fischer et al., 2011; Mainwaring et al., 2022). However, one area which has been largely neglected in this exploration is the NBJW.\u003c/p\u003e\n\u003cp\u003eThe Need to Believe in a Just World\u003c/p\u003e\n\u003cp\u003eThe just world hypothesis states that there is a universal need to believe that we live in a generally fair world, where people get what they deserve; a belief which is essential for the pursuit of long-term goals (Hafer, 2000; Hafer \u0026amp; Rubel, 2015; Lerner, 1980; Lerner \u0026amp; Miller, 1978). Given that the world around us can be unfair, this need can be easily threatened by seeing something contradicting this belief, for example, by seeing an innocent person being sexually assaulted. Consequently, given its functional importance, individuals are motivated to defend this belief in the face of threat (Lerner, 1980).\u003c/p\u003e\n\u003cp\u003eAccording to Lerner (1980), individuals can resolve threat to their belief in a just world (BJW) in several ways. One way, is through using a rational strategy in which the individual tries to rebalance the scales by restoring justice to the situation, for example, through compensating a victim or, most crucially to this article, by intervening and stopping the injustice (Harvey et al., 2014; Lerner, 1980). In contrast, people can also use non-rational strategies to resolve the threat to their BJW which involve cognitively distorting the event to deny an injustice occurred to begin with (Hafer \u0026amp; B\u0026egrave;gue, 2005). These strategies include, among others, victim derogation, victim blaming and psychological distancing (Hafer \u0026amp; Rubel, 2015; Lerner, 1980). Victim derogation involves belittling the character of a victim in a way that they seem deserving of the bad thing that is happening to them (Lerner, 1980), and has been demonstrated to be enhanced in response to a threat to the BJW (Correia \u0026amp; Vala, 2003; Dawtry et al., 2020). Similarly, victim blaming works by criticising a victim\u0026rsquo;s behaviour, so that their actions appear to be to blame for the injustice (Lerner, 1980), and indeed personality literature has shown a link between a stronger BJW trait and greater victim blaming of sexual violence victims (Linhares \u0026amp; Torres, 2021; Str\u0026ouml;mwall et al., 2014). Psychological distancing works in a different way however, being a more subtle method of restoring BJW (Hafer \u0026amp; Rubel, 2015). By highlighting how different you are from the victim through disassociating yourself from them, this removes the concern that if this bad thing can happen to them, it could happen to me, and again has been shown to be higher when the threat to the BJW is greatest (Hafer, 2000). Thus, research supports Lerner\u0026rsquo;s (1980) theory that individuals will use rational and non-rational strategies to restore their BJW.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDespite the clear threat to the BJW that witnessing an injustice like a sexual assault could elicit and the high prevalence of victim blaming experienced by sexual assault survivors (Embury-Dennis, 2018, March 30; Finch \u0026amp; Munro, 2007; Oppenheim, 2022 August 18), very little research has explored how these strategies under the just world hypothesis could be related to bystander intervention when witnessing a sexual assault. Research to date suggests that greater victim blaming is related to lower intentions to intervene in rape, online harassment and injury contexts\u0026nbsp;(Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022). However, there are several pivotal limitations with how confidently one can apply existing findings to understand sexual assault bystanders. Firstly, research has often looked at the prevalence of victim blaming after a crime has been committed, for example as a court case. In such cases, rational strategies like intervention, which research suggests would be the preferred option (Lerner \u0026amp; Simmons, 1966), are no longer possible, meaning victim blaming rates could be inflated (Finch \u0026amp; Munro, 2007; Spaccatini et al., 2022). Secondly, a large body of literature in this field conceptualises the BJW as an individual differences trait upon which individuals vary, rather than as the universal NBJW as Lerner (1980) intended (Hafer \u0026amp; B\u0026egrave;gue, 2005), questioning whether this personality literature can be generalised. And most crucially, research to date has not fully explored how multiple BJW strategies are related to Latan\u0026eacute; and Darley\u0026rsquo;s (1970) entire model, instead cherry picking individual concepts\u0026nbsp;(Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022). All of which combined highlights the need for a comprehensive exploration of how multiple strategies which restore the BJW could be related to all stages of\u0026nbsp;Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model within the context of a bystander witnessing a sexual assault occurring.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhile there is limited research to date, it is possible to make some informed predictions of what this relationship might look like. When considering which strategies were preferred, research suggests that people are inclined to use rational strategies over non-rational strategies to resolve a threat to their BJW (Lerner \u0026amp; Simmons, 1966), but that non-rational strategies are used when the individual does not believe they can fully restore justice to the situation\u0026nbsp;(Montada \u0026amp; Lerner, 1998; Walster et al., 1973; White et al., 2012). When a bystander is witnessing a sexual assault already occurring, a bystander may be aware of the negative impacts the assault has likely already had on the victim\u0026rsquo;s mental and physical health (Carey et al., 2018; Luce et al., 2010; Mason \u0026amp; Lodrick, 2013; McFarlane et al., 2005), and as such it could be suggested that they would view the victim\u0026rsquo;s suffering as irreparable. Based on this, it could be argued that they may decide that intervention is futile and resort to non-rational strategies (Montada \u0026amp; Lerner, 1998; Walster et al., 1973; White et al., 2012).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIf proceeding with non-rational strategies, it could be hypothesised that victim derogation and blaming could negatively impact a bystander\u0026rsquo;s ability to see an assault as an emergency in which intervention is needed. If the victim is viewed as either a bad person or as having carried out a bad action which enabled something to happen to them, then the bystander may view the event as just, and therefore not requiring external intervention. These strategies could also impair the bystander\u0026rsquo;s ability to accept responsibility to intervene. After all, if the victim\u0026rsquo;s character or behaviour is responsible for what is happening to them, then the bystander may decide that they themselves are not responsible and therefore decide not to intervene.\u0026nbsp;Psychological distancing could work in a similar way by hindering a bystander\u0026rsquo;s ability to accept responsibility to act. Wang (2020) suggested that sharing similarities with someone fosters a sense of responsibility for them, so feeling dissimilar to the victim through psychological distancing may result in the reverse effect and reduce their sense of responsibility to intervene.\u0026nbsp;Therefore, we propose that derogating, blaming and psychologically distancing oneself from a victim, could collectively affect two stages in\u0026nbsp;Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model, namely, deciding that intervention is needed and accepting personal responsibility to do so. However, this needs to be tested empirically, so the influence on all stages of the bystander model should be considered. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe Role of Empathy and Empathic Concern\u003c/p\u003e\n\u003cp\u003eTwo related variables of interest to develop our understanding of the possible mechanisms behind how the NBJW and bystander intervention are related are empathy and empathic concern. Understood here as two distinct constructs (Jordan et al., 2016), empathy is the cognitive ability to take on another person\u0026rsquo;s perspective (Davis, 1980), whilst empathic concern goes beyond this to be an emotional concern for the welfare of another based on their circumstances (Davis, 1980; Jordan et al., 2016). Research has found links between empathy and perceptions of fairness and lower victim derogation (Aderman et al., 1974; Singer et al., 2006), as well as a link between greater empathic concern and lower victim blaming (Katz et al., 2015), all of which suggests that these empathic variables could be related to the NBJW. In addition to this, research has also shown links between increased empathy and increased helping (Ahn et al., 2013), as well as positive relationships between empathic concern and all five stages in Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model (Katz, 2015; Nickerson, 2014; Patil et al., 2018).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWith empirical evidence suggesting that empathy and empathic concern could play a role within both the NBJW and bystander intervention but little indication of directionality, there is grounds for a novel exploration of whether they are related to both areas in the same sexual assault bystander context, and whether they could act as potential mediators between the influence of the NBJW strategies on Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model. Moreover, given that past research has on occasions conflated empathy and empathic concern into one variable (Davis, 1980), or indeed used the terms interchangeably (Nickerson, 2014), despite evidence showing them to be distinct constructs (Jordan et al., 2016), it is important to understand which empathic variable could play a mediating role here. Based on our predictions related to the NBJW, we explored whether empathy and empathic concern mediated the relationship between BJW non-rational strategies and the stages of Latan\u0026eacute; and Darley\u0026rsquo;s (1970) model, and in particular, a bystander\u0026rsquo;s belief that intervention is needed and their ability to accept responsibility to act.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eResearch Methodology for Studying Bystanders\u003c/p\u003e\n\u003cp\u003eHaving established a theoretical approach for this research, another decision relates to the research methodology. Dawtry et al. (2020) found that inconsistencies in findings of experimental research on victim derogation could be explained by differences in the emotional impact (i.e. emotional engagement or threat) of the context for observers. Studies using methodologies that were more vivid, severe, felt more real, and were physically and chronologically closer, reported larger derogation effects. Thus, it is important to explore our hypothesised relationships using emotionally impactful stimuli. Past BJW research has utilised emotionally impactful methodologies like witnessing in-person suffering (Fine, 1983; Latta, 1976), watching CCTV footage (Lerner \u0026amp; Simmons, 1966), and watching recorded interviews with victims (Warner et al., 2012), but there are key limitations to each of these. Firstly, conducting a sexual assault bystander experiment featuring an in-person experience would present serious ethical concerns, especially as participants would likely believe this event to be real and would not have an opportunity to withdraw from the study if they found it distressing. Secondly, while CCTV footage and recorded interviews have high realism, both feel like a distant event physically and/or chronologically. However, there is another methodology gaining popularity in social psychological research which would score highly on all four of Dawtry et al.\u0026rsquo;s (2020) emotional impact criteria, namely, virtual reality (VR). VR was not included in Dawtry et al.\u0026rsquo;s (2020) meta-analysis.\u003c/p\u003e\n\u003cp\u003eVR involves artificial sensory information being used to perceptually surround a user, often using a head-mounted display (Blascovich, 2002). VR has already been used effectively in both bystander and sexual offense research (Jicol et al., 2022; Slater et al., 2013 ). For example, Slater et al. (2013) exposed participants in VR to a violent attack between two men to investigate if bystanders were more likely to intervene when the victim of the attack was a fan of the participant\u0026rsquo;s own football team (i.e. their ingroup). Using VR enabled the researchers to test a scenario that may not have been ethical to stage in real life nor easy to control, while also allowing for greater mundane realism than a vignette in a traditional experiment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eResearch has demonstrated the strong emotional impact of VR stimuli. Jicol et al. (2022) exposed participants to a sexual harassment victimisation simulation either on a flat-screen monitor or in VR, finding that participants had a strong negative emotional response in VR, as would be expected if being harassed in real life, and that this response was greater than the response to the flat-screen simulation. Moreover, because of the VR setting, they were able to measure this emotional impact both through self-reported measures and more objectively through physiological measures using heart rate (HR) as an indicator. Patil et al. (2014) similarly found that a moral dilemma was more emotionally arousing, measured through skin conductance, when presented in virtual reality, compared to when presented in a text format. Other research has shown VR to be capable of eliciting feelings of emotions like panic and positive affect in general, showing its versatility as an instrument (Ba\u0026ntilde;os et al., 2008; Pan \u0026amp; Slater, 2011).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eResearch has also found this to be the case when using a specific type of VR, namely, filmed virtual reality (Kahn \u0026amp; Cargile, 2021). Filmed VR involves recording a 2D video using a 360-degree camera and displaying this footage on a head-mounted display. Although animated VR is more widely used in research, filmed VR holds the benefits of time and cost efficiency, requiring much less training and specialised knowledge to develop than animated VR (Barreda-\u0026Aacute;ngeles et al., 2021). Kahn and Cargile (2021) found that filmed virtual reality led to greater feelings of awe than watching the same content on a flat-screen. Moreover, Kahn and Cargile (2021) found that this effect was fully mediated by an increased sense of presence when watching the scene in VR. Presence refers to the participant\u0026rsquo;s illusionary sense of being there in the virtual environment (Slater, 2018; Slater et al., 1994). Indeed, Jicol et al. (2023) and Ba\u0026ntilde;os et al. (2008) found similar positive relationships between emotional intensity and presence in animated VR. A benefit of this relationship is that, in a similar way to emotional intensity (Dawtry et al., 2020), greater presence has been linked to behavioural responses we would expect to see in real life. For example, a replication of Milgram\u0026rsquo;s infamous obedience shock experiment, but this time conducted in immersive VR, found the same pattern of results as the original study (Slater et al., 2006). Van Gelder et al. (2017) also found that in a VR burglary simulation, users reported high levels of presence, and factors which would be expected to be related to stealing behaviour in real life, like the level of deterrence, consciousness and sensation seeking, were related to stealing behaviours in VR too. Thus, feelings of presence in VR, as well as its greater emotional intensity, appears to be beneficial for eliciting behaviours we would expect to see in real life. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTherefore, VR could be a highly suitable method for studying the effect of the NBJW on bystander intervention due to its high emotional impact and presence, while maintaining a controlled and safe setting for participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCurrent Research\u003c/p\u003e\n\u003cp\u003eWe tested the model outlined in Fig. 1. Within this model, a sexual assault in the presence of a bystander causes a threat to the BJW, which in turn results in the use of non-rational strategies of derogation, blame and/or psychological distancing from the victim. In turn, these strategies may result in a reduction in empathy or empathic concern. Meanwhile, the sexual assault means a bystander should be engaged in the five stages outlined by Latan\u0026eacute; and Darley (1970). The use of non-rational strategies and corresponding reduction in empathy and empathic concern may in turn result in the individual being less likely to believe that intervention is needed and a reduced sense of responsibility to act, in turn, reducing their likelihood of intervening. As outlined above, research provides some evidence linking the aspects of this model together. However, no research to date has explored it as a whole, and through the lenses of both Latan\u0026eacute; and Darley\u0026rsquo;s (1970) bystander model and multiple BJW non-rational strategies from Lerner\u0026rsquo;s (1980) theory.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDue to the lack of NBJW research using VR, another aim of this research was to compare the effects of a threat to the BJW on bystander intervention intentions across two studies using different types of research stimuli, comparing a traditional text vignette in Study 1 to a filmed VR scene in Study 2. Taken together, these two studies allow us to test if the model shown in Fig. 1 holds across stimuli varying in emotional impact and presence.\u0026nbsp;\u003c/p\u003e"},{"header":"Study 1 Introduction","content":"\u003cp\u003eIn this first experiment, the severity of the sexual assault, which participants read about in a text vignette, was manipulated. This previously established manipulation was chosen as research suggests that the threat to the BJW is greater when the injustice is more severe (Dawtry et al., 2020). After reading the vignette, participants completed scales on all variables of interest discussed above.\u003c/p\u003e\n\u003ch2\u003eHypotheses\u003c/h2\u003e\n\u003cp\u003eBased on the literature, several hypotheses were formed. Firstly, we proposed that in the more severe condition, victim derogation, blame and psychological distancing would be greater, empathy and empathic concern would be lower, and there would be a reduced ability to complete each of Latan\u0026eacute; and Darley\u0026rsquo;s (1970) stages (e.g. lower acceptance of responsibility to intervene).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn-line with Latan\u0026eacute; and Darley\u0026rsquo;s (1970) bystander model, we predicted that each stage in the model would predict those following it (including intervention intentions), and that earlier stages would indirectly predict intervention intentions via other stages in the model\u0026rsquo;s order. For example, we predicted that a bystander\u0026rsquo;s ability to notice the sexual assault (stage one) would be indirectly predict intervention intentions (both at a general and behaviour-specific level) via an increased perception that the event was an emergency in which intervention was needed (stage two), there would be increased acceptance of responsibility to intervene (stage three) and increased belief in their ability to intervene (stage four).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe also predicted the mediation model outlined in Fig. 1. More specifically, we hypothesised that the severity manipulation would have a negative indirect effect on intervention intentions (both at a general and behaviour-specific level) through serial mediation via increased use of BJW strategies, reduced empathy or empathic concern, reduced success in viewing the event as an emergency in which intervention was needed (stage two) and reduced acceptance of responsibility to intervene (stage three). Each element of this hypothesis would be tested in turn. However, pathways via other bystander stages would be explored.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnd lastly, to confirm whether our two measures of intervention intentions reflected actual intervention behaviour, we predicted that intervention intentions (both at a general and behaviour-specific level) would be correlated with past intervention in real life.\u0026nbsp;\u003c/p\u003e"},{"header":"Study 1 Method ","content":"\u003cp\u003eParticipants\u003c/p\u003e\u003cp\u003eThe sample consisted of full-time UK university students aged between 18 and 24, as research suggests that this occupation and age group are the adults most vulnerable to sexual assault (ONS, 2023b). Sensitivity power analysis conducted using GPower software indicated that a sample size of 300 would provide 80% power to detect at least small-to-medium effect sizes (\u003cem\u003ed\u003c/em\u003e = .32, \u003cem\u003eα\u003c/em\u003e = .05). In total, the study was completed by 327 students. Thirty-three participants were removed from the dataset for the following reasons: four did not meet the screening criteria (i.e. age and occupation), 25 did not give consent to take part after reading the initial information sheet, and a further four participants failed to strongly agree or agree with an attention check item (“I am paying full attention to this survey”). This left 294 participants in the final sample (141 men, 145 women, 8 non-binary individuals). In line with the screening criteria, participants were aged between 18 and 24 (\u003cem\u003eM\u003c/em\u003e = 20.87, \u003cem\u003eSD\u003c/em\u003e = 1.64). 