The Role of Sexual Consent and Past Non-consensual Sexual Experiences on Rape Supportive Attitudes in a Heterosexual Community Sample

The present study aimed to investigate the relationship between (a) sociodemographic variables, past non-consensual experiences and sexual consent and (b) the endorsement of rape supportive attitudes. A sample of 1042 heterosexual adults (42.4% men) was surveyed. A predictive model indicated that the most relevant predictors of rape supportive attitudes were gender (i.e., being male) and most dimensions from sexual consent such as (Lack of) perceived behavioural control, Sexual consent norms, and Awareness and discussion. On the contrary, holding a positive attitude for the establishment of consent in sexual encounters negatively predicted rape supportive attitudes. The positive association between gender (i.e., being male) and (Lack of) perceived behavioural control was stronger among participants who were not victims of past non-consensual sexual experiences than among participants who were victims of past non-consensual sexual experiences, and the same held true when Positive attitude toward establishing sexual consent was involved in the interaction. These findings may help us understand the role of sexual consent on the attitudes that support violence, which can better guide future prevention programs aimed at reducing them.


Introduction
Sexual consent can be defined as the freely given clear communication of a feeling of willingness to engage in sexual activity with a particular person in a particular context (Hickman & Muehlenhard, 1999;. In this framework, sexual activity can only continue if both partners express active interest (Javidi et al., 2022). Thus, sexual consent is a form of sexual communication that has recently acquired a tremendous practical importance in everyday sexual interactions, even as a legal indicator to determine the existence of sexual violence (Humphreys & Brousseau, 2010).
Several studies have investigated sexual consent signalling and people's actual behaviour in communicating consent (King et al., 2021;Muehlenhard et al., 2016;Newstrom et al., 2021). Research has shown that both women and men tend to signal consent nonverbally more than verbally and implicitly more than explicitly (Groggel et al., 2021;Humphreys, 2007;Willis et al., , 2021. For example, both men and women tend to signal consent nonverbally by physically approaching, kissing, touching intimately or not resisting their partner's advances (Willis & Jozkowski, 2022). Young adults also tend to communicate consent indirectly and passively Orchowski et al., 2022;Shumlich & Fisher, 2020). However, research has shown that explicit verbal cues of consent are most accurately interpreted by a partner compared to other types of consent cues, including implicit and nonverbal (Willis et al., 2021). For this reason, sexual consent is considered an important mechanism for reducing sexual assault and coercion (Shumlich & Fisher, 2020).

The Role of Gender and Past Non-consensual Sexual Experiences
Research from a psychological perspective has focused mainly on the individual and interpersonal correlates of sexual consent. For instance, Humphreys and Herold (2007) investigated the impact of gender in a heterosexual population. They found that heterosexual women considered it more important that consent be given regardless of relationship length and that consent be an ongoing process during the sexual encounters rather than a one-time event. In contrast, heterosexual men were more likely to hold the view that consent from a partner can be assumed without direct verbal communication, that consent only needs to be given once at the beginning of sexual activity and that the need for consent decreases as the length of the relationship increases (Humphreys, 2007;Humphreys & Herold, 2007). Other research has shown that, although men and women do not differ in their conceptions of sexual consent, women place more importance on this issue than men (Graf & Johnson, 2021).
Sexual precedents and past non-consensual sexual experiences may also impact one's attitudes, beliefs and behaviours about consent . According to Kilimnik and Humphreys (2018), women's previous sexual experiences may influence their understanding of consent and the ways in which they negotiate sexual consent. A history of non-consensual experiences plays a role in 1 3 consent relevant constructs, such as sexual risk perception and negotiation (Marx & Soler-Baillo, 2005), endorsement of rape myths and risk for re-victimization (Peterson & Muehlenhard, 2011). This is consistent with research on sexual communication strategies, which shows that individuals with non-consensual sexual experiences have less effective strategies for exiting risky dating and social situations compared to individuals without such experiences (Nason & Yeater, 2012).

