Blame and responsibility in the context of youth produced sexual imagery: The role of teacher empathy and rape myth acceptance and Teacher Education

(cid:1) Teachers felt more responsible when confronted with the non-consensual sharing of girls ’ semi-nude pictures. (cid:1) Teachers tended to blame the person whose semi-nude pictures were shared in an attention seeking vignette. (cid:1) Lower empathy and higher rape myth acceptance were related to teachers ’ attribution of blame. This study examined Pre-Service Teachers ’ (PSTs; N ¼ 92; 75% women, 25% men) tendency to blame students or to feel responsibility when confronted with different vignettes of the sharing of youth produced sexual images. Self-report measures of empathy and rape myth acceptance were collected. Findings showed that PSTs felt more responsible for girl targets depending on the vignette type. Moreover, they blamed the target more in the attention seeking vignette compared to other vignette types. Finally, PSTs who tended to blame the target showed lower levels of empathic concern and higher levels of rape myth acceptance. Implications for teaching education are discussed.


Introduction
With the rise of new technologies, the exchange of sexual material among teenagers has moved to the online space, creating a phenomenon called "sexting". This term refers to the sending, receiving and forwarding of texts, photos and videos of sexual content through electronic means (Klettke et al., 2014). In this paper, the term "youth-produced sexual images" will be used, building on previous work by Wolak and Finkelhor (2011), to refer to the exchange of sexual images that portray semi-nude teenagers that are taken by young people themselves. In brief, by this we mean photos that were initially 'self-produced' unless otherwise stated.
Based on an analysis of legal aspects of sexting in the United States, Wolak and Finkelhor (2011) categorised the sharing of youth-produced sexual images according to two different typologies: Experimental and Aggravated. 'Experimental' refers to youth taking sexual pictures of themselves and consensually sharing them, most commonly with a current or potential partner (Symons et al., 2018). Young people might engage in sharing their sexual pictures in order to gain the partner's attention, to prompt sexual and romantic interest, to find validation for their appearance, and/or to initiate sexual activity (Klettke et al., 2014;Symons et al., 2018): In other words, it is a way to maintain and increase closeness in a potential intimate relationship (Albury & Crawford, 2012;Hasinoff, 2015). In addition, consensually self-producing and sharing sexual pictures might allow young people to represent themselves in a sexual way, feel sexy and build their identity, thus enabling them to fulfill their need for self-expression Hasinoff, 2015). This behaviour might also contribute to the exploration of sexuality and sexual identity among LGBTQ þ youth. The private nature of sexting allows young LGBTQ þ individuals to express themselves freely with other known LGBTQ þ individuals as compared to the offline context, where they might be vulnerable, exposed and discriminated against because of their sexual orientation (Hertlein et al., 2015;Ybarra & Mitchell, 2014). Thus, consensual sharing of sexual images can be a fun and pleasurable way for young people to express their sexuality, and it can be conceived of as part of the normative adolescent affective and sexual development (Klettke et al., 2014;Symons et al., 2018;Patchin & Hinduja, 2019).
The second typology suggested by Wolak and Finkelhor (2011) is called Aggravated. The Aggravated typology refers to the criminal elements that are in some cases involved in the creation, sending or possession of youth produced sexual imagery. For instance, these additional elements could include: Adults or other youth coercing minors to send sexual images of themselves; criminal or abusive behaviour by minors such as sexual abuse; extortion, deception or threat; and/or the creation and sending of sexual images without the consent of the youth who were pictured. Using sexual imagery as a tool to harass the person depicted, threaten them, and have control over them is part of the wider spectrum of intimate partner violence (Henry & Powell, 2015). In other words, the nonconsensual sharing of sexual images (or the threat to share them) can be considered as a form of sexual violence (Maddocks, 2018;Walker & Sleath, 2017). This conduct has been associated with other types of sexual violence in adolescents, such as pressuring a partner to have sex even if they don't want to, or to have sex without a condom (Kernsmith et al., 2018), and more generally to physical, sexual and psychological violence within a relationship . The non-consensual sharing of sexual images has also been defined as digital or cyber rape highlighting how this type of online violence can mirror the impact of "offline" abusive behaviours (Bothamley & Tully, 2018;Maddocks, 2018).