38.44% of the sample reported having previously been sexually assaulted, and 60.88% had been previously sexually harassed.\u003c/p\u003e\u003cp\u003eDesign\u003c/p\u003e\u003cp\u003eA between-participants design was used whereby the severity of the sexual assault was manipulated, as an indirect manipulation of the threat to a BJW. Participants read a text description about a woman, Emily, being sexually assaulted. In the less severe condition (\u003cem\u003en\u003c/em\u003e = 146), the man in the description grabbed the woman by the waist, whereas in the more severe condition (\u003cem\u003en\u003c/em\u003e = 148), the man grabbed the woman by the waist and then forced his hands up her dress (for the full text vignette, see the Procedure section below). As outlined in hypotheses one to three, this study examined the effect of this severity manipulation on the stages of Latané and Darley’s (1970) bystander intervention model, BJW strategies, and empathic variables. All other analysis used a correlational design.\u0026nbsp;\u003c/p\u003e\u003cp\u003eMeasures\u003c/p\u003e\u003cp\u003eDescriptive statistics for each measure are provided in Table 1, along with Cronbach’s alphas for multi-item measures.\u0026nbsp;\u003c/p\u003e\u003cp\u003eManipulation Checks.\u0026nbsp;Taken from Dawtry et al. (2018), participants were asked to rate how unfair what happened to Emily was on a scale from one (\u003cem\u003eslightly unfair\u003c/em\u003e) to seven (\u003cem\u003ea great deal unfair\u003c/em\u003e).\u003c/p\u003e\u003cp\u003eBystander Stages of Intervention\u003c/p\u003e\u003cp\u003eStages One to Four of Latané and Darley (1970).\u0026nbsp;The four precursor stages of Latané and Darley’s (1970) model were each measured with a single new item, as no suitable existing scale could be identified. \u0026nbsp;Participants were asked to “rate the extent to which you agree with the following statements” on a scale from one (\u003cem\u003estrongly disagree\u003c/em\u003e) to seven (\u003cem\u003estrongly agree\u003c/em\u003e), with a “don’t know” option. For stage one, noticing the emergency, the item read “At the party I would have noticed what Jack was doing to Emily”. For stage two, recognising the event as an emergency, the item read “Emily needed help in this situation”. For stage three, accepting responsibility to intervene, the item was “As someone at the party, I would feel personally responsible for helping Emily”. And lastly, for stage four, belief in your ability to intervene, the item was “I believe I would have the ability to successfully intervene and help Emily”.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIntervention Intentions.It was not possible in this study to measure actual bystander intervention behaviour (stage five of Latané and Darley’s (1970) model) in response to a sexual assault, and indeed is very difficult to measure in an experimental setting for ethical reasons. This study measured intervention intentions insteadin several ways. Firstly, on an identical scale to stages one to four, a single item was used to measure overall intervention intentions in general: “In this situation, I would intervene in some way to help Emily”. Secondly, participants were asked to “rate the likelihood of you carrying out the following actions in response to Jack’s actions at the party” on a scale from one (\u003cem\u003enot at all likely\u003c/em\u003e) to five (\u003cem\u003eextremely likely\u003c/em\u003e; a “don’t know” option was included, but excluded from analysis). The 12 items included “Covertly trying to keep Jack away from Emily” and “Confronting Jack after the incident”. The question phrasing and scale was taken from Banyard and Moynihan (2011), and the items were adapted from Bowes-Sperry and O'Leary-Kelly’s (2005) theoretical work. Item scores were averaged to create an overall intervention intention scale whereby higher scores indicated increased likelihood of intervening.\u0026nbsp;\u003c/p\u003e\u003cp\u003eEmpathy.\u0026nbsp;This seven-item scale was adapted from Smith and Frieze’s (2003) Rape-Victim Empathy Scale’s subscale specific to empathy during a rape. This scale was adapted by making it specific to the victim and more broadly to sexual assault instead of rape, and by removing one item which was irrelevant to this context: “I get really involved with the feelings of a rape victim in a movie”. Participants were asked to “Please rate the extent to which you agree with the following statements”. For example, the statements included “I can imagine how Emily feels during the incident” and “It’s not hard to understand the feelings of Emily”. The scales ranged from one (\u003cem\u003estrongly disagree\u003c/em\u003e) to five (\u003cem\u003estrongly agree\u003c/em\u003e), with a “don’t know” option. After reverse coding one negative item, these scores were summed and divided by seven to create an empathy score. Higher scores indicated higher levels of empathy, and more specifically perspective-taking, with the victim.\u0026nbsp;\u003c/p\u003e\u003cp\u003eEmpathic Concern.\u0026nbsp;This eight-item scale was used to measure the participant’s concern for the victim and was adapted from Davis’s (1980) Interpersonal Reactivity Index (IRI)’s Empathetic Concern Subscale. Participants were asked to indicate how well each item described them on a scale from one (\u003cem\u003edoes not describe me well)\u0026nbsp;\u003c/em\u003eto five (\u003cem\u003edescribes me very well\u003c/em\u003e) with a “don’t know” option. Items included “I have tender, concerned feelings for Emily” and “When I think of Emily being taken advantage of, I feel kind of protective towards her”. Three negative items were included, such as “Emily’s misfortunes do not disturb me a great deal”, which were reverse coded before analysis. An attention check item was also included here “I am paying full attention to this survey”, which was not included in analysis for this measure. Item scores were averaged, creating an overall empathic concern score, whereby higher scores indicated a higher level of concern for the victim.\u0026nbsp;\u003c/p\u003e\u003cp\u003eVictim Derogation.\u0026nbsp;This relative scale was adapted from Dawtry et al.’s (2018) Relative Derogation items and asked participants to rate the victim (Emily) on certain traits (carelessness, irresponsibility, \u0026amp; foolishness) in comparison to the average university student. Participants were asked to rate the victim on a scale from zero (\u003cem\u003enot at all careless / irresponsible / foolish compared to the average university student\u003c/em\u003e) to 100 (\u003cem\u003ea great deal careless / irresponsible / foolish compared to the average university student\u003c/em\u003e). A relative scale was used to measure derogation because relative (vs. absolute) scales permit more negative, counter-normative responses to victimisation like derogation (see Dawtry et al., 2018). Item scores were averaged to form an overall victim derogation score, with higher scores indicating more derogation of Emily’s character.\u0026nbsp;\u003c/p\u003e\u003cp\u003eVictim Blame.\u0026nbsp;This six-point scale measured the level of blame put on Emily for what happened at the party in the scenario and was adapted from Gerber et al.’s (2004) Questions Involving Victim Blame. Item phrasing varies but was always measured on a seven-point scale from one (\u003cem\u003enot at all\u003c/em\u003e) to seven (\u003cem\u003ecompletely\u003c/em\u003e). Items included “To what extent was Emily’s behaviour responsible for what happened at the party?” and “How much do you consider the incident at the party to be Emily’s fault?”. Item scores were averaged, with higher overall scores showing more blame placed on the victim.\u0026nbsp;\u003c/p\u003e\u003cp\u003ePsychological Distancing from the Victim.\u0026nbsp;Hafer’s (2000) Disassociation Scale was adapted to measure psychological distancing from the victim (Emily). The three items were “To what extent do you think Emily is like you?”, “In general, how similar would you say you are to Emily?” and “Overall, how much do you identify with Emily?”. Scale wording varied but all items were measured on a seven-point scale, with higher scores indicating that Emily was more similar to the participant. In order for the scale to reflect disassociation, the scores were reverse coded and then averaged to create an overall psychological distancing score, whereby higher scores indicated greater psychological distancing from the victim.\u0026nbsp;\u003c/p\u003e\u003cp\u003ePast Bystander Intervention.\u0026nbsp;As behavioural intentions do not always equate to actual behaviour (Ajzen, 1985), our questionnaire included a measure of past bystander behaviour in real life to compare intentions to.\u0026nbsp;This 12-item scale asked participants to indicate if they had ever carried out any of the bystander intervention behaviours listed. All items were specific to sexual assault and taken from Bowes-Sperry and O'Leary-Kelly (2005), mirroring the behaviour-specific intervention intention items mentioned previously. Items included “Covertly trying to keep a sexual assault perpetrator / potential perpetrator away from the sexual assault victim / potential victim” and “Confronting a sexual assault perpetrator after the assault”. When answering if they had previously carried out these behaviours, participants could select from the following options: “Yes”, “No”, “Don’t know / Can’t remember” or “Prefer not to say”. “Yes” answers selected in response to each item were summed to create an overall past bystander intervention score, with higher scores indicating that the participant has intervened in more ways in the past. \u0026nbsp;\u003c/p\u003e\u003cp\u003eMeasures of several extraneous variables were included to control for: past sexual assault victimisation, past sexual harassment victimisation and social desirability bias. Research suggests prior assault victimisation and social desirability bias as factors that could increase intervention intentions in questionnaires (Banyard et al., 2007; Labhardt et al., 2017; Mauer et al., 2022; Moschella-Smith et al., 2022; Vonderhaar, 2015). No literature could be identified suggesting the role of past sexual harassment victimisation, but given the potential role of past assault victimisation, past harassment was included to err on the side of caution.\u0026nbsp;\u003c/p\u003e\u003cp\u003ePast Sexual Assault.\u0026nbsp;Participants were asked if they had ever been sexually assaulted (following a detailed definition shown in Supplementary Materials; coded in data as 1 = assault experienced, 0 = no assault experienced; 3=prefer not to say).\u003c/p\u003e\u003cp\u003ePast Sexual Harassment.\u0026nbsp;An extensive definition of sexual harassment (see Supplementary Materials), was provided for participants, who were asked to indicate if they had ever been sexually harassed (coded in data as 1 = harassment experienced, 0 = no harassment experienced; 3= prefer not to say).\u0026nbsp;\u0026nbsp;\u003c/p\u003e\u003cp\u003eSocial Desirability Bias.\u0026nbsp;This 16-point scale was directly taken from Stöber (2001). Participants were asked to “read each statement carefully and decide if that statement describes you or not. If it describes you, check the word ‘true’; if not, check the word ‘false’”. Statements included “In traffic I am always polite and considerate of others” and “During arguments I always stay objective and matter-of-fact”. Six reverse coded questions were also included, such as “I take out my bad moods on others now and then”. After these six items were transformed into positive scores, all “True” item answers were summed to create an overall social desirability bias score. Higher scores implied a higher tendency to answer questions in ways deemed more socially desirable.\u0026nbsp;\u003c/p\u003e\u003cp\u003eProcedure\u003c/p\u003e\u003cp\u003eThe study took place during March 2022. When taking part in the study, participants were taken from the Prolific website via a hyperlink to the online questionnaire, hosted by Qualtrics (https://www.qualtrics.com/). The questionnaire could only be completed on a desktop computer or laptop, to ensure the desired format was presented.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn the online questionnaire, participants were presented with an information sheet about the study, and on the following pages were asked to give consent to take part and to confirm that they were aged between 18-24 years old and were a full-time student at a UK university. Following this screening process, basic age and gender demographics were collected. It was at this point that participants were asked if they had ever been sexually assaulted. If they answered yes, they were shown more information about the focus of the study, specifically that it related to sexual assault (previously only “distressing sexual experiences” had been mentioned) and they were asked to reconfirm if they would like to take part.\u0026nbsp;\u003c/p\u003e\u003cp\u003eFollowing this, participants were asked to read the text vignette. The less severe vignette was as follows:\u003c/p\u003e\u003cp\u003e“Imagine you are at a party at a new friend’s student house. There’s music playing and lots of people. \u0026nbsp;A while into the party, you notice a man, Jack, who keeps staring at a woman, Emily, who is standing alone across the room. Emily looks visibly uncomfortable when she notices Jack staring and quickly looks down at her phone. After a few minutes, Jack walks over to Emily, standing very closely next to her. You can overhear some of their conversation: “I like your dress” Jack says, “It’d look better on my bedroom floor”. He laughs. \u0026nbsp;“No thanks” Emily replies uneasily and looks down. “What? Can’t take a joke?” Jack says, moving even closer. She turns to walk away. Jack quickly wraps his arm around her waist, pulling her closer despite her efforts to lean away from him. \"I'm not interested\" Emily says. Ignoring her, he continues “Hey all I mean is you’ve got a really hot body”.”\u003c/p\u003e\u003cp\u003eThe more severe vignette was identical except for a final additional line “He pauses and then moves a hand up her dress and between her legs, still holding her waist so tightly that she cannot move.”\u0026nbsp;On the next page, participants were asked to describe what happened in the scenario, to check comprehension.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAfter this, participants completed the measures described above, with an emphasis included on answering honestly, as if they had been present at the party in the scenario.\u003c/p\u003e\u003cp\u003eOn completion of these measures, participants were shown a debrief form, providing information regarding the nature of the study, the researchers’ contact information, and advice to contact their university’s wellbeing team or relevant charities (listed in the form) if they required additional support. At this point participants were asked to give final consent for their data from this questionnaire to be kept and processed by the [Redacted Name of Institution].\u0026nbsp;\u003c/p\u003e"},{"header":"Study 1 Results and Discussion ","content":"\u003cp\u003eData Preparation\u0026nbsp;\u003c/p\u003e\u003cp\u003eQuality checks were completed to check if the same response on a scale was selected regardless of reverse-coded questions, and checking for gibberish answers in the open end questions. No participants were removed as a result of these checks. Comprehension checks were also conducted on the open-end question of what happened at the party, but no participants were removed as a result. No participants were removed for completing the study in under five minutes.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAll “don’t know” and “prefer not to say” response options were coded as missing data to allow the bystander stages, intervention intentions, empathy and empathic concern data to be treated as continuous, and so that past bystander intervention, sexual assault victimisation and sexual harassment victimisation items could be treated as “Yes or No” dichotomous data, all of which would then be suitable for multiple regression analysis.\u003c/p\u003e\u003cp\u003eThe Effect of the Severity Manipulation\u0026nbsp;\u003c/p\u003e\u003cp\u003eManipulation Checks. To check if the severity manipulation had successfully threatened the BJW, Welch’s t-test was used to test for a significant effect of the condition on the perceived unfairness of what happened to the victim (see Table 2). The condition had a significant medium-sized effect on perceived unfairness. Participants in the more severe condition rated what happened to the victim as significantly more unfair (\u003cem\u003eM\u003c/em\u003e = 6.91, \u003cem\u003eSD\u003c/em\u003e = 0.28) than those in the less severe condition (\u003cem\u003eM\u003c/em\u003e = 6.65, \u003cem\u003eSD\u003c/em\u003e = 0.63).\u003c/p\u003e\u003cp\u003eThe Effect of the Manipulation on BJW Strategies, Empathic Variables and the Bystander Stages. In line with our hypotheses, Welch’s t-test was used to test the effect of the severity of the scenario on the use of BJW strategies, each of the five bystander stages, and the empathic variables. As shown in Table 2, Welch’s t-tests found that the severity manipulation had no significant effect on any of the three non-rational BJW-threat-reduction strategies: victim derogation, victim blaming and psychological distancing. The severity manipulation also had no significant effect on empathy or empathic concern, thus not supporting our hypotheses.\u003c/p\u003e\u003cp\u003eHowever, the condition did have a significant effect on stages two and five of Latané and Darley’s (1970) model. Although the average level of agreement was high to begin with, participants in the more severe condition had significantly higher levels of agreement with the statement “Emily needed help in this situation” (stage two; \u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.81, \u003cem\u003eSD\u003c/em\u003e = 0.47) compared to those in the less severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.34, \u003cem\u003eSD\u003c/em\u003e = 0.94). This suggests that the severity of the situation affected whether it was recognised by bystanders as an emergency in which intervention was needed. Again, although agreement with the stage five general statement, “In this situation, I would intervene in some way to help Emily” was high overall, participants in the more severe condition gave significantly higher agreement scores (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.39, \u003cem\u003eSD\u003c/em\u003e = 0.90) than those in the less severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.13, \u003cem\u003eSD\u003c/em\u003e = 1.14). The condition also had a significant effect on intervention intentions relating to specific behaviours, whereby participants in the more severe condition stated that they were significantly more likely to intervene in a number of different ways (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e3.85, \u003cem\u003eSD\u003c/em\u003e = 0.53) than those in the less severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e3.64, \u003cem\u003eSD\u003c/em\u003e = 0.59). The effect of the condition on all other stages of the model were nonsignificant, therefore providing mixed support for the predicted effect on the bystander decision-making process. Moreover, we had hypothesised that agreement that intervention was needed (stage two) and intervention intentions would be lower in the more severe condition due to the negative mediating influence of BJW non-rational strategies (e.g. Li et al., 2018) but in fact found the effect to be in the opposite direction.\u0026nbsp;\u003c/p\u003e\u003cp\u003eExploring the Predictive Model\u003c/p\u003e\u003cp\u003eTo explore the proposed relationships outlined in the hypotheses, hierarchical multiple regression was conducted to test pathways between variables, allowing us to examine the predictive abilities of variables without taking account of the influence of potential mediators, thus measuring the total effects. Mediation analysis was then conducted using Models 4 and 6 of PROCESS (Hayes, 2022; 1,000 bootstrapped samples; percentile bootstrapping used) to test the significance of indirect relationships. The analysis in this section adopted the more cautious approach of Rohrer (2019), only performing mediation analysis where there was a significant total path from the predictor variable to dependent variable, to minimise the risk of spurious correlations. In the case of serial mediation, Rohrer’s (2019) approach was inferred to mean that there should be significant total effects of earlier mediators on later mediators and dependent variables, which is the approach taken here. Thus, mediation was only deemed appropriate to explore where all pathways were significant. To reduce this risk further, the potential extraneous variables of past experiences of sexual assault and sexual harassment, and social desirability bias, were entered into regression and mediation analysis as covariates to control for their influence, along with all other variables of interest (excluding potential mediators, e.g. excluding stages two to four from the covariates when examining how stage one predicts stage five).