Sexual Consent and Sexual Violence
Although most research views sexual consent as an individual and interpersonal issue, attitudes and behaviours on sexual consent may be related to cultural and social factors. Indeed, sexual consent is associated with traditional sexual scripts, i.e., men as initiators of sex and women as sexual gatekeepers (Hirsch et al., 2019). Men are typically assigned the role of actively seeking activity and sexual consent, while women are typically assigned the role associated with the responsibility to give or withhold consent, and thus determine whether or not sexual activity occurs (Jozkowski & Peterson, 2013). The belief, among many adults and college students, that men are socially and sexually dominant leads to double standards and men to assume consent in the absence of clear signals otherwise (Jozkowski et al., 2017). Studies have found that many men indeed stop sexual advances only when there is a clear and explicit (usually verbal) signal to stop (Goodcase et al., 2021;Marcantonio et al., 2018), which often leads to unwanted sexual behaviours (Orchowski et al., 2018). Sexual assault, then, is undoubtedly tied to sexual consent, especially as sexual assault is defined as sexual acts without consent (Jozkowski & Peterson, 2013).
This can have an impact not only on an individual level, but also on social perceptions of sexual violence and rape. Some studies have found that college students who had greater knowledge of sexual consent issues reported higher rates for bystander helping intentions and behaviour in cases of sexual violence than those with less understanding (McMahon et al., 2019;Moschella-Smith et al., 2022). Similarly, Orchoswki et al. (2022) examined bystander attitudes among young adult male active-duty soldiers and found that those who placed more importance on obtaining verbal sexual consent from a partner showed more favourable intentions to intervene when violence occurred. In addition, the intentions to respect communication of consent appear to be negatively related to rape myths, i.e., beliefs that serve to deny and justify men's sexual aggression against women (McMahon & Farmer, 2011;Shafer et al., 2018).
In this sense, sexual consent may be one of the key factors influencing the perception of violence. According to Willie and Kershaw (2019), violence against women is not a merely individual issue, and a social-ecological perspective can provide a comprehensive framework to better understand and address it. If some personal variables may be risk factors at the individual and relational level (e.g., CDC, 2019; Conley et al., 2017;Kelley et al, 2016;Tyler et al., 2017), some social attitudes may be conceived as risk factors at the cultural level. Indeed, a large body of research has used the gender stereotype framework to investigate how cultural values and gender stereotypes may affect the perception of violence 1 3 The Role of Sexual Consent and Past Non-consensual Sexual… against women (e.g., CDC, 2019; Rollero, 2020;Rollero & De Piccoli, 2020;Saucier et al., 2015;Yapp & Quayle, 2018). Gender-role stereotypical attitudes may influence the recognition of some behaviours, such as restriction of freedom and emotional abuse, as actual forms of violence against women (Rollero et al., 2021). Moreover, people who hold sexist attitudes are more likely to blame the victim of a sexual assault than women and men who hold more non-traditional views (e.g., Canto et al., 2014;Chapleau et al., 2007;Moyano et al., 2017;Rollero & Tartaglia, 2019).
Consistent with this framework, we argue that attitudes and behaviours related to sexual consent may be associated with rape supportive attitudes. Indeed, being less concerned with seeking and receiving sexual consent prior to sexual encounters may be associated with more positive attitudes toward rape (Warren et al., 2015), not only at the interpersonal level, but also in relation to the social perception of rape. Based on this reasoning, the purpose of our study was to extend past research on sexual consent by investigating its role on the social perception of rape. To our knowledge, previous literature has considered sexual consent as mainly an individual or interpersonal issue, and its association with attitudes toward violence has not been tested yet.