Legal aspects
From a legal point of view, youth-produced sexual pictures could potentially be considered as child pornography in some jurisdictions. For example, in Ireland under the Child Trafficking and Pornography Act (1998) it is a crime to download naked or sexual images of someone under 17 years, and under the Non-Fatal Offences Act (1997) it is a crime to share someone else's intimate image without their consent with the intention of causing harm to that person. Furthermore, the recent Harassment, Harmful Communications and Related Offences Bill (2017) will soon make it illegal to take or share intimate images without consent. Similar crimes exist in the UK under Section 1 of the Protection of Children Act (1978) regarding those under 18 years. Thus, theoretically leading to penal consequences for the teenagers who own, possess or share intimate images of minors. Such laws do not take into consideration the context in which the sexual pictures were produced, whether it was consensual or not, and whether the pictures were shared only among minors or with an adult (Hasinoff, 2015). Therefore, they risk punishing the consensual creation and sharing of sexual images, thus criminalising young people's sexuality (Albury & Crawford, 2012;Hasinoff, 2015;Salter et al., 2013). This is paradoxical considering that minors above the age of consent but younger than 18 are deemed old enough to engage in sexual activity, but cannot legally create and share self-produced sexual pictures (Albury & Crawford, 2012;Salter et al., 2013). Moreover, in the case of non-consensual dissemination of sexual images, the young person depicted in the photos might be considered prosecutable for having produced child pornography: this creates a paradox where the person who is the target of the abuse is held as prosecutable as the person who committed it (Hasinoff, 2015). Worryingly, this approach could result in blaming the target, instead of protecting them (Albury & Crawford, 2012;Hasinoff, 2015).

Gender
Research shows that both boys and girls experience the same need to be popular, particularly among their peer group. This is evident in both online and offline contexts (Wright, 2018;Del Rey et al., 2019). However, the creation and sharing of sexual images seem to have a gendered nature that envisages specific roles for girls and boys. While there is little research about scripted roles in same-sex couples, in heterosexual relationships, it may be more socially expected for boys to be the ones who ask for sexual pictures and for girls to be the senders (Klettke et al., 2014). Sometimes girls can feel increased pressure to send nude or semi-nude images of themselves even within a consensual relationship (Ricciardelli & Adorjan, 2019). In addition, girls are sometimes expected not to be responsive to requests in order to protect their own reputation (Symons et al., 2018). In fact, the literature shows that girls undergo higher criticism and "slut-shaming" than boys when they accept an invitation to send a sexual picture (Barrense-Dias et al., 2017). At the same time, the sharing of sexual images is more often normalised among boys, both heterosexual and homosexual: they are less at risk of public humiliation and instead their masculinity is celebrated (Albury & Byron, 2014;Ricciardelli & Adorjan, 2019). Even when girls are the targets of non-consensual dissemination of sexual pictures, the blame is sometimes placed on them as they are perceived as the ones who should have prevented the incident by not sharing the picture in the first place (Hadwin, 2017;Hasinoff, 2015). This logic is often applied to targets of other types of sexual violence as well, such as rape (Hadwin, 2017;Hasinoff, 2015). Gender differences can also exist in respect to how sexting is interpreted and socially constructed. For instance, boys have been reported as less concerned by public dissemination of nude pictures, while public dissemination may be more likely to have negative emotional consequences for girls (Ricciardelli & Adorjan, 2019;Del Rey et al., 2019).

Teachers' perceived responsibility
In the context of schools, sexting and/or the non-consensual sharing of images between young people is considered a form of bullying and is an issue that is becoming more and more of a concern for schools (The Key, 2015). Although these incidents often happen outside of school, students can also engage in sexting behaviour at school and the fall out and consequences in terms of changes to peer relationships, victimisation and mental health implications can filter into the school environment in much the same way as incidences of cyberbullying (Dobson & Ringrose, 2016;UK Council for Child Internet Safety, 2016). Similar to other forms of bullying, teachers do not always feel automatically responsible for helping students deal with such matters (Mazzone et al., in press). For instance, there is much evidence showing that school staff is divided on the issue of where their responsibility lies when dealing with student relationship problems that do not occur at school (Green et al., 2017). In some cases teachers feel they have a pastoral role in this regard (Mura et al., 2014), while in other cases they feel that punishing the students involved would mean "overstepping the school authority" (Young et al., 2017). Furthermore, when measures are taken to intervene, they run the risk of being ineffective and harmful. For example, schools may adopt a victimblaming approach in their sexting prevention campaign, to warn students about what the consequences they might incur when sending explicit pictures (as reported in Jørgensen et al., 2019), and avoid working with the perpetrators in trying to reduce the non-consensual sharing of images.