\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe results of our hierarchical regression analysis are shown in Table 3 and our mediation pathways are shown in Table 4.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe Bystander Model. Firstly, we examined our prediction that each bystander stage would predict the stages that followed it in the sequence outlined in Latané and Darley’s (1970) model, and could even mediate the relationship between earlier and later stages. In line with this, noticing the event (stage one), positively, significantly predicted recognising the event as an emergency where intervention was needed (stage two). This meant that those who gave higher ratings of agreement that they would have noticed the sexual assault also had higher levels of agreement that intervention was needed to help the victim. Stage three, accepting responsibility to intervene, was in turn significantly predicted by the previous two stages. Higher levels of agreement that the participant would have noticed the event and that intervention was needed both predicted higher levels of agreement that they were personally responsible to intervene. Stage four, belief in one’s ability to intervene, was predicted by stages one and three, whereby those who gave higher scores to agreeing that they would notice the assault and that they were personally responsible to intervene both predicted stronger belief in their ability to intervene successfully. However, stage two, recognising the event as an emergency, did not significantly predict stage four.\u003c/p\u003e\u003cp\u003eAs discussed in the methods section, stage five, intervention intentions, was measured in two ways: firstly, as a measure of general intentions, and secondly as intentions to perform specific bystander intervention behaviours. When looking at the general measure, this was positively predicted by all previous stages. This meant that those with higher levels of agreement that they would have noticed the event, that intervention was needed to help the victim, that they were personally responsible for helping, and higher levels of belief in their ability to intervene, had higher levels of intervention intentions. However, for the behaviour-specific stage five measure, this was only significantly predicted by stages three and four, whereby those with higher levels of responsibility acceptance and belief in their ability to successfully intervene were more likely to intend to intervene in a number of ways. However, stages one and two did not significantly predict the stage five behaviour-specific measure.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn summary, looking at these total effects in isolation provides mixed support for our hypotheses, largely supporting the findings of Nickerson (2014) and Burn (2009). Moreover, because of the significant direct pathways identified, this met the criteria for mediation analysis to explore whether the relationship between earlier stages in the model and intervention intentions is mediated by later stages in the model.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAs shown in Table 4, stage one only had a significant indirect pathway on the general measure of stage five via either stage three, stage four, or stages three and four in sequence. Moreover, the remaining direct pathway indicated that stage one’s ability to predict bystander interventions in general was fully mediated by stages three and four. There were also several cases of partial mediation, whereby the mediation pathway was significant, but could not fully explain the relationship between the earlier bystander stage and intervention intentions (see Table 4). \u0026nbsp;\u003c/p\u003e\u003cp\u003eTherefore, through examining both direct and indirect pathways, we can see that the data provides mixed support for our predictions. All five stages predicted general bystander intentions. However, this was not the case when looking at the behaviour-specific measure of intentions. Moreover, there were mixed results for each stage predicting and mediating each other in the sequence outlined in Latané and Darley’s (1970) bystander intervention model.\u003c/p\u003e\u003cp\u003eThe Role of the Severity Manipulation in our Predictive Model. Our hypotheses included the effect of the severity manipulation on intervention intentions being mediated by other variables in our model, specifically, BJW non-rational strategies, empathic variables and the bystander stages. However, due to the lack of significant effects of the severity manipulation, it was not possible to test for indirect effects in all but one case: the indirect effect of the severity manipulation on general intervention intentions via stage two. However, this pathway was not significant (see Table 4). This suggests that our earlier point that the lessened ambiguity in the more severe condition, and therefore greater recognition of the event as an emergency (stage two), could explain greater intervention intentions is indeed not the case. Thus, the evidence here would suggest that our hypothesis was not supported, that none of these variables mediate or explain the effect of the severity manipulation and intervention intentions.\u003c/p\u003e\u003cp\u003eThe Predictive Role of BJW Non-Rational Strategies. In order to assess the predictive abilities of the BJW non-rational strategies, we explored the relationship these strategies had with the bystander stages, which in turn we know significantly predict general intervention intentions.\u003c/p\u003e\u003cp\u003eAs shown in Table 3, victim derogation did not significantly predict any of the bystander stages. Victim blaming however can be seen to predict stage three negatively and significantly, whereby those who blamed Emily (the victim) more for her actions in the scenario were less likely to accept responsibility to intervene. Victim blaming did not significantly predict any other stages. Psychological distancing, however, negatively predicted stages one, three, four and the general measure of stage five. This means that those who psychologically distanced themselves from Emily (the victim) were less likely to think they would notice the event, accept responsibility to intervene, think they would be able to intervene successfully and intend to intervene in some way. However, it is worth noting that this significant relationship between distancing and intervention intentions was not found when talking about the specific intervention strategies measure. Thus, there is very mixed evidence in relation to our hypothesis that BJW non-rational strategies would predict stages of the bystander theory. More specifically, the BJW strategies did not predict ability to recognise the event as an emergency as hypothesised, but victim blaming and psychological distancing did negatively predict sense of responsibility to intervene, as we had predicted. \u0026nbsp;\u003c/p\u003e\u003cp\u003eDue to psychological distancing significantly predicting multiple stages, there were grounds to explore indirect pathways. As shown in Table 4, psychological distancing had several significant indirect pathways to general intervention intentions, almost all of which involved mediation via stage three, which collectively fully mediated this relationship. This provides some support for our hypotheses, although the involvement of stages one and four were not predicted in this hypothesis, and that these patterns did not occur for the behaviour-specific measure of intervention intentions.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn summary, results on victim derogation do not support our hypotheses related to the role of BJW non-rational strategies, and victim blaming providing limited support, while the patterns surrounding psychological distancing provide some supporting evidence.\u0026nbsp;\u003c/p\u003e\u003cp\u003eSeveral key mediators of psychological distancing predicting intervention intentions became evident: noticing the event (stage one), accepting responsibility to act (stage three) and belief in one’s ability to act (stage four; in that order). Although going against our original predictions that stages two and three would be most crucial, our findings still highlight the importance of stage three, and substantiate past research already linking these variables together (Burn, 2009; Katz et al., 2015; Nickerson, 2014; Wang, 2020).\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe Mediating Role of Empathy and Empathic Concern. This hypothesis predicted that the relationship between BJW non-rational strategies and the bystander stages could be mediated by empathy or empathic concern. As shown in Table 3, psychological distancing negatively predicted empathy, whereby those distancing themselves from the victim to a greater extent showed less empathy towards the victim. However, in turn, empathy did not significantly predict any of the bystander stages, suggesting that it is not a mediator here.\u003c/p\u003e\u003cp\u003eWhen empathic concern was examined, the analysis showed that it was negatively predicted by both victim blaming and psychological distancing. This means that those who blamed the victim more, and psychologically distanced themselves more so from the victim, displayed lower levels of empathic concern for the victim. And unlike empathy, empathic concern in turn predicted all the bystander stages (including both the general and specific behaviour measures of intervention intentions). This means that those showing greater empathic concern for the victim were more likely to agree that they would notice the event, that intervention was needed, that they were personally responsible for intervening, and that they would be able to successfully intervene, as well as being more likely to intervene in general and using specific behavioural strategies. This would suggest that empathic concern could mediate the relationship between BJW non-rational strategies and the bystander stages.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAs victim blaming did not predict intervention intentions, it was only possible to examine whether the pathway from victim blaming to stage three was mediated by empathic concern. We found this relationship to be fully mediated by empathic concern, thus providing evidence in support of our hypothesis that a BJW strategy would predict a bystander stage via empathic concern.\u0026nbsp;\u003c/p\u003e\u003cp\u003eOf the indirect pathways from psychological distancing, via empathic concern to the bystander stages prior to intervention, several, shown in Table 4, predicting stages one, three and four were found to be significant. Moreover, empathic concern could fully explain psychological distancing’s ability to predict stages one and four. In addition, empathic concern, followed by stages one, three and four of the bystander model (or combinations of these stages) significantly fully mediated the relationship between psychological distancing and general intervention intentions (see Table 4 for each combination of mediators). However, it is worth highlighting that stage three was the only bystander stage that was capable of significantly mediating this process without any other bystander stages in the mediation chain, suggesting perhaps that it is the most crucial bystander stage.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn summary, although not providing full support for the hypothesised mediating role of empathy and empathic concern, the data supports the importance of empathic concern in line with the findings of Katz et al. (2015), Nickerson (2014), and Patil et al. (2018). They suggest that it is the compassion element specific to empathic concern that is key to the bystander process, and not the perspective-taking element of empathy. This supports Jordan et al.’s (2016) argument that empathy and empathic concern are two distinct concepts.\u003c/p\u003e\u003cp\u003eThe Relationship between Bystander Intentions and Behaviour\u003c/p\u003e\u003cp\u003eIn line with our hypothesis, past bystander intervention was positively correlated with both the general stage five measure of intervention intentions, and the behaviour-specific stage five measure (see Table 5). This means that those who answered that they had stronger intentions to intervene to help the victim (Emily) in the scenario reported having previously engaged in more real-life bystander intervention methods related to sexual assault. This casts some validity onto our novel measures, meaning we can have some level of confidence regarding predictions of intervention intentions in this study, although more research would be beneficial to support this.\u003c/p\u003e\u003cp\u003eExploratory Analysis\u003c/p\u003e\u003cp\u003eOur hierarchical regression analysis discussed above was analysed from the perspective of our theoretical model shown in Fig. 1, whereby variables were entered in order of their position in the sequence so that slopes and p-values would not take account of the shared variance with later potential mediators in the model. However, due to the theoretical nature of the model, as well as a lack of evidence directly testing the sequential order of the stages of Latané and Darley’s (1970) bystander model, and indeed evidence of later stages affecting earlier ones (cycling, Latané and Darley, 1970), we also conducted exploratory regression analysis of all variables of interest predicting intervention intentions, free from our theoretical approach (see Table 6). With all variables entered simultaneously, we were also able to directly compare semi-partial correlations squared in order to compare how much variance in intervention intentions each variable could account for. This analysis was not pre-registered and was included upon reflection during analysis.\u0026nbsp;\u003c/p\u003e\u003cp\u003eWe found that when all variables were entered together, the only variables which significantly predicted greater general intervention intentions were, a greater belief in your ability to intervene (stage four), greater acceptance of responsibility to intervene (stage three), greater recognition that intervention was needed (stage two) and lower empathic concern (in descending order of variance in intervention intentions explained). This means that participants who had a greater belief that they could successfully intervene, a greater belief that they were personality responsible for intervening in the assault, agreed to a greater extent that intervention was needed or showed lower emotional upset for the victim’s situation, were more likely to intervene in some way.\u0026nbsp;\u003c/p\u003e\u003cp\u003eFor behaviour-specific intervention intentions, we found that it was positively predicted by responsibility to intervene (stage three), the severity manipulation, and past intervention intentions in real life (again in descending order of variance in intervention intentions explained). This meant that participants who had a greater belief that they were responsible for intervening, who read the more severe vignette, or had intervened in more ways in real life were more likely to intervene in the party scene in a variety of ways. \u0026nbsp;\u003c/p\u003e\u003cp\u003eThis analysis again highlights the importance of accepting responsibility to intervene, being the only variable which significantly predicted both intervention intention measures, when controlling for all other variables. This supports a finding from Burn’s (2009) regression analysis that when you control for the influence of all other stages in the bystander model, accepting responsibility to intervene was the only stage that remained significant in predicting intervention in both men and women.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn summary, this study set out to explore how BJW non-rational strategies might be linked to the bystander decision-making process, in the context of sexual assault. Our analysis suggests that intervention intentions are greater when the sexual assault being witnessed is more severe, although the mechanisms by which this effect occurs are unclear. Our mediation models suggests that those who psychologically distance themselves from a victim show lower levels of empathic concern to the victim, believe they would be likely to notice the events, have a reduced sense of responsibility to intervene, have less confidence in their ability to intervene, and in turn, have lower intentions to intervene in some way.\u003c/p\u003e\u003cp\u003eHowever we found a lack of significant effects of the severity manipulation on BJW non-rational strategies, despite past research suggesting significant effects (Dawtry et al., 2020). A possible explanation for this could lie in the methodology, namely the use of a text vignette. Although this method helped us to perform an initial exploration of the relationship between these variables, it lacks emotional impact compared to other stimuli formats, such as filmed VR (Dawtry et al., 2020; Patil et al., 2014).\u003c/p\u003e"},{"header":"Study 2 Introduction","content":"\u003cp\u003eStudy 2 aimed to replicate Study 1, using VR stimuli to explore whether the use of BJW non-rational strategies which may not have occurred when using the relatively emotionally-unstimulating text vignette in Study 1 (Dawtry et al., 2020; Patil et al., 2014).\u003c/p\u003e\n\u003cp\u003eWe also added several new measures. Firstly, we added a measure of the participant’s emotional response to the stimuli to capture the emotional impact Dawtry et al. (2020) discussed. This measure included three dimensions: emotional arousal, which is the intensity of the emotional response (from calm to excited), emotional valence, meaning the pleasantness of the type of emotions experienced (from unpleasant, e.g. angry, to pleasant, e.g. happy), and emotional dominance, which encapsulates the level of control you feel you have over your surroundings (from submissive to in control, Bradley \u0026amp; Lang, 1944; Russell \u0026amp; Mehrabian, 1977; Sutton et al. 2019). Secondly, we added measures of physiological arousal, specifically HR, heart rate variability (HRV) and electrodermal activity (EDA) as indirect indicators of emotional arousal (Fernández et al., 2012; Greco et al., 2016; Kreibig et al., 2007; Navarrete et al., 2012; Schaaff \u0026amp; Adam, 2013 ), allowing us to compare the emotional arousal in several stages of the study. We also added a measure of presence in the VR environment to explore how this may be related to the BJW and bystander variables, as previously unconnected, but theoretically relevant, constructs. And lastly, we added a measure of subtle victim blame, to use in addition to the existing victim blame measure, but this one designed to capture more indirect methods of blame that may have been missed in Study 1 (Hafer et al., 2019). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHypotheses\u003c/p\u003e\n\u003cp\u003eAs well as the hypotheses from Study 1, the addition of the new measures mentioned above resulted in several additional hypotheses. Firstly, we predicted that the emotional response (arousal, negative valence and dominance) and the increase in physiological arousal compared to baseline (HR, HRV and EDA) would be higher in the more severe condition. This hypothesis was based on Dawtry et al.’s (2020) suggestion that the emotional impact of stimuli is greater with more severe stimuli.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe also predicted a within-participants effect of the measurement period on physiological arousal (HR, HRV and EDA) would be highest when participants are watching the staged sexual assault video in VR, followed by when they were watching a sight test in VR, and lowest during the baseline, following the pattern of the expected emotional impact.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAlso based on Dawtry et al.’s (2020) research, we predicted that the three dimensions of emotional response to the stimuli, especially emotional arousal would be positively related to the use of BJW non-rational strategies. We examined correlations between the emotional response and presence with all other variables, in an exploratory manner (see Supplementary Materials).\u003c/p\u003e\n\u003cp\u003eSubtle victim blame predictions followed the same hypotheses as the victim blame measure used in both studies (see Hypothesis section for Study 1).