The Present Study
In this study we analysed the relation among gender, involvement in past nonconsensual sexual experiences, beliefs, attitudes and behaviours concerning sexual consent, and justification of violence, as measured by rape-supportive attitudes. Specifically, based on the literature above, we tested a moderated-mediated predictive model aimed at examining the following hypotheses: H1 Compared to women, men would show stronger rape supportive attitudes.
H2 Involvement in past non-consensual experiences (as victims) would be associated with lower rape supportive attitudes.
H3 Adaptive sexual consent attitudes, beliefs and behaviours related to sexual consent would be negatively related to rape supportive attitudes. Therefore, a low (Lack of) perceived behavioural control, a strong Positive attitude towards establishing consent, a low Indirect behavioural approach, strong Norms of sexual consent and a strong Awareness of consent would be positively associated with lower rape supportive attitudes.
H4 Regarding the likely mediating/moderating role of sexual consent between the sociodemographic variables and rape supportive attitudes, we expected that attitudes, beliefs and behaviours related to sexual consent may be influenced by gender and past non-consensual sexual experiences (Graf & Johnson, 2021;Kilimnik & Humphreys, 2018;.

Participants
We surveyed 1226 Italian adults. We excluded 182 who declared a non-exclusive heterosexual sexual orientation, as each group was too small to be included in the analysis, or they did not answer the question. Thus, we performed our analyses on a dataset composed of 1042 participants (men = 42.4%, women = 57.6%, no participants declared themselves as non-binary/trans, M age = 30.54, SD = 11.60, min = 18, max = 60). Among them, 44% referred to have been victims of nonconsensual sexual experiences (e.g., unwanted touching or stroking).

Past Non-consensual Sexual Experiences
We measured involvement in past non-consensual sexual experiences using a dummy variable (1 = yes, 0 = no). Based on the Sexual Experiences Survey (SES; Koss & Gidycz, 1985), participants were asked if they have ever had unwanted sexual experiences.

Consent Attitudes, Beliefs and Behaviours
We used an Italian adaptation of Humphreys and Brousseau's (2010) Sexual Consent Scale-Revised (SCSR) that provides information regarding attitudes, beliefs and behaviours related to sexual consent. In its original version, the scale is composed of 39 7-category Likert items Humphreys and Brousseau's (2010) analyses showed that SCSR has a 5-dimension factorial structure: (a) (Lack of) perceived behavioural control (measured by items such as, 'I think that verbally asking for sexual consent is awkward' and 'I am worried that my partner might think I'm weird or strange if I asked for sexual consent before starting any sexual activity'), (b) Positive attitude toward establishing consent (measured by items such as, 'I feel that sexual consent should always be obtained before the start of any sexual activity' and 'I believe that asking for sexual consent is in my best interest because it reduces any misinterpretations that might arise'), (c) Indirect behavioural approach to consent (measured by item such as, 'Typically I communicate sexual consent to my partner using nonverbal signals and body language' and 'It is easy to accurately read my current (or most recent) partner's nonverbal signals as indicating consent or non-consent to sexual activity'), (d) Sexual consent norms (measured by item such as, 'I believe that sexual intercourse is the only sexual activity that requires explicit verbal consent' and 'If consent for sexual intercourse is established, petting and fondling can be assumed'), and (e) Awareness and discussion (measured by item such as, 'I have discussed sexual consent issues with my current (or most recent) partner at times other than during sexual encounters' and 'I have discussed sexual consent issues with a friend').

3
The Role of Sexual Consent and Past Non-consensual Sexual… After a series of Confirmatory Factor Analyses (CFAs) (see Table S1 in the Supplementary material), based on the combination of theoretical reasoning and of the inspection of the modification indices the software provided, we discarded 24 of the scale's items, in that their inclusion would have led to an unsatisfactory fit and/or it would have led to a different number of items measuring the five factors of the scale. The remaining 15 items (reported in Table S2 in the Supplementary material) showed the expected 5-factor structure. The fit of the model was good, CFI = 0.97, RMSEA = 0.0 (90% CI 0.03-0.05). Using CFA, we computed five SCSR scores as factorial scores of the scale. Cronbach´s alpha values ranged from 0.59 to 0.82.