Teachers' attribution of blame and empathy
Evidence demonstrates that the teachers' role or perception of such incidences are complex and there may not be one straightforward opinion on the matter. In particular, a teacher's perception of a student who 'sexts' might be biased by the gender and the specific actions of the youth involved. For instance, some educational authorities have started campaigns to prevent the dissemination of sexual pictures by targeting girls and encouraging them to set limits to their peers' sexual behaviour (Jørgensen et al., 2019). Some researchers believe that this type of "risk management" campaign frame girls who are targets of non-consensual sexting as too provocative and as blameworthy for not having "thought better", as it often occurs with victims of rape (Albury & Crawford, 2012;Salter et al., 2013). Even in the case of consensual sharing of sexual images, these educational campaigns blame girls' willingness to engage in such activity, thus stigmatising their sexual expression rather than the non-consensual dissemination conducted by their peers (Karaian, 2013). A qualitative study with Canadian adolescents showed that teachers tend to discuss incidents of non-consensual sharing of youth-produced sexual images longer with girls compared to boys, which highlights the tendency to consider girls as more responsible and vulnerable to this issue (Ricciardelli & Adorjan, 2019). Besides stigmatising girls, this approach trivialises the experiences of boys, while neglecting that boys' images could be shared without their consent, and forcing them to suffer in silence in order not to violate existing masculine norms (Ricciardelli & Adorjan, 2019).
Teachers' sense of responsibility and their attribution of blame for targets of non-consensual sharing of sexual imagery could also be connected to some personal attributes, such as empathy. Empathy is typically defined as a construct encompassing a cognitive and an emotional aspect. The former consists of understanding the other person's experience by adopting their point of view; the latter implies concern for the other person and the experience of their emotions or of one's own feelings of distress (Hodges and Myers, 2007). A strong negative correlation between empathy and attribution of blame has been found in a variety of social situations. For instance, in cyberbullying incidents, lower levels of empathy correlate with higher attribution of blame towards the target, which also predicts less positive behaviours to support the target (Schacter et al., 2016). A similar association has been shown in rape cases where higher levels of empathy towards the target predicted less blame (Muller et al., 1994).

Rape Myth Acceptance
Given that the non-consensual sharing of sexual images can be placed on a continuum of sexual violence that includes rape as well (Maddocks, 2018), it can be assumed that false beliefs about rape play a role in attributing blame in such contexts. Rape myths involve justifying the perpetrator's actions by denying their responsibility and minimising the harm caused, and placing responsibility on the target by insinuating that either they provoked the rape or they faked it (Payne et al., 1999). In one study, participants displayed higher rape myth acceptance and attribution of blame towards a target of non-consensual dissemination of sexual pictures rather than towards a survivor of sexual assault (Hadwin, 2017). Moreover, people with high rape myth acceptance tend to minimise the unpleasantness of the non-consensual sharing for the depicted person, possibly because they consider them partially responsible of the incident (

Purpose of the present study
Taken the above literature into account, the present study had five main aims. The first was to investigate Pre-Service Teachers (PSTs) perceived sense of responsibility and their levels of attribution of blame when dealing with vignettes where their students were involved in the exchange of youth produced sexual images. The vignettes presented in this study were based on the typology illustrated by Wolak and Finkelhor (2011), which differentiated between experimental and aggravated exchanges of youth produced sexual images. More specifically, we were interested in determining whether feelings of responsibility and levels of blame would change depending on the type of imaginary vignette the PSTs were presented with. No specific hypothesis was formulated regarding this relationship as there were no previous studies investigating this specific question.
The second and third aim of the research were to examine how PSTs' perceived sense of responsibility and attribution of blame varied based on the gender of the target of the vignette and on the gender of the PSTs themselves. As such, we hypothesised that teachers would attribute more blame to female targets in the vignettes and that they would feel more responsible for female targets (H1), whereas no specific hypothesis was formulated with regards to the teachers' gender.
The fourth and fifth aim were to determine the role of PSTs' empathy and rape myth acceptance in their levels of blame and responsibility. We hypothesised that PSTs' empathy levels and rape myth acceptance would be related with their tendency to blame the target or to their feelings of responsibility. In particular, we expected that both perspective taking and empathic concern would be negatively associated with attribution of blame and positively related to perceived sense of responsibility (H2). As for rape myth acceptance, we expected it to be positively associated with attribution of blame and negatively associated with perceived sense of responsibility (H3).

Participants
The sample comprised of 92 Pre-Service Teachers (75% women and 25% men) attending the first and second year of a graduate post-primary teacher education programme at a University in Ireland. The age range of the sample was 21e53 years of age [M(SD): 26.07(6.40)]. The majority of the sample identified as heterosexual (85.9%), followed by bisexual (6.5%), asexual (2.2%), lesbian (2.2%) and homosexual (1.1%). Two participants marked their sexual orientation as "other" without specifying it.

