Examining perceptions of blame and homophobia: the impact of sexual orientation on non-consensual intimate image dissemination

ABSTRACT

Our aim was to investigate whether blame attribution toward same-sex victims compared to heterosexual victims of NCII and offense severity evaluation were associated with homophobic attitudes. Two-hundred-and-sixty men and women, who described themselves as heterosexual, were presented with vignettes depicting an NCII offense in which the intimate material was released by the victim’s ex-intimate partner. Victim’s and perpetrator’s sex were manipulated (i.e. gay, lesbian, heterosexual). We found that higher levels of homophobic attitudes were linked with lower severity attributions for same-sex perpetrator-victim pairs, but not for opposite-sex pairs. Additionally, higher levels of homophobic attitudes were associated with lower blame attributed to the perpetrator, but higher blame attributed to the victim. Sex differences were found, revealing that men tended to attribute higher blame to same-sex victims compared to heterosexual victims, whereas women showed no significant differences in blame attribution based on victim’s sexual orientation. Men also exhibited higher levels of homophobic attitudes than did women. The study reveals the impact of homophobic attitudes on blame allocation in NCII cases, emphasizing the need for targeted interventions to reduce victim-blaming and promote inclusivity. The role of technology in shaping behavior was also underscored, urging consideration of sex and sexual orientation in response strategies.

KEYWORDS:

• Non-consensual intimate image dissemination (NCII)
• homophobic attitudes
• blame attribution
• sexual orientation

Introduction

The intersection of technology and societal interactions significantly shapes the exchange of information and the safeguarding of privacy (Anderson et al., 2017; Umanailo et al., 2019; Zvi & Shechory Bitton, 2020). This technological evolution has also given rise to a new type of sexual violence – technology-facilitated sexual violence (TFSV; Henry et al., 2018). TFSV encompasses a range of sexually aggressive behaviors perpetrated using digital technologies, including non-consensual intimate image (NCII) dissemination (Snaychuk & O’Neill, 2020). NCII is a form of image-based sexual abuse that has gained an increasing amount of public and legal attention in recent years. NCII refers to the sharing of sexually explicit or intimate photos or videos of individuals without their consent (McGlynn, Rackley, & Houghton, 2017) or under coercive circumstances (Serpe & Brown, 2022). NCII can occur across various digital platforms, including social media, smartphones, and dedicated websites (Bates, 2017). Perpetrators leverage these technologies to share private material without consent, driven by motives such as revenge, coercion, blackmail, entertainment, social status, and sexual satisfaction (McGlynn et al., 2017).

There is compelling evidence that NCII is widespread globally (e.g. Henry et al., 2023; Lenhart et al., 2016). However, quantifying its prevalence and incidence poses challenges, as many individuals are unaware of whether their intimate images have been taken or shared (Bond & Tyrrell, 2021; Henry & Powel, 2018). A few studies have shed light on the issue's scale. In the USA, 4% of Internet users, around 10.4 million people, have faced threats or harm from the non-consensual distribution of intimate materials online (Lenhart et al., 2016). Another survey found that 8% of 3,044 respondents were victims of such distribution (Ruvalcaba & Eaton, 2020). In New Zealand, 5% of 1,001 adults experienced unauthorized sharing or threats regarding intimate images (Netsafe, 2019). Another study involving 2,028 New Zealand respondents aged 16–64 revealed that 35.1% experienced someone taking a nude or sexual image of them without their consent, 21.7% experienced unauthorized sharing, and 19.9% faced threats of such actions. Additionally, gay, lesbian, bisexual, and younger respondents were significantly more likely to report such abuse (Henry et al., 2017). In Israel, legislation was enacted in 2014 specifically prohibiting the distribution of intimate images, films, or tapes of another person without their consent, and by the end of 2017, 643 such police cases had been opened, with 83% of the victims being women (Yachimowitz-Cohen, 2018). Since then, the number of cases has been increasing. For example, according to the annual report (2021) of the Association of assistance centers for victims and victims of sexual assault in 2020, 377 cases were opened.