Rape Supportive Attitudes
We used an Italian adaptation of Lottes' (1991) Rape Supportive Attitude Scale (RSAS). In its original version, the scale is composed of 20 5-category Likert items, such as 'Being roughed up is sexually stimulating to some women' and 'If a girl engages in making out or petting and she lets things get out of hand, it is her fault if her partner forces sex on her', and has a unidimensional structure. After a series CFAs (see Table S3 in the Supplementary material), we deleted 5 items to maximize the fit of the factorial solution. The resulting 15-item battery (reported in Table S4 in the Supplementary material) showed a unidimensional structure and a satisfactory fit, CFI = 0.92, RMSEA = 0.06 (90% CI 0.06-0.07). Using CFA, we computed one RSAS score, as the factorial score of the scale (α = 0.82).

Sociodemographic and Control Variables
We used gender (dummy variable: 1 = men, 0 = women) and past non-consensual sexual experiences (dummy variable: 1 = yes, 0 = no). Moreover, in our analyses we controlled for participants' age and years of education. Table 1 reports the descriptive statistics for our variables and the bivariate correlations between them.

Procedure
Participants were recruited using a snowball sampling procedure, begun with postings by the researchers and their students in several social networks, email groups and web pages in which information about this research was provided. The link from the posting took participants to a secure, anonymous online questionnaire where they read an informed consent before beginning the study. Participants were also informed that their participation was voluntary and that they could discontinue the study at any time. The questionnaire took approximately 20 min to complete. No compensation was given for their enrolment. The Ethics Committee of the University of Torino approved the study protocol (protocol number: MAR.20/11.PRY).  The Role of Sexual Consent and Past Non-consensual Sexual…

Data Analyses
Using MPLUS, we resorted to a structural equation modelling approach, combining CFA (the SCSR and the RSAS were measured as latent variables) and a path analysis aimed to test the hypotheses above (see Fig. 1). As the CFAs described above, this predictive model was estimated using the Maximum Likelihood method. We evaluated the fit of all of our models via the Comparative Fit Index (CFI) and the Root Mean Square Error of Approximation (RMSEA). Based on Xia and Yang (2019), we considered the models to be satisfactory with a CFI value > 0.90 and an RMSEA value < 0.08. Table 2 shows the results of the moderated-mediated structural equations model aimed at testing our hypotheses. The first 10 columns report the results for the left portion of the model, aimed at predicting the five SCSR factors, modelled as latent variables. Among the control variables, age showed a positive association with (Lack of) perceived behavioural control and a negative association with Awareness and discussion, while education was positively associated with Positive attitude toward establishing consent and negatively associated with Indirect behavioural approach to consent and Sexual consent norms. More interestingly as concerns our goals, substantially consistent with our hypotheses, being male had a positive association with (Lack of) perceived behavioural control and with Sexual consent norms and a negative association with Positive attitude toward establishing consent. On the other hand, past non-consensual sexual experience Fig. 1 The model we tested  The Role of Sexual Consent and Past Non-consensual Sexual… had a positive association with Awareness and discussion and a negative association with Indirect behavioural approach to consent. Having been victim of non-consensual sexual experiences moderated the associations between being male and (Lack of) perceived behavioural control (see Fig. 2). A simple slope analysis showed that the being male-(Lack of) perceived behavioural control positive association was stronger among participants who were not victims of past non-consensual sexual experiences, simple slope = 0.97, SE = 0.15, than among those who experienced unwanted sexual contacts, simple slope = 0.58, SE = 0.12. The difference between the two slopes was statistically significant, t(1006) = 2.03, p = 0.04.