Procedure
Participants were recruited from a two-year Professional Master of Education (PME) programme. All students in the PME programme (N ¼ 188) were invited to participate in the study. Data were collected between February and March 2019 in two separate sessions, so that approximately half of the participants completed the survey on the first day and the other half completed it on the second day. In both cases, participants were provided with the option to complete the questionnaire online (by following a link provided to them) or in hardcopy in a lecture setting. The first author was present in a lecture during the 15 min of the administration of the survey, to provide the student teachers with the instructions and to answer any questions they had.
This research was approved by the authors' university ethics committee. All participants received information about the study in advance and immediately before participation. Informed consent was obtained from all participants. The students were informed that their participation in the study was voluntary, and that they could withdraw from the study at any time.

Instruments
There were four sections to the survey used in this study. The first related to demographics, and asked participants about their age, gender, and sexual orientation. The second section included two standard validated measures to assess empathy and rape myth acceptance. The third section involved vignettes depicting various scenarios where youth-produced sexual imagery had been shared. The fourth section involved questions related to attribution of blame and feelings of responsibility in reaction to the vignettes. These instruments are described in more detail below.
Empathy. Two subscales of the Interpersonal Reactivity Index (IRI; Davis, 1983) were used to measure perspective taking (PT subscale) and empathic concern (EC subscale). Both sub-scales included seven statements and participants were asked to indicate how well these described them on a Likert scale ranging from 1 ¼ does not describe me well to 5 ¼ describes me very well. Sample items are "I try to look at everybody's side of a disagreement before I make a decision" (i.e. perspective taking) and "I am often quite touched by things that I see happen" (i.e. empathic concern). PT and EC scores were computed for each participant, by averaging their responses across all items (a ¼ 0.64. for perspective taking and a ¼ 67, for empathic concern). Rape myth acceptance. The 22-item version of the Illinois Rape Myth Acceptance scale (IRMA; McMahon & Farmer, 2011) was used to measure participants' agreement with rape myth statements. The updated version of the questionnaire, originally built by Payne, Lonsway and Fitzgerald (1999), includes four subscales. The first sub-scale ("she asked for it") consists of six items and measures the belief that the victim is responsible for provoking the perpetrator into raping her (e.g. "If a girl acts like a slut, eventually she is going to get into trouble"). The subscale named "he didn't mean to" includes six items and assesses the belief that it wasn't the assaulter's intention to rape (e.g. "If a guy is drunk, he might rape someone unintentionally"). The five items of the subscale "it wasn't really rape" measure the denial of the rape (e.g. "If a girl doesn't say "no" she can't claim rape"). Finally, the subscale "she lied" includes five items reflecting the belief that the victim invented the assault for a personal advantage (e.g. "Rape accusations are often used as a way of getting back at guys"). Participants' answers ranged from 1 ¼ strongly agree to 5 ¼ strongly disagree, with higher values indicating greater rejection of rape myths. Rape myth acceptance scores were computed for each participant, by averaging their responses across all items (a ¼ 93).
Vignettes. In order to assess participants' attitudes to youth who produce and share sexual imagery, participants were asked to read five vignettes. The vignettes were adapted from previous research by Wolak and Finkelhor (2011) and included the following categories (see Appendix A for full details): (1) aggravated incident, adult involved (i.e., when an adult asks for a sexual picture from a minor); (2) aggravated incident, youth with intent to harm (i.e., when a minor spreads a sexual photo of a peer (or threatens to do it) as an extortion or as vengeance for an interpersonal conflict); (3) aggravated incident, youth reckless misuse (i.e., when a minor's photo is taken and shared by another youth without consent but with no intent to harm); 4) experimental incident, romantic relationship (i.e. when the picture is consensually shared between two people who are in a relationship together); (5) experimental incident, attention seeking (i.e. when a sexual photo is consensually shared by the creator with one or multiple people, not in the context of a relationship, for a range of reasons that go from selfexpression to sexual attention). The sixth category listed by Wolak and Finkelhor (i.e., experimental incident, other; was not included as it was outside the scope of the current study. All vignettes, regardless of the category, involved one young person (the "target") as a central figure who first produced a sexual image of themselves but the reasons or consequences for such an action vary across the vignettes. The vignettes were presented in random order, as were the questions. Approximately half of all participants (n ¼ 47) read five vignettes where the main character or target was a girl, whereas for the other half (n ¼ 45) the target was always a boy. The gender of the other characters in the vignettes was not specified. Neutral pronouns (i.e. they/them) as well as the expression "boyfriend/girlfriend" were adopted to refer to them, so that the target's sexual orientation could not be assumed, thus not influencing participants' responses. It is important to note that some vignettes contained situations of self-produced sexual imagery where 'sexual imagery' related specifically to semi-nude photos that were non-consensually shared or seen by persons they were not intended for, whereas other vignettes were consensually shared.
Attribution of blame and feelings of responsibility. Following each vignette, participants were asked to express their agreement with six statements concerning: (1) attribution of blame for the targets in the vignettes (three items) and (2) their perceived responsibility for dealing with the incidents (three items; see Appendix B for description of all the items). The answers were given on a Likert scale ranging from 1 ¼ disagree strongly to 5 ¼ agree strongly. These were developed for the current study and based on two previous studies by Holfeld (2014) and Tangney, Dearing, Wagner and Gramzow (2000). Attribution of blame explored the blame put on the character of the vignette that created the picture, while perceived responsibility items evaluated to what degree participants felt responsible to deal with the situation illustrated in the vignette. Scores for attribution of blame and perceived responsibility were computed for each participant by averaging their responses across (a ¼ 0.89, for attribution of blame and a ¼ 0.90, for perceived responsibility).