\u0026nbsp;\u003c/p\u003e\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n"},{"header":"Study 2 Method","content":"\u003cp\u003eParticipants\u003c/p\u003e\u003cp\u003eFor this study, we aimed for a sample of 120 full-time UK university students aged between 18 and 24, as a feasible compromise between trying to maximise statistical power and being realistic with recruitment for in-person VR studies. Sensitivity power analyses conducted in G*Power indicated that our sample size provided 80% power to detect at least medium effect sizes (\u003cem\u003ed\u003c/em\u003e = .52,\u003cem\u003e\u0026nbsp;α\u003c/em\u003e = .05) in our t-test analysis. Due to the potentially higher emotional impact of Study 2, we excluded people who had experienced sexual assault, and people who had experienced six or more symptoms of PTSD in the past week related to past sexual harassment. In line with the [Redacted Name of Institution]’s Standard Operating Procedure for VR headsets, we also excluded individuals with health conditions which would prevent them from using VR, such as those with epilepsy or a heart condition. We also advised potential participants that they should not take part if tired, in need of sleep, under the influence of drugs or alcohol, hungover or stressed.\u0026nbsp;\u003c/p\u003e\u003cp\u003eParticipants were recruited in two ways. Eighty-one were recruited through the research participation scheme for undergraduate psychology students at the [Redacted Name of Institution] and received course credits as an incentive. Of these 81, 13 had also participated in another VR sexual assault study although the focus of these studies differed enough that we do not believe this would have affected responses. Forty-one were recruited through either the research community panel or flyers (displayed around campus, on university websites and on social media), and were given £10 for completing both the eligibility survey and the 45-minute in-person session,.\u0026nbsp;\u003c/p\u003e\u003cp\u003eFrom our recruited sample of 122, the data from five participants were removed from the dataset for various reasons: one participant completed the 45-minute study in less than 15 minutes (a cut off we deemed ahead of data collection to be too short to give meaningful answers), two participants answered “not sure” to the closed-end comprehension question, one participant failed to accurately describe the VR scene in the open-end comprehension check, and one participant experienced serious technical issues during the VR scene (which prevented them from seeing the stimulus as intended). This left us with a final sample size of 117 (16 men, 96 women, 5 gender unknown), aged between 18 and 24 (\u003cem\u003eM\u003c/em\u003e = 19.38, \u003cem\u003eSD\u003c/em\u003e = 1.40; age unknown for 6 but within inclusion age range). 28.21% of the sample reported having previously been sexual harassed.\u003c/p\u003e\u003cp\u003eDesign\u003c/p\u003e\u003cp\u003eA between-participants design very similar to Study 1 was used, in which we manipulated the severity of the sexual assault in the filmed VR scene (see Fig. 2, full scenes can be viewed here: [INSERT LINK: https://youtube.com/playlist?list=PLCfqtCP4BMd2z1ljF7wAQhTXZKKtV4lss\u0026amp;feature=shared] ), as an indirect way of manipulating the threat to the BJW. In the less severe condition (\u003cem\u003en\u0026nbsp;\u003c/em\u003e= 50), as well as sexually harassing comments and stares, the man in the video also grabbed the woman’s waist. In the more severe condition (\u003cem\u003en\u0026nbsp;\u003c/em\u003e= 67), he also stroked her upper thigh.\u0026nbsp;\u003c/p\u003e\u003cp\u003eIn addition, within-participant analysis was used to explore the effect of the task (baseline, sight test in VR, staged sexual assault in VR) on physiological arousal.\u003c/p\u003e\u003cp\u003eMeasures\u003c/p\u003e\u003cp\u003eAll measures from Study 1 were used in this study, as well as several additional measures outlined here and included in Table 1.\u003c/p\u003e\u003cp\u003eTrauma Related to Past Sexual Harassment.\u0026nbsp;This measure was adapted from the Trauma Screening Questionnaire (Brewin et al., 2002). Participants were asked “Please consider the following reactions which sometimes occur after a traumatic event. This question is concerned with your personal reactions to previous experiences of sexual harassment. Please indicate (Yes/No) whether or not you have experienced any of the following at least twice in the past week, specifically in relation to past experiences of sexual harassment”. Items included “Upsetting dreams about the event” and “Being jumpy or being startled at something”. Participants could respond to the 10 items with “yes”, “no”, or “prefer not to say”. Participants were deemed ineligible to take part in the study if “Yes” or “Prefer not to say” were selected for six or more items.\u003c/p\u003e\u003cp\u003eSubtle Victim Blame.\u0026nbsp;Hafer et al.’s (2019) subtle victim blame measure was adapted for this study. This four-item measure is designed to capture a less explicit form of victim blaming, one which is less susceptible to the influence of social norms. More specifically, participants were asked “How likely do you think it is that the incident at the party happened because…” followed by two factors within the victim’s control, and two factors outside of the victim’s control. For this study, the two factors within the victim’s control were that the victim was standing on her own and her choice of clothing. The two factors outside of the victim’s control were the perpetrator’s intoxication, and the victim being in “the wrong place at the wrong time”. Participants were asked to respond on a seven-point scale from one (not at all likely) to seven (very likely). The scores for the two questions with factors within the victim’s control were averaged to create a subtle victim blame measure, with higher scores indicating greater blame on the victim.\u003c/p\u003e\u003cp\u003ePresence. This was adapted from Slater and Steed’s (1994) measure of presence in VR. Participants were asked three questions: “Please rate your sense of being there in the party”, “To what extent were there times during the experience when the party became the \"reality\" for you, and you almost forgot about the \"real world\" outside?” and “When you think back about your experience, do you think of the party more as something that you saw, or more as somewhere that you visited?”. Scale wording varied but all items were measured on a seven-point scale, with higher scores indicating a greater sense of presence in the VR setting. The scores for each question were averaged to create an overall presence score.\u0026nbsp;\u003c/p\u003e\u003cp\u003eEmotional Responses to VR Scene. Bradley and Lang’s (1994) Self-Assessment Manikin (SAM) was used to measure participants’ emotional reactions to the VR scene, more specifically capturing three variables: the emotional valence, arousal and dominance. Participants were asked to “Please rate how you felt while you were watching the scene at the party”, and responded on three 9-point scales. The scales were visual, with images of the SAM, showing varying emotions (see Fig. 3). These SAM images came with descriptive cues at the extremes of the scales, for example, for emotional arousal, the lower end of the scale had cues for “Stimulated, excited, frenzied, jittery, wide awake, aroused” and the other end had cues for “Relaxed, calm, sluggish, dull, sleepy, unaroused”. A higher score on the emotional valence scale indicated greater negative emotions. A higher score on the arousal SAM scale indicated lower emotional arousal, and was therefore be reverse coded so that a greater score in our analysis indicates greater arousal. A higher score on the emotional dominance scale indicates a greater feeling of control. \u0026nbsp;\u003c/p\u003e\u003cp\u003ePhysiological Arousal. HR (interbeat-interval in raw form; IBI), HRV (IBI in raw form), and EDA were measured using the Empatica E4 wristband. We recorded during a seated 5-minute baseline, when watching the sight test in VR, and when watching the staged sexual assault condition video in VR. Mean HR, HRV and EDA for each task were calculated from individual participant files using R. To compare the effect of condition and correlations, the baseline measurement was subtracted from the measurement when watching the condition stimuli video in VR, to create an increase score for each measure to account for individual baseline differences.\u0026nbsp;\u003c/p\u003e\u003cp\u003eProcedure\u003c/p\u003e\u003cp\u003eData collection took place between March and December 2023. Participants completed an eligibility survey (hosted by Qualtrics) before they were able to sign up to take part in the laboratory study. In this survey, participants were given an information sheet, followed by questions on age, gender, student status, past sexual assault and harassment experience, trauma related to past sexual harassment (if selecting that they had experienced it or preferred not to say) and relevant health conditions.\u003c/p\u003e\u003cp\u003eIn the in-person session, participants were given a detailed verbal walkthrough of what they would be doing in the session and were given the opportunity to ask questions. They then read an information sheet in an online questionnaire and confirmed their consent to take part.\u0026nbsp;\u003c/p\u003e\u003cp\u003eParticipants then had an Empatica E4 wristband attached to them and a five-minute seated baseline physiological recording was collected, and this wristband remained on for both VR videos.\u003c/p\u003e\u003cp\u003e\u0026nbsp;Participants were given instructions for the VR scenes, namely that they would be seeing two videos (a sight test and the main video), in both they would be able to look in any direction and rotate on the spot, but would not be able to move around the room or interact with the environment. As part of this, they were also reminded of their right to withdraw if feeling cybersickness in VR or feeling upset by the video content. Participants then had the Oculus Quest 2 head-mounted display fitted and saw a sight test in VR to confirm that they could see details and to get participants used to the VR setting. This sight test depicted a 360-degree photo of a room containing a Snellen chart on the wall in front of them (see Fig. 4; an optical sight chart traditionally used by opticians). Participants were asked to read the top seven lines of random letters on the chart. If unable to do so, the VR headset was adjusted to improve the focus, and if still unable to read the lines, the participant would have been informed that they were not able to take part in the study (although this last step did not happen with any of the participants). Following the sight test, participants were then instructed that they would be watching a two-minute video next and that again they would be able to look around, but could not interact with the environment. Participants then watched the stimuli video from the condition they were assigned to (see Fig. 2), which followed the same story as in Study 1 (except for the more severe condition’s exact form of sexual assault, see design section for details).\u003c/p\u003e\u003cp\u003eFollowing this, the VR headset and wristband were removed. Participants then completed the main questionnaire. As well as the comprehension question used in Study 1, an additional question was added whereby they were asked to confirm if they saw the man in the white t-shirt (perpetrator) and the woman in the red dress (victim). They then completed the measures described above.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAfter completing the questionnaire, participants were given a verbal debrief, explaining the rationale behind the study, it’s equipment and the conditions, and were actively encouraged to ask questions and talk about how they were feeling after watching the stimuli. They were given a debrief form to take away with them, following the same structure as the form in Study 1, and the contact information of the research team and relevant organisations offering support were verbally highlighted. Participants then returned to the VR headset and watched a filmed VR video of the actors from the stimuli VR scene explaining that they are actors and are safe and well. Following this, participants were given the option of watching one of three happy VR videos to ensure that they left the study on a positive note and did not leave with any negative associations with VR due to the study’s subject matter. These videos were obtained from the DeoVR app, and each showed a 180-degree filmed VR video of a relaxing sunset over a river and trees, cherry blossom trees in bloom, and puppies.\u003c/p\u003e"},{"header":"Study 2 Results and Discussion","content":"\u003cp\u003eData Preparation\u0026nbsp;\u003c/p\u003e\u003cp\u003eQuality checks identical to Study 1 were used, and again, did not result in any participants’ data being removed. We were unable to match the screener data to the rest of the data sources for six participants, meaning there is missing data on screening measures including age, gender, and sexual harassment, for those people. As with Study 1, there were several measures which included a “don’t know” options, and these response options were coded as missing data so that the remaining scale could be treated as continuous or dichotomous.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThere were errors with incorrect or missing tagging when using the Empatica E4 wristband for 13 participants, so their physiological data was not included in analysis. There were also issues with substantial noise in the IBI raw data for an additional 28 participants, meaning that their HR and HRV data was unusable (while their EDA data was fine). This left a sample 104 participants for the EDA analysis, and 76 participants for the HR and HRV analysis.\u003c/p\u003e\u003cp\u003eThe Effect of the Severity Manipulation\u0026nbsp;\u003c/p\u003e\u003cp\u003eManipulation Checks.\u0026nbsp;We again used Welch’s t-test to see if the condition had a significant effect on our manipulation check variable (see Table 2). However, in this study, there was no significant effect on perceived unfairness. It is worth noting that the effect size was comparable to that for Study 1 and did not statistically significantly differ (see confidence intervals in Table 2).\u003c/p\u003e\u003cp\u003eThe Effect of the Manipulation on BJW Strategies, Empathic Variables and the Bystander Stages.\u0026nbsp;In our hypotheses, we predicted that in the more severe condition, if BJW non-rational strategies were used, empathy and empathic concern would be lower, and there would be a reduced ability to complete each of the bystander stages (e.g. lower intention to intervene). As shown in Table 2, Welch’s t-test found significant effects of the severity manipulation on both the general and behaviour-specific measures of intervention intentions. However, we found the opposite to what we had predicted: instead of intervention intentions being higher in the less severe condition, general intervention intentions were significantly greater in the more severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.62, \u003cem\u003eSD =\u003c/em\u003e 0.67), than in the less severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e6.14, \u003cem\u003eSD =\u003c/em\u003e 1.08). Similarly, intentions to intervene in specific ways was higher in the more severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e3.71, \u003cem\u003eSD =\u003c/em\u003e 0.55), than in the less severe condition (\u003cem\u003eM =\u0026nbsp;\u003c/em\u003e3.50, \u003cem\u003eSD =\u003c/em\u003e 0.51). Contrary to our hypotheses, no other effects of the severity manipulation were significant.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe Effect of the Manipulation on Emotional and Physiological Responses.\u0026nbsp;In Study 2, we also added the hypothesis that the emotional response to the stimuli (arousal, negative valence, and dominance) would be greater in the more severe condition. However, as shown in Table 2, there was no significant effect of the manipulation on any of the emotion response variables. Similarly, we tested whether the increase in physiological arousal, measured as the difference in HR, HRV and EDA from baseline to the stimuli watching period, was greater in the more severe condition than the less severe one. But we found no significant difference in this across conditions.\u003c/p\u003e\u003cp\u003eExploring the Predictive Model\u003c/p\u003e\u003cp\u003eAn analysis plan identical to Study 1 was used for Study 2, whereby to examine the hypothesised relationships, we used hierarchical multiple regression, conducted in the sequence of steps in our model (Fig. 1), and where appropriate, then conducted mediation analysis. Although we had planned to include it, subtle victim blame was not included in regression and mediation analysis due to its low Cronbach’s alpha (see Table 1) and on reflection to allow a fairer comparison of other BJW regression pathways with those from Study 1 in which this variable was not recorded.\u003c/p\u003e\u003cp\u003eThe results of our hierarchical regression analysis are shown in Table 3 and all mediation pathways we tested are in Table 4.\u003c/p\u003e\u003cp\u003eThe Bystander Model.\u0026nbsp;As with Study 1, we tested our prediction that each of the bystander stages would significantly predict all the stages following it in Latané and Darley’s (1970) model, and most importantly, would predict stage five, intervention intentions. In Study 2 we found that stage one, noticing the event, did not predict stage two, deciding that intervention is needed. Noticing the event and deciding that intervention was needed did both positively predict stage three, accepting responsibility to intervene. Those who gave higher agreement scores that they would notice the sexual assault and that help was needed had a greater sense of responsibility for intervening as someone attending the party. We also found that both noticing the event and accepting responsibility to intervene positively predicted stage four, belief in your ability to intervene. Greater agreement that the participant would have noticed the assault and that they were responsible for intervening was linked to a greater belief in their ability to intervene. But, agreement that intervention was needed did not predict belief in your ability to intervene.\u0026nbsp;\u003c/p\u003e\u003cp\u003eWhen looking at general intervention intentions, we found that this was positively predicted by stages one, three and four of the model. Participants with higher agreement that they would notice the event, that they would be responsible to intervene, and that they believed they could successfully intervene answered that they would be more likely to intervene in some way. However, we found that the second stage, deciding that intervention is needed, did not predict general intervention intentions. For the behaviour-specific intervention intentions measure, we found that this was positively predicted by stages one, two, and three, but not stage four. Those who agreed to a greater degree that they would notice the event, that intervention was needed, and that they were responsible for intervening, agreed to a greater extent that they would intervene in a number of suggested ways. Thus, direct regression pathways provide mixed results for our hypothesis that each stage in Latané and Darley’s (1970) model would predict those following it in sequence. Nonetheless, our findings that several of the bystander stages predicted those following it, and that all stages in the model predicted one or more measures of intervention intentions, supports the works of Burn (2009) and Nickerson (2014).\u003c/p\u003e\u003cp\u003eExamining all possible indirect pathways, we found several significant indirect pathways in which stage three, accepting responsibility to intervene, mediated it. Firstly, stage three fully mediated the pathway from stage one to the general intervention intentions, Secondly, the pathway from stage two to the behaviour-specific measure of bystander interventions via stage three was also fully mediated. And lastly, stage three mediated \u0026nbsp;the pathway from stage one to behaviour specific intervention intentions, but could not fully explain this relationship. Paired with the direct pathways, these indirect pathways provide only partial support for our hypothesis, but highlight the importance of stage three, accepting responsibility to intervene.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe Role of the Severity Manipulation in our Predictive Model.\u0026nbsp;Due to the severity manipulation only having a significant effect on intervention intentions, we were not able to test our hypotheses that this effect would occur via BJW non-rational strategies, empathic variables or the bystander stages (specifically stages two and three). The lack of significant effects does not provide evidence for our hypotheses.\u003c/p\u003e\u003cp\u003eThe Predictive Role of BJW Non-Rational Strategies.\u0026nbsp;Next examining if the BJW non-rational strategies negatively predicted the stages of Latané and Darley’s (1970) model, we found limited support for this hypothesis. We found that victim derogation positively predicted greater intervention intentions. Those who blamed the woman in the red dress (victim)’s character more for the assault were more likely to intend to intervene in some way. This is in direct contrast to past research which had found a negative relationship between BJW non-rational strategies and bystander intervention (Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022), and the lack of a significant prediction from derogation to intervention intentions in Study 1. However, this pattern in Study 2 was not found in our simultaneous regression or correlational analysis (see Tables 5 \u0026amp; 6), suggesting this pattern could be due to controlling for the variance shared with certain variables within our data in the hierarchical sequence. An alternative suggestion could be that critiquing the victim’s character meant that the bystander viewed her as incompetent and therefore not capable of looking after herself, perhaps as a form of benevolent sexism (Glick \u0026amp; Fiske, 1996), although further research would be needed to confirm if this was the mechanism.