Results
Having been victim of non-consensual sexual experiences moderated also the association being male and a Positive attitude toward establishing sexual consent (see Fig. 3). The negative association between these variables was stronger among those who did not have unwanted sexual contacts, simple slope = − 1.02, SE = 0.12, than among those who had unwanted sexual contacts, simple slope = − 0.67, SE = 0.10. The difference between the two slopes was statistically significant, t(1006) = 2.24, p = 0.03. Table 2's last 2 columns report the parameters for the right portion of the model, aimed at predicting RSAS, modelled as latent variable. Being male showed a positive association with RSAS, while age, years of education and past non-consensual sexual experiences were not associated with it. More interestingly as concerns our goals, consistent with our hypotheses, RSAS showed a positive association with (Lack of) perceived behavioural control, Sexual consent norms, and Awareness and discussion and a negative association with Positive attitude toward establishing consent. Its association with Indirect behavioural approach to consent was not significant. The indirect association between being a man and RSAS was positive and significant, indirect effect = 0.12, SE = 0.02, p < 0.001.

Discussion
The present study aimed to examine the relationships between sexual consent, conceptualised as a five-multidimensional construct encompassing attitudes, beliefs and behaviours, and rape supportive attitudes in a community sample of heterosexual men and women. The predictive role of gender, past non-consensual sexual experiences and of their interaction was also considered, while age and education were statistically partialled out. Overall, some dimensions of sexual consent showed a positive association with being male, especially among participants who had been victims of non-consensual sexual experiences in the past. In addition, having difficulties to ask for consent, less positive attitudes towards establishing consent, endorsing certain sexual consent norms and greater awareness and discussion about sexual consent were predictors of higher rape supportive attitudes.
The first relevant finding from our study is that gender plays a role on sexual consent. In particular, men, more than women, indicate higher (Lack of) perceived behavioural control and more Sexual consent norms. That is, they have more difficulties for establishing consent prior sex and hold more sexual norms in relation to consent. Our findings are in line with traditional gender roles in which men are those who should initiate sex, while women should reject sexual activity, even if they want it, while men insist or ignore women's refusal . This is consistent with recent results showing that a sexual double standard still exists and influences whether or not and how we communicate within a sexual context (Gómez Berrocal et al., 2019;Moyano et al., 2022;Sánchez-Fuentes et al., 2020). In addition, an interaction between being male and previous non-consensual experiences was found for both (Lack of) perceived behavioural control and Positive attitude for establishing consent. These results indicate that those men who were victims of violence will be more willing to establish sexual consent, in contrast to those who did not. This greater favourable attitude toward negotiating consent has been recently The Role of Sexual Consent and Past Non-consensual Sexual… shown among those who reported to have been sexual victims (Moyano et al., 2022) which emphasizes the connection between sexual violence and sexual consent (Rousseau et al., 2020).
A second relevant finding is that participants who experienced non-consensual sexual experiences also indicate being more aware of this topic. In addition, when they approach consent in their relationships they are less likely to display indirect behavioural signalling, instead they would use more explicit or direct strategies, such as verbal manifestations. This contradicts previous results from Kilimnik and Humphreys (2018) in undergraduate women with a history of non-sexual consent, who reported fewer positive attitudes toward obtaining consent and more indirect approaches to express consent, while other studies did not find assault history as significant (McKenna et al., 2021). Instead, our study emphasizes that victims increase their awareness after non-consensual activities, which is consistent with a previous study from Spain in which sexual victims, in comparison to non-sexual victims, consider important the establishment of consent (Moyano et al., 2022). Furthermore, victims would not easily assume consent without direct verbal communication, and it should be obtained explicitly to reduce ambiguity (Humphreys & Brousseau, 2010).
A final relevant finding is that the variables that better predict rape supportive attitudes are the following: being male, lower self-efficacy to ask for consent, negative attitudes toward establishing consent, endorsement of more sexual consent norms and greater awareness about consent. Therefore, most attitudes, beliefs and behaviours related to how sex is negotiated are linked to attitudes about rape. This association is of great relevance, given that, as Moyano et al. (2017) have previously shown, supportive attitudes toward rape are related to violent or aggressive behaviours. In their study, Moyano and colleagues showed that male aggressors have more supportive attitudes toward rape than non-aggressors. As hypothesized, in our study, gender differences have also emerged. Similarly, Camp et al. (2018) found that women, compared to men, endorse more positive attitudes towards explicit consent. On the contrary, hypermasculine attitudes and acceptance of the rape myth have been related to unfavourable attitudes towards communication about sexual consent (Shafer et al., 2018). Moreover, men who are more exposed to sexualized women also accept more myths about rape and myths of sexual abuse, gender stereotypes and interpersonal violence (Sáez-Díaz, 2016). Furthermore, undesired explicit advances that are supported by sexual scripts would lead to greater invisibility and refusal of consent. Taken together, maintaining a culture that supports certain ideas associated with the perpetuation of sexual violence behaviours reduces the responsibility for seeking sexual consent (Argiero et al., 2010;Kilimnik & Humphreys, 2018).
Our study adds support to the multidimensionality of sexual consent, by distinguishing between beliefs, attitudes and behaviours. Therefore, our findings show that having certain knowledge about sexual consent, discussing about it does not necessarily lead to the endorsement of being more favourable to consent. On the contrary, behavioural control and positive attitudes do mean being favourable to consent, as they are probably core factors of consents, showing a stronger connection with particular sexual behaviours.