Data analysis
The study adopted a cross-sectional design. The data was analysed using SPSS 24 version. Descriptive statistics and bivariate correlations were conducted for all measures. A three-way 5 Â 2 Â 2 mixed ANOVA was implemented to investigate any differences in teachers' attribution of blame and feelings of responsibility (outcome variables) in terms of vignette type (within-subject variable), participants' gender and target's gender (between-subjects variables). Pairwise comparisons with Bonferroni adjustment were used to interpret the main effects of the independent variables. When the assumption of sphericity was not met, the Greenhouse-Geisser correction was reported.
Hierarchical multiple regression analyses were performed to test the associations between participants' attitudes to youthproduced and shared sexual imagery vignettes (outcome variables), empathy and rape myth acceptance (predictor variables). Furthermore, two hierarchical regression models were performed respectively for perceived responsibility and attribution of blame as outcome variables. Given the wide age-range sampled in this study, age was entered in the first step as a control variable; perspective taking and empathic concern were entered in the second step, and rape myth acceptance was entered in the third step.

Descriptive analyses and bivariate correlations
Descriptive statistics for all study variables are provided in Table 1. The average ratings for all the study variables (except for attribution of blame) were above the midpoint, thus indicating high levels of empathic concern, perspective taking and perceived responsibility for the sample, as well as low rape myth acceptance and attribution of blame. Table 1 also displays the bivariate correlations between the study variables. There was a negative correlation between attribution of blame and the empathic concern subscale, and the IRMA score, meaning that higher levels of attribution of blame were correlated to lower empathic concern and lower rape myth rejection (i.e., higher rape myth acceptance). Perceived responsibility did not show any significant correlation with any of the variables, indicating no association between responsibility and attribution of blame, nor with being empathic or believing in rape myths. The IRMA score was unrelated to empathic concern but positively correlated with perspective taking, thus indicating that higher rape myth rejection (i.e., lower rape myth acceptance) was correlated to higher perspective taking.

Attribution of blame and perceived responsibility in terms of vignette type and gender (H1)
The results of the three-way mixed ANOVA for attribution of blame revealed only a significant main effect of the vignette type, F(2.76, 217.75) ¼ 14.23, p < .001. Analysis of the pairwise comparisons showed that participants reported significantly higher levels of attribution of blame in the attention seeking vignette (M ¼ 2.41, SE ¼ 0.15) compared to all the others: adult involved (M ¼ 1.74, SE ¼ 0.14); reckless misuse (M ¼ 1.55, SE ¼ 0.10); intention to harm (M ¼ 1.75, SE ¼ 0.14); romantic relationship (M ¼ 1.78, SE ¼ 0.14). Namely, participants tended to blame the target more when the self-produced sexual images were shared to seek attention, compared to other contexts.
With regards to teachers' perceived responsibility, the ANOVA outcome indicated a significant effect of the vignette type, F(3.58, 113.62) ¼ 4.50, p ¼ .002 as well as a significant effect of the interaction between vignette type and target's gender, F(3.58, 113.62) ¼ 2.63, p ¼ .04. In other words, the type of vignette and the target's gender were both related with PSTs' perceived responsibility.
Follow-up analyses of the interaction showed that there was no significant difference between the two groups of participants (boy target vs. girl target) within the same vignette. However, the levels of perceived responsibility did differ significantly among the vignettes within the group of participants who were assigned a girl target. Participants in this group reported significantly lower levels of perceived responsibility in the vignettes involving a romantic relationship (M ¼ 2.48, SE ¼ 0.15) compared to the intention to harm vignette (M ¼ 2.98, SE ¼ 0.17) and the reckless misuse vignette (M ¼ 2.93, SE ¼ 0.15). Moreover, the responses to the attention seeking vignette indicated significantly lower scores of perceived responsibility (M ¼ 2.31, SE ¼ 0.16) compared to the adult involved vignette (M ¼ 2.87, SE ¼ 0.14), the intention to harm vignette and the reckless misuse vignette. PSTs who read vignettes where the target was a girl felt more responsible to deal with the incident based on the type of vignettes. More specifically, participants felt more responsible to act when the non-consensual sharing of sexual images happened in an aggravated context, where the target was threatened by an adult, when the images were spread as revenge against the target, and when they were showed to others breaking a promise. Instead, participants felt less responsible to deal with vignettes where the sexual images where consensually shared, be it within a couple or not. No significant differences in perceived responsibility among the vignettes were found within the group who was assigned a boy target, meaning that PSTs' levels of perceived responsibility did not differ based on the vignette type when the target was a boy.