\u0026nbsp;\u003c/p\u003e\u003cp\u003eAgain in contrast to our hypothesis, psychological distancing positively predicted greater agreement that they would notice the event. Those who said they were more different from the woman in the red dress believed they were more likely to notice the harassment and assault taking place. Victim blaming however, negatively predicted the participant’s belief that they would notice the event and their general intervention intentions. Those who blamed the victim’s actions more believed that they would be less likely to have noticed the assault and less likely to intervene in some way.\u0026nbsp;\u003c/p\u003e\u003cp\u003eMediation analysis was run to explore the only possible mediation path from BJW strategies to intervention intentions via the bystander stages, namely whether victim blaming predicted general intervention intentions via reducing noticing ability, but this was not significant (see Table 4). Overall, Study 2 provides no evidence to support derogation and psychological distancing negatively predicting the stages of Latané and Darley’s (1970) model either directly or indirectly, but evidence of victim blaming doing so to a degree directly. Although this last finding provides evidence in line with past research\u0026nbsp;(Boppana, 2021; Li et al., 2018; Spaccatini et al., 2022) linking victim blaming and intervention intentions, our predictions that this relationship would be mediated by the stages of the bystander model were not found to be the case.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThe Mediating Role of Empathy and Empathic Concern.\u0026nbsp;We also predicted that empathy and empathic concern may act as a mediator between BJW non-rational strategies and the bystander stages (specifically stages two and three), and then intervention intentions. In line with our hypothesis, we found that empathy and empathic concern were both negatively predicted by psychological distancing. This means that participants who believed they were very different from the woman in the red dress had lower levels of empathy and empathic concern for her. We also found that victim blaming negatively predicted empathic concern, meaning that those who blamed the victim more had lower levels of empathic concern for her. These patterns support our hypothesis, and past research (Katz et al., 2015), \u0026nbsp;suggesting that there is a negative relationship between some BJW non-rational strategies and empathic variables, although this was not the case with victim derogation.\u0026nbsp;\u003c/p\u003e\u003cp\u003eExamining if our empathic variables could predict the bystander stages, we found limited evidence to support this hypothesis, with most predicted pathways being nonsignificant. We did however find two significant regression pathways. Firstly, empathy positively predicted general intervention intentions. This meant that those who were better at taking the perspective of the woman in the red dress rated themselves as more likely to intervene in some way. And secondly, we found that empathic concern positively predicted stage two of the bystander model, recognising the event as an emergency whereby intervention is needed. Under this regression pathway, those who felt more empathic concern for the woman in the red dress were more likely to agree that intervention was needed. Thus, there is some evidence from this study connecting our empathic variables with the bystander decision-making process, but this is more limited than we had predicted.\u0026nbsp;\u003c/p\u003e\u003cp\u003eDue to a lack of significant pathways between the severity manipulation, BJW non-rational strategies, empathy and empathic concern, and the bystander stages, it was not possible to test our empathic variables as mediators in this process.\u003c/p\u003e\u003cp\u003eThe Relationship between Bystander Intentions and Behaviour\u003c/p\u003e\u003cp\u003eWe predicted that both our general and behaviour-specific measures of intervention intentions would be correlated with past bystander intervention against sexual assault in real life. However, we found that neither of these correlations were significant (see Table 5).\u0026nbsp;\u003c/p\u003e\u003cp\u003ePhysiological Arousal Across Tasks\u0026nbsp;\u003c/p\u003e\u003cp\u003eNew to Study 2, we predicted a within-participants effect on physiological arousal, whereby physiological arousal (HR, HRV \u0026amp; EDA) would be highest when participants were watching the staged sexual assault in VR, followed by when they are watching the VR sight test, and then during the seated baseline.\u0026nbsp;\u003c/p\u003e\u003cp\u003eThree repeated-measure ANOVAs revealed that the task had a significant effect on HR, HRV and EDA (see Table 7). Planned comparisons showed that HR during the sight test and sexual assault video were both significantly higher than during the baseline. However, there was no significant difference in HR between during the sight test and sexual assault video. Similarly, HRV was significantly higher in the sight test and sexual assault video than in the baseline. But, like HR, HRV during the sight test and sexual assault video did not significantly differ. For EDA however, we found significant differences between all three tasks. More specifically, EDA during the sight test and sexual assault video was significantly higher than during the baseline, and EDA during the sexual assault video was significantly higher than using the sight test.\u003c/p\u003e\u003cp\u003eThese findings are largely in support of our hypotheses, showing that the task influenced physiological arousal, although failing to provide support for circulatory indicators having heightened levels during the sexual assault stimuli VR video compared to the VR sight tests. This could suggest either issues with measurement quality, or that HR and HRV were more greatly impacted by standing up and being in VR, rather than by the content within VR, while EDA was able to capture the impact of the emotional nature of the content too. \u0026nbsp; \u0026nbsp;\u003c/p\u003e\u003cp\u003eExploratory Analysis\u003c/p\u003e\u003cp\u003eSimultaneous Regression.\u0026nbsp;As with Study 1, we conducted simultaneous regression analysis to compare the relative contribution of our variables of interest in accounting for intervention intentions, free from our theoretical preconceptions. This analysis was not pre-registered and was added as an exploratory addition during analysis. In Study 2, we found that greater general intervention intentions were significantly predicted by a greater sense of responsibility to intervene (stage three), greater belief in their ability to intervene (stage four), lower victim blaming, and higher social desirability bias (listed in descending order of the amount of variance in intentions they could account for). This means that participants who had a greater sense of responsibility, who had a greater belief in their ability to intervene, blamed the victim’s behaviour less or displayed a greater bias towards answering questions in a socially desirable way, were more likely to answer that they would intervene in some way. Greater behaviour-specific intervention intentions however were predicted by greater responsibility to intervene (stage three) and greater belief that they would notice the assault (stage one; in descending order of variance explained). This means participants who agreed to a stronger degree that they were personally responsible for intervening, or had a greater belief that they would notice the harassment and assault take place were more likely to think they would intervene in a variety of ways.\u0026nbsp;\u003c/p\u003e\u003cp\u003eInfluence of Sample Differences.\u0026nbsp;We also ran unplanned analysis to explore differences in our findings between Studies 1 and 2. There were several crucial differences in the samples of our two studies, namely, the gender split (49.32% female, 47.96% male, 2.72% non-binary vs. 82.05% female, 13.68% male, 4.27% unknown), percentage of the sample who had previously been sexually assaulted (38.44% vs 0%), and percentage who had previously been sexually harassed (60.88% vs 28.21%). Additional analysis was therefore performed to understand if these sample differences could account for differences in effects and relationships. Comparing confidence intervals for our t-tests and hierarchical regression results across both studies revealed that there was no overlap between confidence intervals for, and therefore key differences in, regression pathways of victim blaming predicting the ability to notice the event (stage one), psychological distancing predicting the ability to notice the event, and belief in their ability to intervene (stage four) predicting general intervention intentions (see Table 3). Moderation analysis was conducted using model 1 of PROCESS (Hayes, 2022) on data from Study 1 to examine whether gender, past sexual assault or past sexual harassment moderated these patterns.\u003c/p\u003e\u003cp\u003eFirstly, looking at victim blaming predicting their ability to notice the event, we found that there was a significant victim blaming x sexual assault history interaction, \u003cem\u003eb\u003c/em\u003e = -2.11, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.78, \u003cem\u003et\u003c/em\u003e (251) = -2.71, \u003cem\u003ep =\u003c/em\u003e .01, 95% CI [-3.64, -0.58]. Among those who had previously been sexually assaulted, victim blaming significantly, negatively predicted belief that they would notice the event \u003cem\u003eb\u003c/em\u003e = -1.91, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.76, \u003cem\u003et\u003c/em\u003e (251) = -2.50, \u003cem\u003ep =\u003c/em\u003e .01, 95% CI [-3.40, -0.41]. This meant that sexual assault survivors who blamed the victim to a greater extent were more likely to believe that they would not notice the assault take place. Whereas, among those who had not been previously sexually assaulted, victim blame did not significantly predict agreement that they would notice the event, \u003cem\u003eb\u003c/em\u003e = 0.21, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.16, \u003cem\u003et\u003c/em\u003e (251)= 1.25, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .21, 95% CI [-0.12, 0.53]. Gender and sexual harassment victimisation did not however significantly moderate the relationship between victim blaming and noticing the event, \u003cem\u003eb\u003c/em\u003e = -.20, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.42, \u003cem\u003et\u003c/em\u003e (253) = -0.48, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .63, 95% CI [-1.03, 0.63],\u003cem\u003e\u0026nbsp;\u003c/em\u003eand\u003cem\u003e\u0026nbsp;b\u003c/em\u003e = -.71, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.47, \u003cem\u003et\u003c/em\u003e (256) = -1.53, \u003cem\u003ep\u003c/em\u003e = .13, 95% CI [-1.63, 0.20]\u003cem\u003e,\u0026nbsp;\u003c/em\u003erespectively. Thus, the inclusion of sexual assault survivors in Study 1, but not in Study 2, may account for differences between victim blaming’s ability to predict noticing the event between our two studies.\u0026nbsp;\u003c/p\u003e\u003cp\u003eLooking at potential moderators of psychological distancing predicting noticing the event, we found that gender, past sexual assault and past sexual harassment all did not interact with this pathway, \u003cem\u003eb\u003c/em\u003e = 0.19, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.12, \u003cem\u003et\u003c/em\u003e (253) = 1.58, \u003cem\u003ep\u003c/em\u003e = .11, 95% CI [-0.05, 0.42]\u003cem\u003e, b\u003c/em\u003e = 0.01, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.11 \u003cem\u003et\u003c/em\u003e (251) = 0.07, \u003cem\u003ep\u003c/em\u003e = .94, 95% CI [-0.20, 0.22], and\u003cem\u003e\u0026nbsp;b\u003c/em\u003e = -.005, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.11, \u003cem\u003et\u003c/em\u003e (256) = -0.04, \u003cem\u003ep\u003c/em\u003e = .97\u003cem\u003e,\u0026nbsp;\u003c/em\u003e95% CI [-0.22, 0.22],\u003cem\u003e\u0026nbsp;\u003c/em\u003erespectively. Similarly, we found that gender, past sexual assault and past sexual harassment all did not moderate the relationship between belief in your ability to intervene and general intervention intentions, \u003cem\u003eb\u003c/em\u003e = -.03, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.07, \u003cem\u003et\u003c/em\u003e (273) = -0.37, \u003cem\u003ep\u003c/em\u003e = .71\u003cem\u003e,\u003c/em\u003e 95% CI [-0.17, 0.12],\u003cem\u003e\u0026nbsp;b\u003c/em\u003e = -.03, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.08, \u003cem\u003et\u003c/em\u003e (269) = -.31, \u003cem\u003ep\u003c/em\u003e = .76, 95% CI [-0.19, 0.14], and\u003cem\u003e\u0026nbsp;b\u003c/em\u003e = .02, \u003cem\u003eSE =\u0026nbsp;\u003c/em\u003e0.07, \u003cem\u003et\u003c/em\u003e (276) = 0.28, \u003cem\u003ep\u003c/em\u003e = .78\u003cem\u003e,\u003c/em\u003e 95% CI [-0.13, 0.17],\u003cem\u003e\u0026nbsp;\u003c/em\u003erespectively. Thus, these sample differences cannot account for differences in these pathways between Studies 1 and 2. \u003c/p\u003e"},{"header":"General Discussion","content":"\u003cp\u003eWe conducted two experiments, exploring the effect of indirectly manipulating the threat to the NBJW on BJW non-rational strategies, empathic variables and Latané\u0026nbsp;and Darley’s (1970) bystander decision-making process when witnessing a sexual assault. We expanded on our initial text vignette exploration in Study 1 by replicating the study using emotionally impactful (Patil et al., 2014), realistic and immersive filmed VR in Study 2.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eDifferences Between our Studies\u003c/h2\u003e\n\u003cp\u003eBetween the two studies, we found differences in the statistical significance of the success of the manipulation, the predictive ability of the BJW non-rational strategies and the role of empathic variables as mediators between non-rational strategies and bystander stages including intervention, albeit most of these differences between the studies were not themselves statistically significant. The most obvious explanation for these differences is the use of filmed VR instead of a text description, which research suggests would have increased the emotional impact of the stimuli and presence (Baños et al., 2008; Dawtry et al., 2020; Jicol et al., 2023; Jicol et al., 2022; Kahn \u0026amp; Cargile, 2021; Pan \u0026amp; Slater, 2011; Patil et al., 2014). However, future research is needed to confirm this influence in this context, experimentally manipulating the stimuli type using a randomised sample. There are also other explanations for the differences in our findings, which will be discussed here.\u003c/p\u003e\n\u003cp\u003eWe found differences in which BJW non-rational strategy was the most important in predicting intervention intentions. An explanation for psychological distancing being the key BJW predictor of intervention in Study 1, while victim blaming was the key predictor in Study 2, may lie in BJW literature. Dawtry et al. (2020) found that victim derogation effects were greater in response to more emotionally impactful, realistic stimuli, where the threat to the need to BJW is likely greater. This may suggest that more extreme counter-normative non-rational strategies like derogation and victim blaming are reserved for these greater belief threats. Comparing our two studies, it may have been the case that where participants were responding to a less emotionally evocative text stimuli (Dawtry et al., 2020; Patil et al., 2014), the simpler strategies of psychological distancing from the victim may have been sufficient in resolving the threat. Whereas, in Study 2, where participants were witnessing the assault in an arguably more highly emotionally arousing immersive VR context, more drastic measures, like victim blaming, may have played a crucial role in determining bystander decision-making.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt could also be argued that there was a greater opportunity for victim blaming in the second study. In the first study, there is no mention of Emily’s (the victim’s) appearance. Whereas, in the second study, there are several factors of her appearance which could be used under rape myths as reasons to blame the victim. A commonly held rape myth is the idea that a victim is ‘asking for it’ because of the way they were dressed (Burt, 1980). Under this logic, the victim in Study 2 is wearing a low cut, short dress, and due to the lack of names used in the scene, is only referred to in the questionnaire as ‘the woman in the red dress’, a colour associated with sexual receptivity (Pazda et al., 2012). In this way, it could be suggested that participants adhering to rape myths may have used these components as fuel for victim blaming, which therefore may have become a greater factor in either influencing or justifying their bystander intervention intentions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHowever, relevant to this argument, is the subtle victim blame measure in Study 2. This measure was not included in our regression analysis, partially due to it’s poor Cronbach’s alpha (see Table 1). The reason for this low alpha was that, on average, participants were willing to engage in subtle victim blaming based on the woman standing on her own (\u003cem\u003eM\u003c/em\u003e = 5.00, \u003cem\u003eSD\u0026nbsp;\u003c/em\u003e= 1.44), but reluctant to blame her based on what she was wearing (\u003cem\u003eM\u003c/em\u003e = 2.87, \u003cem\u003eSD\u003c/em\u003e = 1.48). This could possibly reflect a rejection of more widely known rape myths among younger adults (Chauhan, 2022), such as those based on clothing, or stronger social norms against this (Dawtry et al., 2018). Regardless, the low agreement with this second item would suggest that the clothing of the victim in our stimuli did not account for the more prominent role of victim blaming in predicting intervention intentions in Study 2. Moreover, as shown in Table 2, the mean victim blaming score did not drastically increase in Study 2, but instead all BJW non-rational strategy use slightly increased, regardless of their predictive strength. This suggests that the differences in psychological distancing and victim blaming predictions between the two studies cannot be explained by an increased use of victim blaming and a resultant reduction in the need for distancing.\u003c/p\u003e\n\u003cp\u003eMoving our focus onto our potential mediators, empathy and empathic concern, Study 2 found a quite stark difference in significant patterns compared to Study 1. In Study 1, we found that greater empathy was only related to reduced psychological distancing, while greater empathic concern was a more crucial variable in our model, being predicted by both lower victim blaming and lower psychological distancing, as well as predicting greater agreement with all stages in the bystander model, including both measures of intervention intentions. In Study 2, the empathic variables were both still predicted by the same BJW strategies as in Study 1. However, their predictive patterns were very different, with greater empathy predicting greater general intervention intentions, and empathic concern only predicting greater agreement that intervention was needed (stage two). Although the reason for the changes in empathic concern’s ability to predict the bystander stages remain unclear, a possible explanation for the increase in empathy’s importance may involve the use of VR in Study 2. The very nature of VR involves the use of perspective taking, putting yourself in another person’s shoes (either literally or figuratively). Due to this, VR has been referred to as the “ultimate empathy machine” (TED, 2015, 3:12), and has indeed been harnessed for this specific purpose of increasing empathy towards particular groups or individuals (De La Peña et al., 2010; Seinfeld et al., 2018; Tassinari et al., 2022; Ventura et al., 2021). And although our average level of empathy did not increase from Study 1 to Study 2 (likely due to already nearing the scale’s ceiling in Study 1), we can see an increase in the importance of empathy in our VR study, through it’s now significant prediction of general intervention intentions. Returning to our hypotheses, we can see that both our empathic variables are related to some BJW strategies and parts of the bystander model, but did not act as mediators in Study 2 in this instance.