Limitations and Future Directions
There are some limitations that should be noted. First, as it often happens in social-psychological research, a convenience method for sampling was used and, therefore, our findings cannot be generalized to the general Italian population. Second, our study is cross-sectional, thus the direction of the relationships among the examined variables cannot be established. Longitudinal studies could help us to better understand this phenomenon and its changes across life span. Third, we used a self-report method. Thus, we have no actual information about participants' past sexual experiences. These limitations are typical of the large majority of the studies performed in this field. Fourth, when fitting the RSAS and, especially, the SCSR we had to discard some items. This was one of the first times these scales have been used in the Italian context. Moreover, our analyses were much more demanding that the original ones. A deep methodological study could be conducted in order to explore whether this was a problem of the Italian adaptation or any other intrinsic problem of the scales. Therefore, it could be interesting to conduct a replication study in order to provide support to the validity of our scales. Finally, due to the skewed distribution of our participants on the sexual orientation items, we have to discard the non-heterosexual participants. A replication of this study performed in more diverse samples and with sexual minorities could be interesting. In addition, future studies could explore how each particular dimension of consent, beliefs, attitudes and behaviours are differently related to particular sexual aspects.
Nonetheless, on the positive side, in contrast to previous research in which college student samples were mostly used, our study provides an analysis of the phenomenon of sexually consent in a heterogeneous sample from different ages and in which men and women were equally distributed.

Conclusion
The present study moves forward to approach the question of how some attitudes, beliefs and behaviours related to sexual consent are a predictor of key attitudes to understand sexual violence, particularly rape supportive attitudes among heterosexual men and women. In this sense, we would like to highlight three main contributions of our present work: First, men have more difficulty than women in asking for and establishing sexual consent, which is a relevant predictor of attitudes toward rape. Prevention programmes should consider this as a key issue and train men and women in their communication strategies and assertiveness when negotiating sex, mitigating some sexual scripts and masculinity standards. Further research could get deep into men´s ideas associated to sexual consent request, as they are likely to feel vulnerable when asking for it. Second, previous non-consensual experiences are a key aspect that makes victims more aware and sensitive to the issue of consent, leading them to show more explicit and direct strategies when it comes to negotiating sex. Third, there is still a culture that emphasises the persistence of a social dominance of men over women. This is noteworthy, given that rape supportive attitudes are crucial to understanding a culture based on violence. Several implications can be drawn for sexual violence prevention education, related to emphasising the need for a culture of consent, even beyond young people and college students. This is necessary not only at the individual level, but also at the social level, as the consent culture can impact on victim blaming and its related cultural processes in case of rape.