Empathy and rape acceptance as predictors of attribution of blame and perceived responsibility (H2 and H3)
Findings from the hierarchical multiple regression analysis showed that age was positively associated with attribution of blame (b ¼ 0.15; p ¼ .11; see Table 2); i.e., older participants tended to attribute more blame to the target. After controlling for age, the results showed that lower levels of empathic concern (b ¼ À0.21; p ¼ .03), and higher levels of rape myths acceptance (b ¼ À0.49; p < .001) were positively associated with participants' tendency to blame the target of the vignettes. The regression for perceived responsibility revealed no significant results, suggesting that empathy and rape myth acceptance did not predict the perceived responsibility scores. That is to say that being empathic and believing in rape myths were not related to feeling responsible to deal with the incidents.

Discussion
The overarching goal of the current study was to assess attribution of blame and feelings of responsibility that Pre-Service Teachers (PSTs) displayed when dealing with cases of youthproduced sexual imagery. We believe this is an important endeavor to get a feel for teachers' opinions on a sensitive topic that is creeping into schools across Ireland and internationally. We postulated that teachers' perceptions of vignettes involving youth produced sexual imagery (based on previous work by Wolak & Finkelhor, 2011), may be related to their own acceptance/rejection of rape myths and levels of empathy. In general, investigating  the teacher's viewpoint on such cases was a useful endeavor which may lend itself to the enhancement of existing prevention methods to reduce online sexual harassment and improve child protection measures. Findings from this study showed a positive correlation between attribution of blame and rape myth acceptance, indicating that PSTs who placed higher blame on the targets in the vignettes were also likely to score high on rape myth acceptance. This result is not surprising, as both these variables express a tendency to view the target as responsible for their victimisation (Grubb & Turner, 2012). Participants in this study displayed levels of empathic concern and perspective taking that were higher than the midpoint, thus indicating that they feel great levels of compassion towards people in distress and that they are particularly able to adopt the other's point of view. This result confirms previous work on Irish teachers (Murphy, Tubritt, & O'Higgins Norman, 2018). Indeed, in the present study a negative correlation was found between empathic concern and attribution of blame, indicating that empathic PSTs are less likely to blame the target in the vignettes. Such associations are documented in previous literature showing how participants with higher empathy tend to blame the target less, both in cyberbullying episodes (Schacter et al., 2016) and in rape cases (Muller et al., 1994). Similarly, higher perspective taking was found to be associated with lower rape myth acceptance, meaning that PSTs who more frequently adopt the other's perspective are less inclined to find rape myths acceptable. This finding confirms results from previous studies, which have reported that higher empathy towards a rape victim is associated with a lower acceptance of rape myths (Miller et al., 2011).
In contrast with H1, the current results show that PSTs' attribution of blame of the main character in the vignette was not affected by the gender of the target, nor by the PSTs' own gender. These findings could indicate that further factors (e.g., specific vignette and nuisances within them), rather than simply gender were more salient in judging where their responsibility lies. Indeed, we found that PSTs' tendency to blame the target or to feel responsible for dealing with the consequences differed according to the vignette type. PSTs tended to attribute more blame to the target in the vignette that referred to attention seeking compared to the reckless misuse, adult involvement, intention to harm and romantic vignettes. One explanation is that PSTs might consider this type of behaviour as provocative. In other words, they might believe that seeking attention through sexting and expressing oneself sexually outside of a relationship might lead to negative consequences, such as the further and unwanted dissemination of the sexual images (Bandura, 2002). In addition, it is possible that the PSTs blamed the target for not being careful enough to consider the possible negative consequences of their actions. Blaming the target could also serve to deny their own responsibility in educating students about online safety. In addition, teachers might not view the sharing of sexual images as a part of adolescent normal affective development, which could explain their inclination towards blaming teenagers for expressing themselves in such a manner. It is worth noting that both the romantic and the attention seeking vignettes belong to the same overarching typology, as they both involve the consensual sharing of youth-produced sexual images. Nevertheless, PSTs may blame the target more in the vignette that aimed to seek attention outside of a relationship rather than within one.