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eConsistent Patterns Across our Findings\u003c/h2\u003e\n\u003cp\u003eHaving considered the different findings between the studies, it is important to also consider the key consistent findings. Most crucially to our hypotheses, we found that manipulating the severity of the sexual assault participants witnessed (as an indirect manipulation of the threat to the BJW) had no significant effect on BJW non-rational strategies, empathy, empathic concern, and stages one, three and four of Latané\u0026nbsp;and Darley’s (1970) bystander model. After Study 2 we can rule out a lack of emotional impactful stimuli as a reason for this lack of effects (Dawtry et al., 2020; Patil et al., 2014). However, another possible explanation of the lack of an effect of the manipulation on BJW non-rational strategies relates to Lerner’s (1980) theory itself. As discussed, a bystander is unlikely to proceed with non-rational strategies to remove a threat to the BJW if it is possible to resolve the threat rationally and they believe they can fully remove the suffering. Thus, the lack of significant effects of condition on non-rational strategies could be due to participants believing they could intervene in this scenario and thus resolve the threat to their BJW rationally.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis explanation is further supported by the high means for measures of belief in their ability to successfully intervene (stage four) and general intervention intentions (stage five) in both studies. The use of rational strategies like intervention was ruled out in our predictions because research suggests that an individual would be unlikely to use these strategies if they believed that they could not fully remove the suffering (Montada \u0026amp; Lerner, 1998; Walster et al., 1973; White, 2012). As the victim had already been assaulted, it may not be possible to remove this existing suffering to their physical and mental health\u0026nbsp;(Carey et al., 2018; Luce et al., 2010; Mason \u0026amp; Lodrick, 2013; McFarlane et al., 2005). However, it could be that participants thought that the situation might escalate into a more severe assault, which would involve far greater suffering, which they could prevent by intervening. Future research could test this explanation by manipulating the personal risks of intervening, a factor Lerner (1980) theorised affects this decision to use a rational or non-rational strategy. It would be insightful to see whether a high-cost condition would result in greater derogation, blame and psychological distancing.\u003c/p\u003e\n\u003cp\u003eWhile it’s possible for this BJW-based explanation to account for the lack of significant effects, another explanation lies in bystander research. Based on our BJW theoretical approach, we proposed that witnessing a more severe sexual assault would elicit reduced intervention intentions due to the use of non-rational BJW strategies. However, we found the exact opposite, that intervention intentions were higher when witnessing a more severe assault. Removing our BJW lens, this finding appears to be consistent with some bystander research. For example, Huang et al. (2023) found greater intervention intentions in response to a more severe incident of cyberbullying than a less severe one. Similarly, Zhao et al. (2023) found that the perceived severity of cyberbullying predicted greater intentions to defend and support the victim. Thus, our findings appear to match existing bystander research. However, the explanation for this effect is still unclear. One would expect that increased severity has an effect on intervention intentions by removing any ambiguity and making it clearer that it is an emergency in which help is needed (stage two). But, the lack of a significant mediation pathway via stage two in Study 1, and lack of any significant effect on stage two in Study 2, casts doubt on this explanation. Huang et al. (2023) found that the effect of increased cyberbullying severity on intervention intentions was mediated by the bystander’s sense of responsibility. However, due to our severity manipulation having no effect on the responsibility to intervene, we can rule out this explanation. Thus, further research is needed to explore the mechanisms behind the effect of severity on intervention intentions in a sexual assault context.\u003c/p\u003e\n\u003cp\u003eAnother key finding across our two studies is the relative importance of the bystander stages. We had predicted that recognising the event as an emergency where help is needed (stage two) and accepting responsibility to intervene (stage 3) would be the key mediators, but instead found that stage two was less crucial than predicted, and that noticing the assault (stage one), and belief in your ability to intervene 4) were also important. On reflection it may not be surprising that stages one and four are involved. Firstly, the means for our bystander stage measures (see Table 1) show a greater hesitance in being able to notice the assault (stage one), perhaps being due to the party setting of our stimuli, where alcohol is likely consumed by bystanders (Haikalis et al., 2018), which research has suggested could hinder this stage (Burn, 2009; Steele \u0026amp; Josephs, 1990). Secondly, relating to belief in your ability to intervene, BJW highlights that it’s the belief in your ability to remove the injustice that determines whether you will use a rational or non-rational strategy (Lerner, 1980; Montada \u0026amp; Lerner, 1998; Walters et al., 2007; White et al., 2012), and much bystander research has demonstrated the importance of self-efficacy in determining bystander intervention (Banyard \u0026amp; Moynihan, 2011; Banyard et al., 2007; Bennett et al., 2014; Burn, 2009; Jouriles et al., 2019; Thornberg, 2013). Thus, the unpredicted role of stages one and four as mediators in Study 1 and predictors in Study 2 appear to fit with the literature.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBut regardless of these other stages, both studies consistently highlighted the importance of a bystander accepting responsibility to intervene, the third stage in Latané and Darley’s (1970) bystander model. In both studies, we found that a greater sense of responsibility for intervening predicted both greater general and behaviour-specific intervention intentions. When all variables of interest were entered simultaneously in our exploratory analysis, we found that acceptance of responsibility was the only variable which consistently predict intervention intentions, regardless of the stimuli type and measure of intervention intentions. Moreover, our mediation analysis showed that it consistently acted as a mediator between noticing the event (stage one) predicting general intervention intentions, and recognising the event as an emergency (stage two) predicting intervention intentions (although which intention measure this predicted differed between studies), further corroborating its importance. Although the role of accepting responsibility to intervene is not a novel finding, it does support the body of empirical research in this area which has found this (Bennett et al., 2014; Burn, 2009; Darley \u0026amp; Latané, 1968; Fischer et al., 2011; Hortensius \u0026amp; de Gelder, 2014; Jouriles et al., 2019; Katz et al., 2015; Levine, 2005). Where our research does add novelty is in highlighting the relative importance responsibility above and beyond other important factors such as severity, BJW strategies, and empathic variables.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA further interesting pattern from our studies was that we found more direct and indirect pathways predicting general intervention intentions than behaviour-specific intervention intentions. There could be several reasons for this. Firstly, it could be that individuals were more willing to say that they would intervene “in some way” than to agree to specific actions, in the same way that it’s easier to say you intend to do something but drop out when it gets down to the practicalities of implementing that intention (Ajzen, 1985).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eLimitations\u003c/h2\u003e\n\u003cp\u003eUnlike in Study 1, the severity manipulation within Study 2 did not have a significant effect on our manipulation check, namely, perceptions of unfairness, nor on deciding that intervention is needed (stage two of Latane and Darley’s (1970) model). A possible explanation for this could be related to the distinction between the two conditions. In Study 1, the difference in severity between conditions was quite prominent, being defined by whether the perpetrator put his hand up the victim’s dress. Whereas in Study 2, the difference between conditions was a more subtle stroke of the victim’s thigh. The reason for this was due to the ethical considerations when working with actors. During filming for Study 2, a priority was ensuring that the student actors felt comfortable and were not pushed to do any actions that they were not comfortable performing, resulting in potentially less of a difference between conditions. In future research using filmed 360-degree videos, we would recommend using more experienced actors (although more difficult in this case where the characters were student age) and hiring an intimacy co-ordinator when filming.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnother related limitation of our Study 2 design worth highlighting at this stage relates to the VR scene filming. In order to make the participant feel as if they are viewing from the first-person perspective of someone at the party, instead of having a fly on the wall point-of-view, and to ensure that they would witness the assault in full, the 360 camera was positioned in the middle of the room. The woman in the red dress (victim) was in one corner, and diagonally across from here in the other corner, was the man in the white t-shirt (perpetrator). As an unforeseen consequence of this, some participants mentioned that when the man was staring at the woman, it felt as if he was staring at the camera instead. As a result, this may have made some participants view themselves as a potential victim, instead of a bystander, which, for better or worse, could have influenced their decisions.\u003c/p\u003e\n\u003cp\u003eThere are several broader limitations worth mentioning. A key difference between our two studies which could explain the different findings is the sample composition. Our two samples differ in proportions representing genders, past sexual assault victimisation and past sexual harassment victimisation. With regards to gender, for Study 1, we were recruiting from a very large online pool, so were able to ensure a roughly equal split between male and female participants. This decision was based on research suggesting that gender could impact factors in the model (Burn, 2009; Katz et al., 2015; Mainwaring et al., 2022; Moschella-Smith et al., 2022). Whereas, in Study 2, we were recruiting from students limited by geographic location and subject, resulting in our final sample that was 82% women. This is likely due to 81% of psychology undergraduate students in the UK (from which we recruited most of our Study 2 sample) being female (Higher Education Statistics Agency, 2023), and women generally being more engaged in gender-based violence research (e.g. Bennett et al., 2014; Katz et al., 2015; Kistler et al., 2021; Pagliaro et al., 2021). Regarding past sexual assault victimisation, 38% of our Study 1 sample had previously been the victim of sexual assault, while we excluded those who had been assaulted from our Study 2 sample due to the emotionally impactful nature of the stimuli (Patil et al., 2014). Moreover, although a similar classification was used, Study 1’s percentage was much higher than what is seen in nationally representative data (16.64%; ONS, 2023a). However, this is likely due to the ONS survey using an age cut off, whereby they only included sexual assaults committed from a victim age of 16 onwards, while the questionnaire used here did not include an age restriction, thus capturing sexual assaults on minors too. We also had a much higher proportion of people who had been sexually harassed in our Study 1 sample (60.88%) compared to our Study 2 sample (28.21%). This difference could be due to the exclusion of people who have experienced sexual assault, which we indeed found in our Study 1 data to be positively correlated with sexual harassment experience (see Table 5), as well as screening out those who have experienced six or more symptoms of trauma related to experiences of sexual harassment in Study 2 only.\u003c/p\u003e\n\u003cp\u003eTaking these differences, as well as findings from Study 1 and past research suggesting that these factors could be related to our variables of interest (Bennett et al., 2017; Brockdorf et al., 2023; Mauer et al., 2022; Moschella-Smith et al., 2022; Vonderhaar, 2015), we ran impromptu exploratory analysis after Study 2 (on Study 1 data) to see if these variables moderated any of the effects and relationships in our predictions for which we found no confidence interval overlap between Studies 1 and 2. The only incidence where we found this to be the case was for sexual assault victimisation moderating the relationship between victim blaming and noticing the event (stage one), so that this negative relationship was only significant for those who had previously been sexually assaulted. Although the logic behind this moderation is unclear at this stage, it is important to take this into account when considering this relationship in our findings. The lack of any other significant moderation pathways helps to strengthen our conclusions when considering explanations of the differences between Study 1’s and Study 2’s findings outside of these sample differences.\u003c/p\u003e\n\u003cp\u003eThis project’s specific sample was chosen to reflect the adult age group and occupation most vulnerable to sexual assault, there are of course issues related to generalising these findings beyond university students. Related to this, there are also issues of generalising beyond the stimuli scenario. In both studies, our stimuli described an assault which bystanders directly witnesses, which covers only up to 29% of sexual assaults (Casper et al., 2022; Hamby et al., 2016; Hart \u0026amp; Miethe, 2008; Planty, 2002; Revolt Sexual Assault \u0026amp; The Student Room, 2018). Indeed it is far more common for bystanders to not be present at all, or only be present to witness events leading up to an assault (Haikalis et al., 2018). Moreover, our scenario featured a victim and perpetrator who appeared not to know each other, so there could be issues generalising to a situation where they have a pre-existing relationship, as there could be greater ambiguity for bystanders in this situation (Shotland \u0026amp; Straw, 1976). There also may be issues generalising from a party setting, to other assault settings such as at the victim’s or perpetrator’s home or in a public outdoor space (ONS, 2021). Thus, further research is needed to explore if factors like severity and accepting responsibility to intervene are as key when witnessing sexual assaults in other contexts too, varying the victim, perpetrator and participant age and occupation, at what stage bystanders are present, pre-existing relationships, and the assault setting. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMoreover, an additional limitation across both studies relates to measurement. The differing patterns according to which intervention intention measure was used could also be due to a methodological limitation of our behaviour-specific measure. As discussed in the Study 1 methods section, the behaviour-specific measure is calculated by listing 12 different possible intervention strategies and asking participants how likely they would be to use each strategy, and then averaging the likelihood score. It may have been the case that some participants would prefer some intervention strategies over others, for example due to the safety concerns of directly intervening. This would have meant that they might have selected that they would have been unlikely to intervene in direct ways, but very likely to intervene in indirect ways (e.g. talking to the victim afterwards, getting help from someone else). So although the participant may have been adamant that they would intervene in some way, this would not have been reflected in their mean score from the behaviour-specific measure. Thus, although adapted from an existing scale (Banyard \u0026amp; Moynihan, 2011), the behaviour-specific scale may need further development before use in future research. In terms of interpretation of our findings, these considerations suggest that perhaps caution should be given to pathways in these studies which were only found to predict one of our intervention intentions measures and not the other.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnd lastly, a key limitation for our research is the lack of causality. Beyond the few effects of our severity manipulation on our variables of interest, and our within-participants examination of changes in physiological arousal, it is important to emphasise that the rest of the analysis was purely correlational. Despite efforts to use a more conservative approach to mediation analysis to minimise the risks of spurious correlations, there is no guarantee of direction or causality. Predicted directions in our mediation models are purely hypothetical until further research has manipulated each element.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eImplications\u003c/h2\u003e\n\u003cp\u003eFirstly, these findings help to further our understanding of the NBJW and the bystander decision-making process, providing initial evidence for a link between two largely unconnected areas of social psychology. Secondly, this research helps to bolster Dawtry et al.’s (2020) arguments on the importance of emotionally impactful stimuli, and Study 2 demonstrates how filmed VR can be used effectively in a social psychological setting. However, although we can see clear distinctions in our findings from Studies 1 and 2, a more direct experimental comparison of text vignettes and filmed VR to confirm effects on emotional impact, BJW strategy use, empathic variables and bystander intervention is needed.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMost importantly, this research has positive implications beyond academic research in informing the design of bystander intervention programmes. Universities are coming under increasing pressure to tackle sexual assault due to campus locations being a hotspot for assaults (Fenton et al., 2015; Revolt Sexual Assault \u0026amp; The Student Room, 2018). In the US, it is now a legal requirement of colleges to include bystander intervention training in their educational programmes designed to combat sexual violence (\"Campus Sexual Violence Elimination Act,\" 2013). Despite a lack of equivalent legislation in the UK, many universities now have their own bystander intervention training schemes in place. Some training programmes like these include a focus on the role of Latané and Darley’s (1970) model (College of Policing, 2022; Fenton et al., 2015), and our findings demonstrating the importance of a bystander’s responsibility to intervene suggest that this could be an effective approach to take. However, our findings on the importance of psychological distancing in Study 1 and victim blaming in Study 2, suggest that there is need to take a more comprehensive approach to account for these roles. Such programmes could also be adapted for use in other settings, such as workplaces, schools, and community centres. Thus, these studies have positive implications for informing real-world efforts to reduce sexual assault, by encouraging bystanders to intervene.\u0026nbsp;\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eTo summarise, we conducted two studies to explore how BJW non-rational strategies might be linked to the bystander decision-making process, in the context of sexual assault. We found key differences in our findings across the two studies and offered potential explanations for these. However there are still several important results across both. Firstly, our severity manipulation had limited effects, but most notably higher severity made participants more likely to intervene, but without an obvious mechanism behind this. Mediation analysis suggested that in both studies, a BJW non-rational strategy was negatively related to the bystander\u0026rsquo;s general intentions to intervene. In study one, this relationship was mediated by reduced empathic concern and acceptance of responsibility to intervene, while in Study 2, there was no mediation among our variables of interest. Although explorative in nature and correlational, this research makes a first attempt at bringing together two previously unconnected areas of social psychology, casting a new perspective on theory and real-world sexual assault prevention.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthical Approval: Prior to fieldwork commencing, this study was given ethical approval by the Psychology Research Ethics Committee at the University of Bath (PREC reference numbers: 22-013 \u0026amp; 22-095).\u003c/p\u003e\n\u003cp\u003eCRediT author statement: Sarah Warbis: conceptualisation, methodology, formal analysis, investigation, resources, writing \u0026ndash; original draft, visualisation, and project administration. Dana\u0026euml; Stanton Fraser: conceptualisation, methodology, writing \u0026ndash; review \u0026amp; editing and supervision. Mitchell Callan: conceptualisation, methodology, writing \u0026ndash; review \u0026amp; editing and supervision.\u003c/p\u003e\n\u003cp\u003eCompeting interests: There are no relevant financial or non-financial competing interests to report.\u003c/p\u003e\n\u003cp\u003eEthical Approval: Prior to fieldwork commencing, this study was given ethical approval by the Psychology Research Ethics Committee at the University of Bath (PREC reference numbers: 22-013 \u0026amp; 22-095).\u003c/p\u003e\n\u003cp\u003eData Availability Statement: Anonymous data created during this research from participants who all consented to data sharing, plus other supporting information, will be made available upon publication from the University of Bath Research Data Archive.\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank Dr Harry Farmer, Jonathan Yip, Niamh Mulhern, Yeeun Kim, Ben Robins, Nathan Taylor, as well as the actors who starred in our filmed scene, and the University of Bath for funding this research.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAderman, D., Brehm, S. S., \u0026amp; Katz, L. B. (1974). Empathic observation of an innocent victim: The just world revisited. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;29\u003c/em\u003e(3), 342-347. https://doi.org/10.1037/h0036001\u003c/li\u003e\n \u003cli\u003eAhn, S. J., Le, A. M. T., \u0026amp; Bailenson, J. (2013). The effect of embodied experiences on self-other merging, attitude, and helping behavior. \u003cem\u003eMedia Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;16\u003c/em\u003e(1), 7-38. https://doi.org/10.1080/15213269.2012.755877\u003c/li\u003e\n \u003cli\u003eAjzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl \u0026amp; J. Beckmann (Eds.), \u003cem\u003eAction Control\u003c/em\u003e (pp. 11-39). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-69746-3_2\u003c/li\u003e\n \u003cli\u003eBandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. \u003cem\u003ePsychological Review\u003c/em\u003e,\u003cem\u003e\u0026nbsp;84\u003c/em\u003e(2), 191-215. https://doi.org/https://doi.org/10.1037/0033-295X.84.2.191\u003c/li\u003e\n \u003cli\u003eBa\u0026ntilde;os, R. M., Botella, C., Rubi\u0026oacute;, I., Quero, S., Garc\u0026iacute;a-Palacios, A., \u0026amp; Alca\u0026ntilde;iz, M. (2008). Presence and emotions in virtual environments: The influence of stereoscopy. \u003cem\u003eCyberPsychology \u0026amp; Behavior\u003c/em\u003e,\u003cem\u003e\u0026nbsp;11\u003c/em\u003e(1), 1-8. https://doi.org/10.1089/cpb.2007.9936\u003c/li\u003e\n \u003cli\u003eBanyard, V. L., \u0026amp; Moynihan, M. M. (2011). Variation in bystander behavior related to sexual and intimate partner violence prevention: Correlates in a sample of college students. \u003cem\u003ePsychology of Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;1\u003c/em\u003e(4), 287-301. https://doi.org/10.1037/a0023544\u003c/li\u003e\n \u003cli\u003eBanyard, V. L., Moynihan, M. M., \u0026amp; Plante, E. G. (2007). Sexual violence prevention through bystander education: An experimental evaluation. \u003cem\u003eJournal of Community Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;35\u003c/em\u003e(4), 463-481. https://doi.org/10.1002/jcop.20159\u003c/li\u003e\n \u003cli\u003eBanyard, V. L., Plante, E. G., \u0026amp; Moynihan, M. M. (2004). Bystander education: Bringing a broader community perspective to sexual violence prevention. \u003cem\u003eJournal of Community Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;32\u003c/em\u003e(1), 61-79. https://doi.org/10.1002/jcop.10078\u003c/li\u003e\n \u003cli\u003eBarreda-\u0026Aacute;ngeles, M., Serra-Blasco, M., Trepat, E., Pereda-Ba\u0026ntilde;os, A., P\u0026agrave;mias, M., Palao, D., . . . Cardoner, N. (2021). Development and experimental validation of a dataset of 360\u0026deg;-videos for facilitating school-based bullying prevention programs. \u003cem\u003eComputers \u0026amp; Education\u003c/em\u003e,\u003cem\u003e\u0026nbsp;161\u003c/em\u003e, 104065. https://doi.org/10.1016/j.compedu.2020.104065\u003c/li\u003e\n \u003cli\u003eBennett, S., Banyard, V. L., \u0026amp; Edwards, K. M. (2017). The Impact of the Bystander\u0026apos;s Relationship With the Victim and the Perpetrator on Intent to Help in Situations Involving Sexual Violence. \u003cem\u003eJ Interpers Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;32\u003c/em\u003e(5), 682-702. https://doi.org/10.1177/0886260515586373\u003c/li\u003e\n \u003cli\u003eBennett, S., Banyard, V. L., \u0026amp; Garnhart, L. (2014). To act or not to act, that is the question? Barriers and facilitators of bystander intervention. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;29\u003c/em\u003e(3), 476-496. https://doi.org/10.1177/0886260513505210\u003c/li\u003e\n \u003cli\u003eBlascovich, J., Loomis, J., Beall, A.C., Swinth, K.S., Hoyt, C.L., \u0026amp; Bailenson J.N. (2002). Immersive Virtual Environment Technology as a Methodological Tool for Social Psychology. \u003cem\u003ePsychological Inquiry\u003c/em\u003e,\u003cem\u003e\u0026nbsp;13\u003c/em\u003e(2), 103-125.\u003c/li\u003e\n \u003cli\u003eBoppana, S. (2021). \u003cem\u003eWhen men are the victims: Factors affecting rape victim blame \u0026amp; bystander aid\u0026nbsp;\u003c/em\u003e[Doctoral dissertation, University of Mississippi]. eGrove. https://egrove.olemiss.edu/etd/2149\u003c/li\u003e\n \u003cli\u003eBowes-Sperry, L. (2005). To act or not to act: The dilemma faced by sexual harassment observers. \u003cem\u003eAcademy of Management Review\u003c/em\u003e,\u003cem\u003e\u0026nbsp;30\u003c/em\u003e(2), 288-307. https://doi.org/10.5465/AMR.2005.16387886\u003c/li\u003e\n \u003cli\u003eBradley, M. M., \u0026amp; Lang, P. J. (1994). Measuring emotion: The self-assessment manikin and the semantic differential. \u003cem\u003eJournal of Behavior Therapy and Experimental Psychiatry\u003c/em\u003e,\u003cem\u003e\u0026nbsp;25\u003c/em\u003e(1), 49-59. https://doi.org/https://doi.org/10.1016/0005-7916(94)90063-9\u003c/li\u003e\n \u003cli\u003eBrewin, C. R., Rose, S., Andrews, B., Green, J., Tata, P., McEvedy, C., . . . Foa, E. B. (2002). Brief screening instrument for post-traumatic stress disorder. \u003cem\u003eBritish Journal of Psychiatry\u003c/em\u003e,\u003cem\u003e\u0026nbsp;181\u003c/em\u003e(02), 158-162. https://doi.org/10.1017/s0007125000161896\u003c/li\u003e\n \u003cli\u003eBrockdorf, A. N., Eagan, S. R., Zosky, J. E., Grandgenett, H. M., Gervais, S. J., \u0026amp; Dilillo, D. (2023). Associations Between Sexual Objectification and Bystander Efficacy: The Mediating Role of Barriers to Bystander Intervention [Article]. \u003cem\u003eViolence and Victims\u003c/em\u003e,\u003cem\u003e\u0026nbsp;38\u003c/em\u003e(1), 148-164. https://doi.org/10.1891/VV-2021-0168\u003c/li\u003e\n \u003cli\u003eBurn, S. M. (2009). A situational model of sexual assault prevention through bystander intervention. \u003cem\u003eSex Roles\u003c/em\u003e,\u003cem\u003e\u0026nbsp;60\u003c/em\u003e(11/12), 779-793. https://doi.org/10.1007/s11199-008-9581-5\u003c/li\u003e\n \u003cli\u003eBurt, M. R. (1980). Cultural myths and supports for rape. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;38\u003c/em\u003e(2), 217-230. https://doi.org/10.1037/0022-3514.38.2.217\u003c/li\u003e\n \u003cli\u003eCampus Sexual Violence Elimination Act, S.128 (2013). https://www.congress.gov/bill/113th-congress/senate-bill/128/text\u003c/li\u003e\n \u003cli\u003eCarey, K. B., Norris, A. L., Durney, S. E., Shepardson, R. L., \u0026amp; Carey, M. P. (2018). Mental health consequences of sexual assault among first-year college women. \u003cem\u003eJournal of American College Health\u003c/em\u003e,\u003cem\u003e\u0026nbsp;66\u003c/em\u003e(6), 480-486. https://doi.org/10.1080/07448481.2018.1431915\u003c/li\u003e\n \u003cli\u003eCasper, D. M., Witte, T., Gibson, E., \u0026amp; McCulley, E. (2022). \u0026ldquo;I Pulled Them Apart and Told Them to Stop\u0026rdquo;: A Mixed-Methods Examination of Bystander Behavior, Preparedness, and Emotional Reaction. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e. https://doi.org/10.1177/08862605221092071\u003c/li\u003e\n \u003cli\u003eChauhan, M. (2022). Gender and Age Related Differences in Attitude towards Rape and Acceptance of Rape Myths. \u003cem\u003eThe International Journal of Indian Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;10\u003c/em\u003e(3). https://doi.org/10.25215/1003.130\u003c/li\u003e\n \u003cli\u003eCollege of Policing. (2022). \u003cem\u003eBystander Programmes: Evidence briefing\u003c/em\u003e. \u0026nbsp;Retrieved from https://assets.college.police.uk/s3fs-public/2022-03/Bystander-programmes-evidence-briefing.pdf\u003c/li\u003e\n \u003cli\u003eCorreia, I., \u0026amp; Vala, J. (2003). When will a victim be secondarily victimized? The effect of observer\u0026apos;s belief in a just world, victim\u0026apos;s innocence and persistence of suffering. \u003cem\u003eSocial Justice Research\u003c/em\u003e,\u003cem\u003e\u0026nbsp;16\u003c/em\u003e(4), 379-400. https://doi.org/10.1023/A:1026313716185\u003c/li\u003e\n \u003cli\u003eDarley, J. M., \u0026amp; Latan\u0026eacute;, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;8\u003c/em\u003e(4), 377-383. https://doi.org/10.1037/h0025589\u003c/li\u003e\n \u003cli\u003eDavis, M. H. (1980). A multidimensional approach to individual differences in empathy. \u003cem\u003eJSAS catalog of selected documents in psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;10\u003c/em\u003e, 85.\u003c/li\u003e\n \u003cli\u003eDawtry, R. J., Callan, M. J., Harvey, A. J., \u0026amp; Gheorghiu, A. I. (2020). Victims, vignettes, and videos: Meta-analytic and experimental evidence that emotional impact enhances the derogation of innocent victims. \u003cem\u003ePersonality and Social Psychology Review\u003c/em\u003e,\u003cem\u003e\u0026nbsp;24\u003c/em\u003e(3), 233-259. https://doi.org/10.1177/1088868320914208\u003c/li\u003e\n \u003cli\u003eDawtry, R. J., Callan, M. J., Harvey, A. J., \u0026amp; Olson, J. M. (2018). Derogating innocent victims: The effects of relative versus absolute character judgments. \u003cem\u003ePersonality and Social Psychology Bulletin\u003c/em\u003e. https://doi.org/10.1177/0146167217733078\u003c/li\u003e\n \u003cli\u003eDe La Pe\u0026ntilde;a, N., Weil, P., Llobera, J., Spanlang, B., Friedman, D., Sanchez-Vives, M. V., \u0026amp; Slater, M. (2010). Immersive Journalism: Immersive Virtual Reality for the First-Person Experience of News. \u003cem\u003ePresence: Teleoperators and Virtual Environments\u003c/em\u003e,\u003cem\u003e\u0026nbsp;19\u003c/em\u003e(4), 291-301. https://doi.org/10.1162/pres_a_00005\u003c/li\u003e\n \u003cli\u003eEmbury-Dennis, T. (2018, March 30). Judge who asked alleged rape victim if she tried to \u0026lsquo;close her legs\u0026rsquo; charged with misconduct.\u003cem\u003e\u0026nbsp;The Independent\u003c/em\u003e. https://www.independent.co.uk/news/world/americas/judge-rape-victim-close-your-legs-john-russo-new-jersey-complaint-a8281356.html\u003c/li\u003e\n \u003cli\u003eFenton, R. A., Mott, H. L., McCartan, K., \u0026amp; Rumney, P. N. (2015). \u003cem\u003eA review of evidence for bystander intervention to prevent sexual and domestic violence in universities\u003c/em\u003e. Public Health England. https://www.gov.uk/government/publications/sexual-and-domestic-violence-prevention-in-universities-evidence-review\u003c/li\u003e\n \u003cli\u003eFern\u0026aacute;ndez, C., Pascual, J. C., Soler, J., Elices, M., Portella, M. J., \u0026amp; Fern\u0026aacute;ndez-Abascal, E. (2012). Physiological Responses Induced by Emotion-Eliciting Films. \u003cem\u003eApplied Psychophysiology and Biofeedback\u003c/em\u003e,\u003cem\u003e\u0026nbsp;37\u003c/em\u003e(2), 73-79. https://doi.org/10.1007/s10484-012-9180-7\u003c/li\u003e\n \u003cli\u003eFinch, E., \u0026amp; Munro, V. E. (2007). The demon drink and the demonized woman: Socio-sexual stereotypes and responsibility attribution in rape trials involving intoxicants. \u003cem\u003eSocial \u0026amp; Legal Studies\u003c/em\u003e,\u003cem\u003e\u0026nbsp;16\u003c/em\u003e(4), 591-614. https://doi.org/10.1177/0964663907082737\u003c/li\u003e\n \u003cli\u003eFine, M. (1983). The social context and a sense of injustice: The option to challenge. \u003cem\u003eRepresentative Research in Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;13\u003c/em\u003e(1), 15-33.\u003c/li\u003e\n \u003cli\u003eFischer, P., Krueger, J. I., Greitemeyer, T., Vogrincic, C., Kastenm\u0026uuml;ller, A., Frey, D., . . . Kainbacher, M. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous and non-dangerous emergencies. \u003cem\u003ePsychological Bulletin\u003c/em\u003e,\u003cem\u003e\u0026nbsp;137\u003c/em\u003e(4), 517-537. https://doi.org/10.1037/a0023304\u003c/li\u003e\n \u003cli\u003eGalop. (2022). \u003cem\u003eLGBT+ People \u0026amp; Sexual Violence\u003c/em\u003e. https://galop.org.uk/resource/lgbt-people-sexual-violence-report/\u003c/li\u003e\n \u003cli\u003eGerber, G. L., Cronin, J. M., \u0026amp; Steigman, H. J. (2004). Attributions of blame in sexual assault to perpetrators and victims of both genders. \u003cem\u003eJournal of Applied Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;34\u003c/em\u003e(10), 2149-2165. https://doi.org/10.1111/j.1559-1816.2004.tb02694.x\u003c/li\u003e\n \u003cli\u003eGlick, P., \u0026amp; Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;70\u003c/em\u003e(3), 491-512. https://doi.org/10.1037/0022-3514.70.3.491\u003c/li\u003e\n \u003cli\u003eGreco, A., Valenza, G., Citi, L., \u0026amp; Scilingo, E. P. (2016). Arousal and valence recognition of affective sounds based on electrodermal activity. \u003cem\u003eIEEE Sensors Journal\u003c/em\u003e,\u003cem\u003e\u0026nbsp;17\u003c/em\u003e(3), 716-725.\u003c/li\u003e\n \u003cli\u003eHafer, C. L. (2000). Investment in long-term goals and commitment to just means drive the need to believe in a just world. \u003cem\u003ePersonality and Social Psychology Bulletin\u003c/em\u003e,\u003cem\u003e\u0026nbsp;26\u003c/em\u003e(9), 1059-1073. https://doi.org/10.1177/01461672002611004\u003c/li\u003e\n \u003cli\u003eHafer, C. L., \u0026amp; B\u0026egrave;gue, L. (2005). Experimental research on just-world theory: Problems, developments, and future challenges. \u003cem\u003ePsychological Bulletin\u003c/em\u003e,\u003cem\u003e\u0026nbsp;131\u003c/em\u003e(1), 128-167. https://doi.org/10.1037/0033-2909.131.1.128\u003c/li\u003e\n \u003cli\u003eHafer, C. L., \u0026amp; Rubel, A. N. (2015). The why and how of defending belief in a just world. \u003cem\u003eAdvances in Experimental Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;51\u003c/em\u003e(1), 41-96. https://doi.org/10.1016/bs.aesp.2014.09.001\u003c/li\u003e\n \u003cli\u003eHafer, C. L., Rubel, A. N., \u0026amp; Drolet, C. E. (2019). Experimental evidence of subtle victim blame in the absence of explicit blame. \u003cem\u003ePLOS ONE\u003c/em\u003e,\u003cem\u003e\u0026nbsp;14\u003c/em\u003e(12), e0227229. https://doi.org/10.1371/journal.pone.0227229\u003c/li\u003e\n \u003cli\u003eHaikalis, M., Leone, R. M., Parrott, D. J., \u0026amp; Dilillo, D. (2018). Sexual Assault Survivor Reports of Missed Bystander Opportunities: The Role of Alcohol, Sexual Objectification, and Relational Factors. \u003cem\u003eViolence Against Women\u003c/em\u003e,\u003cem\u003e\u0026nbsp;24\u003c/em\u003e(10), 1232-1254. https://doi.org/10.1177/1077801218781941\u003c/li\u003e\n \u003cli\u003eHamby, S., Weber, M. C., Grych, J., \u0026amp; Banyard, V. (2016). What difference do bystanders make? The association of bystander involvement with victim outcomes in a community sample. \u003cem\u003ePsychology of Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;6\u003c/em\u003e(1), 91-102. https://doi.org/10.1037/a0039073\u003c/li\u003e\n \u003cli\u003eHart, T. C., \u0026amp; Miethe, T. D. (2008). Exploring bystander presence and intervention in nonfatal violent victimization: When does helping really help? \u003cem\u003eViolence and Victims\u003c/em\u003e,\u003cem\u003e\u0026nbsp;23\u003c/em\u003e(5), 637-651. https://doi.org/10.1891/0886-6708.23.5.637\u003c/li\u003e\n \u003cli\u003eHarvey, A. J., Callan, M. J., \u0026amp; Matthews, W. J. (2014). How much does effortful thinking underlie observers\u0026rsquo; reactions to victimization? \u003cem\u003eSocial Justice Research\u003c/em\u003e,\u003cem\u003e\u0026nbsp;27\u003c/em\u003e(2), 175-208. https://doi.org/10.1007/s11211-014-0209-3\u003c/li\u003e\n \u003cli\u003eHayes, A. F. (2022). \u003cem\u003eIntroduction to mediation, moderation, and conditional process analysis: A regression-based approach\u003c/em\u003e (3rd ed.). The Guilford Press.\u003c/li\u003e\n \u003cli\u003eHigher Education Statistics Agency. (2023, January 19). \u003cem\u003eHigher Education Student Statistics: UK, 2021/22\u003c/em\u003e. https://www.hesa.ac.uk/news/19-01-2023/sb265-higher-education-student-statistics/subjects\u003c/li\u003e\n \u003cli\u003eHome Office. (2023). \u003cem\u003eCrime outcomes in England and Wales, year ending March 2023: Data tables\u003c/em\u003e [Data tables]. Retrieved December 21, 2023 from https://www.gov.uk/government/statistics/crime-outcomes-in-england-and-wales-2022-to-2023\u003c/li\u003e\n \u003cli\u003eHortensius, R., \u0026amp; de Gelder, B. (2014). The neural basis of the bystander effect \u0026mdash; The influence of group size on neural activity when witnessing an emergency. \u003cem\u003eNeuroImage (Orlando, Fla.)\u003c/em\u003e,\u003cem\u003e\u0026nbsp;93\u003c/em\u003e(1), 53-58. https://doi.org/10.1016/j.neuroimage.2014.02.025\u003c/li\u003e\n \u003cli\u003eHuang, L., Li, W., Xu, Z., Sun, H., Ai, D., Hu, Y., . . . Zhou, Y. (2023). The Severity of Cyberbullying Affects Bystander Intervention Among College Students: The Roles of Feelings of Responsibility and Empathy [Article]. \u003cem\u003ePsychology Research and Behavior Management\u003c/em\u003e,\u003cem\u003e\u0026nbsp;16\u003c/em\u003e, 893-903. https://doi.org/10.2147/PRBM.S397770\u003c/li\u003e\n \u003cli\u003eJicol, C., Cheng, H. Y., Petrini, K., \u0026amp; O\u0026rsquo;Neill, E. (2023). A predictive model for understanding the role of emotion for the formation of presence in virtual reality. \u003cem\u003ePLOS ONE\u003c/em\u003e,\u003cem\u003e\u0026nbsp;18\u003c/em\u003e(3), e0280390. https://doi.org/10.1371/journal.pone.0280390\u003c/li\u003e\n \u003cli\u003eJicol, C., Feltham, J., Yoon, J., Proulx, M. J., O\u0026apos;Neill, E., \u0026amp; Lutteroth, C. (2022, April 29-May 5, 2022). Designing and Assessing a Virtual Reality Simulation to Build Resilience to Street Harassment. CHI Conference on Human Factors in Computing Systems, New Orleans, LA, USA.\u003c/li\u003e\n \u003cli\u003eJordan, M. R., Amir, D., \u0026amp; Bloom, P. (2016). Are empathy and concern psychologically distinct? \u003cem\u003eEmotion\u003c/em\u003e,\u003cem\u003e\u0026nbsp;16\u003c/em\u003e(8), 1107-1116. https://doi.org/10.1037/emo0000228\u003c/li\u003e\n \u003cli\u003eJouriles, E. N., McDonald, R., Rosenfield, D., \u0026amp; Sargent, K. S. (2019). Increasing bystander behavior to prevent adolescent relationship violence: A randomized controlled trial. \u003cem\u003eJournal of Consulting and Clinical Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;87\u003c/em\u003e(1), 3-15. https://doi.org/10.1037/ccp0000355\u003c/li\u003e\n \u003cli\u003eKahn, A. S., \u0026amp; Cargile, A. C. (2021). Immersive and Interactive Awe: Evoking Awe via Presence in Virtual Reality and Online Videos to Prompt Prosocial Behavior. \u003cem\u003eHuman Communication Research\u003c/em\u003e,\u003cem\u003e\u0026nbsp;47\u003c/em\u003e(4), 387-417. https://doi.org/10.1093/hcr/hqab007\u003c/li\u003e\n \u003cli\u003eKatz, J. (2015). Effects of Group Status and Victim Sex on Male Bystanders\u0026apos; Responses to a Potential Party Rape. \u003cem\u003eJournal of Aggression, Maltreatment \u0026amp; Trauma\u003c/em\u003e,\u003cem\u003e\u0026nbsp;24\u003c/em\u003e(5), 588-602. https://doi.org/10.1080/10926771.2015.1029184\u003c/li\u003e\n \u003cli\u003eKatz, J., Olin, R., Herman, C., \u0026amp; DuBois, M. (2013). Spotting the signs: First‐year college students\u0026apos;responses to bystander‐themed rape prevention posters. \u003cem\u003eJournal of Community Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;41\u003c/em\u003e(4), 523-529.\u003c/li\u003e\n \u003cli\u003eKatz, J., Pazienza, R., Olin, R., \u0026amp; Rich, H. (2015). That\u0026rsquo;s what friends are for. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;30\u003c/em\u003e(16), 2775-2792. https://doi.org/10.1177/0886260514554290\u003c/li\u003e\n \u003cli\u003eKistler, T., Stuart, G. L., \u0026amp; Cornelius, T. L. (2021). Barriers to Bystander Interventions for Sexual Assault: Impact of Personal History of Sexual Victimization. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;37\u003c/em\u003e(17-18), NP16727-NP16749. https://doi.org/10.1177/08862605211023491\u003c/li\u003e\n \u003cli\u003eKreibig, S. D., Wilhelm, F. H., Roth, W. T., \u0026amp; Gross, J. J. (2007). Cardiovascular, electrodermal, and respiratory response patterns to fear‐ and sadness‐inducing films. \u003cem\u003ePsychophysiology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;44\u003c/em\u003e(5), 787-806. https://doi.org/10.1111/j.1469-8986.2007.00550.x\u003c/li\u003e\n \u003cli\u003eLabhardt, D., Holdsworth, E., Brown, S., \u0026amp; Howat, D. (2017). You see but you do not observe: A review of bystander intervention and sexual assault on university campuses. \u003cem\u003eAggression and Violent Behavior\u003c/em\u003e,\u003cem\u003e\u0026nbsp;35\u003c/em\u003e, 13-25. https://doi.org/10.1016/j.avb.2017.05.005\u003c/li\u003e\n \u003cli\u003eLatan\u0026eacute;, B., \u0026amp; Darley, J. M. (1970). \u003cem\u003eThe unresponsive bystander: Why doesn\u0026rsquo;t he help?