The findings did not show any differences in teachers' perceived sense of responsibility depending on the gender of the target, in contrast with H2. However, when the vignette's main character was a girl, participants reported different levels of responsibility depending on the type of vignette. Specifically, they felt more responsible for dealing with the scenario involving an adult as compared to the romantic and attention seeking vignettes.
Confirming our hypothesis (H2), there were no difference related to the target gender for attribution of blame, however significant differences were found for perceived responsibility. In the situation where an adult first seduces and then threatens a student to receive sexual pictures, it is likely that the PSTs perceived the target as dealing with someone more powerful than them and thus in need of their help.
Moreover, the PSTs felt more responsible for dealing with the reckless misuse and intention to harm vignettes (i.e. the aggravated typologies) compared to the attention seeking and romantic relationship vignettes (i.e. the experimental typologies). A possible explanation to this is that the PSTs detected the additional criminal element typical of the aggravated vignettes, which could make them perceive the target as more in danger and could make them feel more responsible for the safety of the student in the vignette. On the other hand, the exchange of self-produced sexual images could be considered as less harmful in the vignettes where the target sent pictures to get the attention from their partner or of other peers (i.e. experimental typology). Such behaviour could be seen as part of the private peer to peer relationships where both parties are equal in one sense and there may be less need for inference from an authority figure.
Such differences in perceived responsibility were observed only when the main character was a girl; instead, when the main character was a boy target, the levels of perceived responsibility did not differ depending on the consensual or non-consensual character of the vignette. PSTs could perceive as particularly serious the situations in which girls are the target of non-consensual sharing of sexual images, such as when an adult threatens to make public a girl's semi-nude picture, or when a peer shares such a picture with their classmates without the girl's consent. Indeed, in such contexts, the semi-nude pictures could potentially reach a wide audience against the girls' will (as opposed to when she consensually shares the picture with a romantic partner or to express herself). Possibly, PSTs realised that non-consensual public dissemination might be an issue particularly for girls rather than for boys. When a girl's sexual pictures go public, she is the one who risks being blamed rather than the person who shared them without her consent, because she is expected to resist boys' requests and not send sexual pictures in the first place ( Barrense-Dias et al., 2017). Therefore, when a girl breaks this gender role she is marked as a "slut" and ostracised (Ricciardelli & Adorjan, 2019;Ringrose et al., 2013), thus risking to experience internalising problems (Bates, 2017;Mori et al., 2019;Ringrose et al., 2013). On the contrary, much research suggests that boys' sending sexual pictures is normalised, and they might even gain admiration from their peers if their sexual pictures go public (Ricciardelli & Adorjan, 2019;Ringrose et al., 2013). Considering these gender differences, PSTs might feel more responsible to intervene when the sexual pictures of a teenage girl are publicly shared.
Findings of the regression analysis showed that attribution of blame was positively predicted by rape myth acceptance, in accordance with H3. In previous studies rape myths have been shown to influence the perception of the victim, at least in rape cases (Grubb & Turner, 2012;Russell & Hand, 2017). Such beliefs can minimise the responsibility put on the perpetrator and place the blame on the victim instead. Rape myths may be a product of the cognitive bias known as "just world theory" (Hayes et al., 2013). This concept posits that every event is a morally balanced consequence to a previous action: In other words, everyone gets what they deserve (Lerner & Matthews, 1967). In a similar way, rape myths insinuate that the victim did something to deserve the rape (e.g. they did not say "no" clearly), and therefore they should be blamed (Grubb & Turner, 2012). In the case of the exchange of sexual images, the just world belief intrinsic in rape myth acceptance may lead one to think that the target of the unwanted dissemination is actually to be held guilty for the incident because they did something to deserve it (e.g. they were too naive; Sakallı-Uǧurlu et al., 2007).

Strengths and limitations
Some limitations in the present study need to be acknowledged. The vignettes adopted here were originally inspired by over 550 real-life cases obtained from a law enforcement survey in the United States (Wolak & Finkelhor, 2011) and as such they may not be exact replicas of real-life scenarios that teachers in Ireland may encounter. Moreover, the original cases described by Wolak and Finkelhor (2011) always involve an adult finding out about the exchange of sexual pictures, and later reporting it to the police. However, such incidents are not always known to adults, especially considering that teenagers tend not to tell adults about episodes of sexual harassment (Lipson, 2001). A further limitation to the present study is that PSTs who participated in this study might have had limited experience in dealing with the unwanted dissemination of youth-produced sexual images; therefore, their responses to the hypothetical vignette may not be fully representative of how they would behave when confronted with real life cases. It would be interesting if future research could assess the perceived responsibility and attribution of blame of teachers with more years of service and compare their responses to teachers who have limited practical experience.
A methodological limitation of the present study is that all the statements in the Rape Myth Acceptance scale relate to situations where the target is a girl (although the instrument did show good internal reliability with the current sample). The same measure with a boy target should be included in future studies to investigate the role of gender in rape myths. Moreover, the items used to assess perceived responsibility and attribution of blame were adopted for the current study and had not been previously validated as standardised assessments for these variables. Indeed, these measurements were adopted because of the lack of instruments specifically assessing these concepts in the field of the unwanted sharing of sexual images. Further research should work on the creation of a validated instrument for such purposes. Another limitation of the present research is given by its correlational nature, which prevents us from inferring a causal relationship in our results. For instance, we can neither ensure that higher acceptance of rape myths causes greater attribution of blame towards the target of the dissemination of sexual pictures, nor the opposite. However, we can state that a significant relationship between these variables exists and that it should be investigated by further research with a longitudinal design in order to make causality inferences. A final limitation of this study is the fact that we did not provide the PSTs with any information about the ages of the hypothetical students discussed in the vignettes. Future research would need to take this into account to investigate if teachers' responses in terms of either blame or responsibility differ in this regard.
Despite these limitations, this study offers a novel contribution to the literature. The present study showed that PSTs' attitudes differ according to each different subtype of sexting, indicating the need to raise awareness about teachers' roles and responsibilities in all kind of incidents involving the unwanted dissemination of students' sexual pictures. An additional contribution of the present study is represented by the assessment of the Pre-service Teachers' rape myth acceptance. Besides predicting victim-blame in rape cases, as shown by previous research, this variable now seems to play a role in predicting blame towards the target of the nonconsensual sharing of sexual pictures as well, suggesting that the perception of the target/victim may be similar in episodes of rape and unwanted dissemination of sexual images.

Practical implications
Future prevention and intervention programmes should increase teachers' awareness regarding all forms of unwanted sharing of youth-produced sexual images and the role that teachers can play in preventing such incidents from happening. Since the consequences of non-consensual sharing of sexual images can affect young people's school lives as well (as reported in Hadwin, 2017), it is also teachers' responsibility to intervene in these situations. Previous research does show that teachers feel responsible for educating their students about this phenomenon and its impact, with a view to primary prevention and to the avoidance of legal consequences (Parmenter, 2016). Training on how to correctly handle incidents of youth-produced sexual images would benefit teachers, who otherwise might feel unprepared (Mura et al., 2014). Our results demonstrate that individual factors such as empathy and rape myth acceptance also appear to be important in to influencing PSTs' response to episodes involving the unwanted dissemination of sexual images, by affecting their attribution of blame towards the target. Hence, future programmes and university education should encourage and promote empathy. Increasing empathy towards the target can indeed enhance the likelihood of intervening to support the target (Schacter et al., 2016). Future interventions could focus on educating PSTs' about sexual violence and deconstructing the myths connected to it, with a view to reducing rape myth acceptance and therefore victim-blaming.

Funding
This work was supported by the Irish Research Council and the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 713279.
promised not to show it to anyone else. Your student was really annoyed because the next day, a friend of hers said that loads of people had seen the picture and it was spreading around school.
Experimental incident: romantic relationship. One of your students texted a semi-nude photo of herself to her boyfriend/girlfriend (an ex pupil). The image was discovered on the boyfriend/ girlfriend's mobile phone by his/her mother who deleted the image and then contacted the girl's parents. The girls' parents approached you for advice. The girl and her boyfriend/girlfriend assured you that no physical sexual activity took place between them prior to this event on or offline.
Experimental incident: attention seeking. One of the students in your class posted semi-nude pictures of herself on a social networking site. The website operator eventually took down the photos, but by then the other students and the teachers had already seen them.
Note. The target of the vignettes was a girl for about half of the participants (n ¼ 47); the other half (n ¼ 45) read vignettes where the target was a boy.