\u003c/em\u003e Appleton Century Crofts.\u003c/li\u003e\n \u003cli\u003eLatta, R. (1976). \u003cem\u003eAn attribution theory interpretation of reactions to victims\u003c/em\u003e [Doctoral Dissertation, Iowa State University]. Iowa State University Digital Repository. https://dr.lib.iastate.edu/handle/20.500.12876/78338\u003c/li\u003e\n \u003cli\u003eLerner, M. (1980). \u003cem\u003eThe Belief in a Just World : A Fundamental Delusion\u003c/em\u003e (1st ed.). New York, NY : Springer US : Imprint: Springer.\u003c/li\u003e\n \u003cli\u003eLerner, M. J., \u0026amp; Miller, D. T. (1978). Just world research and the attribution process: Looking back and ahead. \u003cem\u003ePsychological Bulletin\u003c/em\u003e,\u003cem\u003e\u0026nbsp;85\u003c/em\u003e(5), 1030-1051. https://doi.org/10.1037/0033-2909.85.5.1030\u003c/li\u003e\n \u003cli\u003eLerner, M. J., \u0026amp; Simmons, C. H. (1966). Observer\u0026apos;s reaction to the \u0026quot;innocent victim\u0026quot;: Compassion or rejection? \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;4\u003c/em\u003e(2), 203-210. https://doi.org/10.1037/h0023562\u003c/li\u003e\n \u003cli\u003eLevine, M. (2005). Identity and Emergency Intervention: How Social Group Membership and Inclusiveness of Group Boundaries Shape Helping Behavior. \u003cem\u003ePersonality \u0026amp; Social Psychology Bulletin\u003c/em\u003e,\u003cem\u003e\u0026nbsp;31\u003c/em\u003e(4), 443-454. https://doi.org/10.1177/0146167204271651\u003c/li\u003e\n \u003cli\u003eLi, X., Lu, H., Wang, H., Zhu, P., \u0026amp; Zhang, J. (2018). General belief in a just world, moral disengagement, and helping propensity in emergencies. \u003cem\u003eSocial Behavior \u0026amp; Personality: An International Journal\u003c/em\u003e,\u003cem\u003e\u0026nbsp;46\u003c/em\u003e(11), 1923-1937. https://doi.org/10.2224/sbp.7407\u003c/li\u003e\n \u003cli\u003eLinhares, L., \u0026amp; Torres, A. R. R. (2021). She deserved it: Analysis of variables that influence the accountability of victims of sexual violence. \u003cem\u003eActa Colombiana de Psicolog\u0026iacute;a\u003c/em\u003e,\u003cem\u003e\u0026nbsp;25\u003c/em\u003e(1), 218-229.\u003c/li\u003e\n \u003cli\u003eLuce, H., Schrager, S. B., \u0026amp; Gilchrist, V. (2010). Sexual assault of women. \u003cem\u003eAmerican Family Physician\u003c/em\u003e,\u003cem\u003e\u0026nbsp;81\u003c/em\u003e(4), 489-495.\u003c/li\u003e\n \u003cli\u003eMainwaring, C., Gabbert, F., \u0026amp; Scott, A. J. (2022). A systematic review exploring variables related to bystander intervention in sexual violence contexts. \u003cem\u003eTrauma, Violence, and Abuse\u003c/em\u003e,\u003cem\u003e\u0026nbsp;0\u003c/em\u003e(0), 1-16. https://doi.org/10.1177/15248380221079660\u003c/li\u003e\n \u003cli\u003eMason, F., \u0026amp; Lodrick, Z. (2013). Psychological consequences of sexual assault. \u003cem\u003eBest Practice \u0026amp; Research Clinical Obstetrics \u0026amp; Gynaecology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;27\u003c/em\u003e(1), 27-37. https://doi.org/10.1016/j.bpobgyn.2012.08.015\u003c/li\u003e\n \u003cli\u003eMauer, V. A., Edwards, K. M., Waterman, E. A., Dardis, C. M., Dworkin, E. R., Rodriguez, L. M., \u0026amp; Ullman, S. E. (2022). Disclosure recipients\u0026rsquo; perceptions related to helping victims of dating and sexual violence. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;0\u003c/em\u003e(0), 1-24. https://doi.org/10.1177/08862605211063004\u003c/li\u003e\n \u003cli\u003eMcFarlane, J., Malecha, A., Watson, K., Gist, J., Batten, E., Hall, I., \u0026amp; Smith, S. (2005). Intimate partner sexual assault against women: Frequency, health consequences, and treatment outcomes. \u003cem\u003eObstetrics \u0026amp; Gynecology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;105\u003c/em\u003e(1), 99-108. https://doi.org/10.1097/01.AOG.0000146641.98665.b6\u003c/li\u003e\n \u003cli\u003eMontada, L., \u0026amp; Lerner, M. J. (1998). \u003cem\u003eResponses to Victimizations and Belief in a Just World\u003c/em\u003e (1st ed.). Plenum Press.\u003c/li\u003e\n \u003cli\u003eMoschella-Smith, E. A., Potter, S. J., \u0026amp; Moynihan, M. M. (2022). Correlates of bystander intentions and behavior among community college students in situations of sexual violence. \u003cem\u003eJournal of Prevention\u003c/em\u003e,\u003cem\u003e\u0026nbsp;43\u003c/em\u003e, 277-295. https://doi.org/10.1007/s10935-021-00663-y\u003c/li\u003e\n \u003cli\u003eNavarrete, C. D., McDonald, M. M., Mott, M. L., \u0026amp; Asher, B. (2012). Virtual morality: Emotion and action in a simulated three-dimensional \u0026ldquo;trolley problem\u0026rdquo;. \u003cem\u003eEmotion\u003c/em\u003e,\u003cem\u003e\u0026nbsp;12\u003c/em\u003e(2), 364-370. https://doi.org/10.1037/a0025561\u003c/li\u003e\n \u003cli\u003eNickerson, A. B. (2014). Measurement of the bystander intervention model for bullying and sexual harassment. \u003cem\u003eJournal of Adolescence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;37\u003c/em\u003e(4), 391-401. https://doi.org/10.1016/j.adolescence.2014.03.003\u003c/li\u003e\n \u003cli\u003eOffice for National Statistics. (2021).\u0026nbsp;\u003cem\u003eNature of sexual assault by rape or penetration, England and Wales: Year ending March 2020\u003c/em\u003e. \u0026nbsp;Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/\u003cbr\u003enatureofsexualassaultbyrapeorpenetrationenglandandwales/yearendingmarch\u003cbr\u003e2020#sources-of-support-sought-by-victims\u003c/li\u003e\n \u003cli\u003eOffice for National Statistics. (2023a).\u0026nbsp;\u003cem\u003eSexual offences prevalence and victim characteristics, England and Wales\u003c/em\u003e [Data tables]. Retrieved\u0026nbsp;\u003cbr\u003eJanuary 26, 2022, from https://www.ons.gov.uk/peoplepopulationandcommunity/\u003cbr\u003ecrimeandjustice/datasets/sexualoffencesprevalenceandvictimcharacteristicsenglandandwales\u003c/li\u003e\n \u003cli\u003eOffice for National Statistics. (2023b).\u0026nbsp;\u003cem\u003eSexual offences victim characteristics, England and Wales: year ending March 2022\u003c/em\u003e. \u0026nbsp;Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/\u003cbr\u003esexualoffencesvictimcharacteristicsenglandandwales/yearendingmarch2022/pdf\u003c/li\u003e\n \u003cli\u003eOppenheim, M. (2022 August 18). Judge dismisses woman\u0026rsquo;s sexual assault complaint because she wore \u0026lsquo;provocative clothes\u0026rsquo;.\u003cem\u003e\u0026nbsp;The Independent\u003c/em\u003e. https://www.independent.co.uk/asia/sexual-assault-judge-india-provocative-clothes-b2147722.html\u003c/li\u003e\n \u003cli\u003eOrchowski, L. M., Oesterle, D. W., \u0026amp; Haikalis, M. (2022). What Stops Unwanted Sexual and Social Advances Made by Heavy Drinking College Men? \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e. https://doi.org/10.1177/08862605211072157\u003c/li\u003e\n \u003cli\u003ePagliaro, S., Paolini, D., \u0026amp; Pacilli, M. G. (2021). Intimate Partner Violence and Same-Sex Couples: Examining the Antecedents of the Helping Intentions of Bystanders. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;36\u003c/em\u003e(21-22), NP11593-NP11617. https://doi.org/10.1177/0886260519888530\u003c/li\u003e\n \u003cli\u003ePan, X., \u0026amp; Slater, M. (2011, 2011). Confronting a Moral Dilemma in Virtual Reality: A Pilot Study.\u003c/li\u003e\n \u003cli\u003ePatil, I., Cogoni, C., Zangrando, N., Chittaro, L., \u0026amp; Silani, G. (2014). Affective basis of judgment-behavior discrepancy in virtual experiences of moral dilemmas. \u003cem\u003eSocial Neuroscience\u003c/em\u003e,\u003cem\u003e\u0026nbsp;9\u003c/em\u003e(1), 94-107. https://doi.org/10.1080/17470919.2013.870091\u003c/li\u003e\n \u003cli\u003ePatil, I., Zanon, M., Novembre, G., Zangrando, N., Chittaro, L., \u0026amp; Silani, G. (2018). Neuroanatomical basis of concern-based altruism in virtual environment. \u003cem\u003eNeuropsychologia\u003c/em\u003e,\u003cem\u003e\u0026nbsp;116\u003c/em\u003e(Pt A), 34-43. https://doi.org/10.1016/j.neuropsychologia.2017.02.015\u003c/li\u003e\n \u003cli\u003ePazda, A. D., Elliot, A. J., \u0026amp; Greitemeyer, T. (2012). Sexy red: Perceived sexual receptivity mediates the red-attraction relation in men viewing woman. \u003cem\u003eJournal of Experimental Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;48\u003c/em\u003e(3), 787-790. https://doi.org/https://doi.org/10.1016/j.jesp.2011.12.009\u003c/li\u003e\n \u003cli\u003ePlanty, M. (2002). \u003cem\u003eThird-party involvement in violent crime, 1993-99\u003c/em\u003e. U.S. Department of Justice: Office of Justice Programs. https://www.ojp.gov/ncjrs/virtual-library/abstracts/third-party-involvement-violent-crime-1993-99\u003c/li\u003e\n \u003cli\u003eRevolt Sexual Assault, \u0026amp; The Student Room. (2018). \u003cem\u003eSexual Violence at Universities Statistical Report\u003c/em\u003e. https://revoltsexualassault.com/research/\u003c/li\u003e\n \u003cli\u003eRussell, J. A., \u0026amp; Mehrabian, A. (1977). Evidence for a three-factor theory of emotions. \u003cem\u003eJournal of Research in Personality\u003c/em\u003e,\u003cem\u003e\u0026nbsp;11\u003c/em\u003e(3), 273-294. https://doi.org/https://doi.org/10.1016/0092-6566(77)90037-X\u003c/li\u003e\n \u003cli\u003eSchaaff, K., \u0026amp; Adam, M. T. (2013). Measuring emotional arousal for online applications: Evaluation of ultra-short term heart rate variability measures. 2013 Humaine association conference on affective computing and intelligent interaction,\u003c/li\u003e\n \u003cli\u003eSeinfeld, S., Arroyo-Palacios, J., Iruretagoyena, G., Hortensius, R., Zapata, L. E., Borland, D., . . . Sanchez-Vives, M. V. (2018). Offenders become the victim in virtual reality: impact of changing perspective in domestic violence. \u003cem\u003eScientific Reports\u003c/em\u003e,\u003cem\u003e\u0026nbsp;8\u003c/em\u003e(1). https://doi.org/10.1038/s41598-018-19987-7\u003c/li\u003e\n \u003cli\u003eShotland, R. L., \u0026amp; Straw, M. K. (1976). Bystander response to an assault: When a man attacks a woman. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;34\u003c/em\u003e(5), 990-999. https://doi.org/10.1037/0022-3514.34.5.990\u003c/li\u003e\n \u003cli\u003eSinger, T., Seymour, B., O\u0026apos;Doherty, J. P., Stephan, K. E., Dolan, R. J., \u0026amp; Frith, C. D. (2006). Empathic neural responses are modulated by the perceived fairness of others. \u003cem\u003eNature\u003c/em\u003e,\u003cem\u003e\u0026nbsp;439\u003c/em\u003e(7075), 466-469. https://doi.org/10.1038/nature04271\u003c/li\u003e\n \u003cli\u003eSlater, M. (2018). Immersion and the illusion of presence in virtual reality. \u003cem\u003eBritish Journal of Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;109\u003c/em\u003e(3), 431-434. https://doi.org/10.1111/bjop.12305\u003c/li\u003e\n \u003cli\u003eSlater, M., Antley, A., Davison, A., Swapp, D., Guger, C., Barker, C., . . . Sanchez-Vives, M. V. (2006). A virtual reprise of the Stanley Milgram obedience experiments. \u003cem\u003ePLOS ONE\u003c/em\u003e,\u003cem\u003e\u0026nbsp;1\u003c/em\u003e(1), e39-e39. https://doi.org/10.1371/journal.pone.0000039\u003c/li\u003e\n \u003cli\u003eSlater, M., Rovira, A., Southern, R., Swapp, D., Zhang, J. J., Campbell, C., \u0026amp; Levine, M. (2013). Bystander Responses to a Violent Incident in an Immersive Virtual Environment. \u003cem\u003ePLOS ONE\u003c/em\u003e,\u003cem\u003e\u0026nbsp;8\u003c/em\u003e(1), e52766-e52766. https://doi.org/10.1371/journal.pone.0052766\u003c/li\u003e\n \u003cli\u003eSlater, M., Usoh, M., \u0026amp; Steed, A. (1994). Depth of Presence in Immersive Virtual Environments. \u003cem\u003eTeleoperators and Virtual Environments - Presence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;3\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eSmith, C. A., \u0026amp; Frieze, I. H. (2003). Examining rape empathy from the perspective of the victim and the assailant. \u003cem\u003eJournal of Applied Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;33\u003c/em\u003e(3), 476-498. https://doi.org/10.1111/j.1559-1816.2003.tb01907.x\u003c/li\u003e\n \u003cli\u003eSpaccatini, F., Pacilli, M. G., Pagliaro, S., \u0026amp; Giovannelli, I. (2022). Victim blaming 2.0: Blaming sexualized victims of online harassment lowers bystanders\u0026rsquo; helping intentions. \u003cem\u003eCurrent Psychology\u003c/em\u003e. https://doi.org/10.1007/s12144-022-02884-8\u003c/li\u003e\n \u003cli\u003eSteele, C. M., \u0026amp; Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous effects. \u003cem\u003eAmerican Psychologist\u003c/em\u003e,\u003cem\u003e\u0026nbsp;45\u003c/em\u003e(8), 921-933. https://doi.org/10.1037/0003-066X.45.8.921\u003c/li\u003e\n \u003cli\u003eSt\u0026ouml;ber, J. (2001). The Social Desirability Scale-17 (SDS-17): Convergent validity, discriminant validity, and relationship with age. \u003cem\u003eEuropean Journal of Psychological Assessment\u003c/em\u003e,\u003cem\u003e\u0026nbsp;17\u003c/em\u003e(3), 222-232. https://doi.org/10.1027/1015-5759.17.3.222\u003c/li\u003e\n \u003cli\u003eStr\u0026ouml;mwall, L. A., Landstr\u0026ouml;m, S., \u0026amp; Alfredsson, H. (2014). Perpetrator characteristics and blame attributions in a stranger rape situation. \u003cem\u003eThe European Journal of Psychology Applied to Legal Context\u003c/em\u003e,\u003cem\u003e\u0026nbsp;6\u003c/em\u003e(2), 63-67. https://doi.org/10.1016/j.ejpal.2014.06.002\u003c/li\u003e\n \u003cli\u003eSutton, T. M., Herbert, A. M., \u0026amp; Clark, D. Q. (2019). Valence, arousal, and dominance ratings for facial stimuli. \u003cem\u003eQuarterly Journal of Experimental Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;72\u003c/em\u003e(8), 2046-2055.\u003c/li\u003e\n \u003cli\u003eTassinari, M., Aulbach, M. B., \u0026amp; Jasinskaja-Lahti, I. (2022). The use of virtual reality in studying prejudice and its reduction: A systematic review [Review]. \u003cem\u003ePLOS ONE\u003c/em\u003e,\u003cem\u003e\u0026nbsp;17\u003c/em\u003e(7 July), Article e0270748. https://doi.org/10.1371/journal.pone.0270748\u003c/li\u003e\n \u003cli\u003eTED. (2015, April 22). \u003cem\u003eChris Milk: How virtual reality can create the ultimate empathy machine\u003c/em\u003e [Video]. YouTube. https://www.youtube.com/watch?v=iXHil1TPxvA\u003c/li\u003e\n \u003cli\u003eThornberg, R. (2013). Bystander behavior in bullying situations: Basic moral sensitivity, moral disengagement and defender self-efficacy. \u003cem\u003eJournal of Adolescence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;36\u003c/em\u003e(3), 475-484. https://doi.org/10.1016/j.adolescence.2013.02.003\u003c/li\u003e\n \u003cli\u003eUllman, S. E. (2007). A 10-year update of \u0026ldquo;Review and critique of empirical studies of rape avoidance\u0026rdquo;. \u003cem\u003eCriminal Justice and Behavior\u003c/em\u003e,\u003cem\u003e\u0026nbsp;34\u003c/em\u003e(3), 411-429. https://doi.org/10.1177/0093854806297117\u003c/li\u003e\n \u003cli\u003eVan Gelder, J.-L., Nee, C., Otte, M., Demetriou, A., Van Sintemaartensdijk, I., \u0026amp; Van Prooijen, J.-W. (2017). Virtual Burglary: Exploring the Potential of Virtual Reality to Study Burglary in Action. \u003cem\u003eJournal of Research in Crime and Delinquency\u003c/em\u003e,\u003cem\u003e\u0026nbsp;54\u003c/em\u003e(1), 29-62. https://doi.org/10.1177/0022427816663997\u003c/li\u003e\n \u003cli\u003eVentura, S., Cardenas, G., Miragall, M., Riva, G., \u0026amp; Ba\u0026ntilde;os, R. (2021). How Does It Feel to Be a Woman Victim of Sexual Harassment? The Effect of 360\u0026deg;-Video-Based Virtual Reality on Empathy and Related Variables. \u003cem\u003eCyberpsychology, Behavior and Social Networking\u003c/em\u003e,\u003cem\u003e\u0026nbsp;24\u003c/em\u003e(4), 258-266. https://doi.org/10.1089/cyber.2020.0209\u003c/li\u003e\n \u003cli\u003eVonderhaar, R. L. (2015). There are no \u0026ldquo;innocent victims\u0026rdquo;: The influence of just world beliefs and prior victimization on rape myth acceptance. \u003cem\u003eJournal of Interpersonal Violence\u003c/em\u003e,\u003cem\u003e\u0026nbsp;30\u003c/em\u003e(10), 1615-1633. https://doi.org/10.1177/0886260514549196\u003c/li\u003e\n \u003cli\u003eWalster, E., Berscheid, E., \u0026amp; Walster, G. W. (1973). New directions in equity research. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;25\u003c/em\u003e(2), 151-176. https://doi.org/10.1037/h0033967\u003c/li\u003e\n \u003cli\u003eWalters, J. T. R., Bisson, J. I., \u0026amp; Shepherd, J. P. (2007). Predicting post-traumatic stress disorder: validation of the Trauma Screening Questionnaire in victims of assault. \u003cem\u003ePsychological Medicine\u003c/em\u003e,\u003cem\u003e\u0026nbsp;37\u003c/em\u003e(1), 143-150. https://doi.org/10.1017/s0033291706008658\u003c/li\u003e\n \u003cli\u003eWang, S. (2020). Standing up or standing by: Bystander intervention in cyberbullying on social media. \u003cem\u003eNew Media \u0026amp; Society\u003c/em\u003e,\u003cem\u003e\u0026nbsp;23\u003c/em\u003e(6), 1379-1397. https://doi.org/10.1177/1461444820902541\u003c/li\u003e\n \u003cli\u003eWarner, R. H., Vandeursen, M. J., \u0026amp; Pope, A. R. D. (2012). Temporal distance as a determinant of just world strategy: Mental Time Travel: Social Psychological Perspectives on a Fundamental Human Capacity. \u003cem\u003eEuropean Journal of Social Psychology\u003c/em\u003e,\u003cem\u003e\u0026nbsp;42\u003c/em\u003e(3), 276-284.\u003c/li\u003e\n \u003cli\u003eWhite, K., MacDonnell, R., \u0026amp; Ellard, J. H. (2012). Belief in a just world: Consumer intentions and behaviors toward ethical products. \u003cem\u003eJournal of Marketing\u003c/em\u003e,\u003cem\u003e\u0026nbsp;76\u003c/em\u003e(1), 103-119. https://doi.org/10.1509/jm.09.0581\u003c/li\u003e\n \u003cli\u003eZhao, Y., Chu, X., \u0026amp; Rong, K. (2023). Cyberbullying experience and bystander behavior in cyberbullying incidents: The serial mediating roles of perceived incident severity and empathy [Article]. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e,\u003cem\u003e\u0026nbsp;138\u003c/em\u003e, Article 107484. https://doi.org/10.1016/j.chb.2022.107484\u003c/li\u003e\n \u003cli\u003eZilkens, R. R., Smith, D. A., Kelly, M. C., Mukhtar, S. A., Semmens, J. B., \u0026amp; Phillips, M. A. (2017). Sexual assault and general body injuries: A detailed cross-sectional Australian study of 1163 women. \u003cem\u003eForensic Science International\u003c/em\u003e,\u003cem\u003e\u0026nbsp;279\u003c/em\u003e, 112-120. https://doi.org/https://doi.org/10.1016/j.forsciint.2017.08.001\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 7 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University of Bath","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"sexual assault, bystander intervention, virtual reality, gender-based violence, victim blaming. ","lastPublishedDoi":"10.21203/rs.3.rs-6008933/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6008933/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper explores how the need to believe in a just world (NBJW) affects bystanders’ intentions to intervene when witnessing a sexual assault. More specifically, we explored how non-rational strategies used to resolve a threat to the belief in a just world (BJW) are related to Latané and Darley’s (1970) bystander decision-making process, and whether empathic variables mediate this. In Study 1, the threat to the BJW was indirectly manipulated by varying the sexual assault severity described in a text vignette. Measures assessing victim derogation, blame, psychological distancing from the victim, empathy, empathic concern and the bystander process were completed by 294 participants. The assault severity had no significant effect on BJW strategies, but increased intervention intentions. There was a lack of evidence for the role of victim derogation and blaming, but greater psychological distancing indirectly predicted lower general intervention intentions via reduced empathic concern, and a reduced ability to succeed in several of the bystander stages, especially accepting responsibility to intervene. We replicated this design in Study 2 (\u003cem\u003eN\u003c/em\u003e = 117), but using filmed virtual reality (VR) stimuli. We again found a lack of effect of the severity manipulation beyond increasing the likelihood of intending to intervene. In Study 2, we also found that victim blaming predicted general intentions, but with no clear mediators. Implications are discussed, including how factors like the emotional impact of stimuli and the empathic nature of VR could help to explain some of the differing results found across our studies.\u003c/p\u003e","manuscriptTitle":"Sexual Assault in a ‘Just World’: An Immersive VR Study of Bystanders","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-24 13:34:15","doi":"10.21203/rs.3.rs-6008933/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"20c058f9-482d-4ce2-ae0e-734cf3ed9bdd","owner":[],"postedDate":"February 24th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":44183830,"name":"Psychology"}],"tags":[],"updatedAt":"2025-02-24T13:34:16+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-24 13:34:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6008933","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6008933","identity":"rs-6008933","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}