Research consistently underlines the detrimental impacts on NCII victims, showcasing parallels between the consequences experienced by victims of physical sexual offenses. Both NCII and physical sexual violence share commonalities as forms of intimate and intrusive acts inducing emotional trauma and rendering victims vulnerable. Contrary to assumptions that non-physical sexual violence might cause less harm, empirical evidence demonstrates that NCII victims endure profound negative psychological repercussions including shame, low self-esteem, reduced trust, depression, and anxiety disorders (Bates, 2017; Ruvalcaba & Eaton, 2020). NCII frequently results in emotional turmoil, compromised privacy, and reputational harm for victims (Henry et al., 2020; Powell et al., 2022), and in some instances, it may escalate into instances of bullying, harassment, and even physical violence (Bates, 2017; McGlynn et al., 2021). Moreover, the unique aspect of NCII resides in the perpetuation and dissemination of intimate images through modern technology, fostering prolonged victimization and societal humiliation (e.g. McGlynn et al., 2021). Although both physical and non-physical forms of sexual violence inflict enduring psychological distress, the digital nature of NCII introduces a distinct dimension: the perpetuation of images across online platforms, often leading to ongoing humiliation and recurrent victimization. The extensive reach and digital permanence of these images might exacerbate the duration and complexity of psychological distress experienced by victims, posing a unique challenge in the contemporary landscape of sexual violence.

For instance, ‘sexting,’ defined as the exchange of intimate images via electronic communication, often contributes to NCII. This behavior, prevalent among adolescents and young adults, can lead to cases of betrayal when what was initially perceived as a consensual act turns into victimization (Gavin & Scott, 2019). Studies indicate varying rates of sexting among young adults, with approximately 14.8% engaging in sending sexts and 27.4% receiving them (Madigan, Ly, Rash, Van Ouytsel, & Temple, 2018). Particularly, sexual minority youth are inclined to participate in sexting activities (Mori et al., 2020).

Although engaging in sexting does not equate to consenting to the distribution of one’s intimate photos (Henry & Powell, 2018), victims face higher blame attributions when the stolen images are self-taken, as opposed to when they are taken without consent (Starr, 2018; Zvi, 2022a, 2022b; Zvi & Shechory Bitton, 2021). This finding aligns with the victim precipitation theory, which suggests that victims may contribute to their own victimization to some extent (Amir, 1967; Schafer, 1968). They bear some responsibility for the circumstances that led to the crime being perpetrated against them (e.g. self-taken images) (Timmer & Norman, 1984). Hence, attributions of blame to NCII victims and assessments of the harm caused can be influenced by the behavior of the victims themselves (Fido et al., 2021).

Victim-blaming is also widely acknowledged to be significantly influenced by sex biases, such as the prevailing sexual double standard that subjects women to harsher judgment than men for expressing their sexuality or engaging in sexual behaviors (Flynn et al., 2023; Zvi, 2022a, 2022b). Recent research has provided evidence of the reinforcement of this double standard in online activities. Women have shared their experiences of how their online activities are frequently sexualized and subjected to negative judgment, resulting in shaming and humiliation. In contrast, the same activities by men are often perceived as humorous and elicit praise and respect (Handyside & Ringrose, 2017). Furthermore, the tendency to blame victims has been found in many studies to characterize men to a greater extent than women. A few studies regarding public perceptions of NCII indicated sex differences in victim-blaming, with men being significantly more likely than women to blame the victims (Bothamley & Tully, 2018; Zvi, 2022a, 2022b; Zvi & Shechory Bitton, 2021).

However, there still exists a significant gap in understanding the influence of sex differences on attributions of blame within cases of NCII (Karasavva & Forth, 2022; Powell et al., 2022; Powell & Henry, 2019). Moreover, alongside this gap, another notable deficiency exists in the research; this lacuna pertains to the public perception of NCII and its association with the sexual orientation of both victims and perpetrators (Karasavva & Forth, 2022; Serpe & Brown, 2022; Zvi & Shechory Bitton, 2020). Compared to heterosexual individuals, sexual minority individuals are at an increased risk of becoming victims of various crimes, and recent evidence suggests that this vulnerability may also extend to NCII (Henry et al., 2017; Karasavva & Forth, 2022; Waldman, 2019).

Considering the heightened victim-blaming experienced by same-sex victims compared to heterosexual victims in cases of sexual violence (Davies & Rogers, 2006; Shechory Bitton & Jaeger, 2020; Van der Bruggen & Grubb, 2014; Walfield, 2018; White & Yamawaki, 2009), it is reasonable to anticipate that same-sex victims of NCII may face even greater levels of blame. Those who adhere to traditional sex roles might interpret same-sex relationships as challenging conventional masculinity or societal expectations, potentially leading to increased blame attribution toward same-sex victims who are seen as violators of traditional sex roles. As such, these victims may face greater blame and stigma. This same-sex blaming tendency may be more pronounced among men, with research suggesting higher attributions of blame and homophobic attitudes among heterosexual men compared to heterosexual women (Anderson, 2004; Davies et al., 2012). As stated, the increased victim-blaming attitudes among heterosexual men might be rooted in a perceived threat to their masculine identity, perceiving same-sex relationships as a challenge to traditional masculinity and societal norms (Herek, 1986; Kite & Whitley, 1996). This in-group/out-group dynamic, influenced by social norms and traditional sex roles, significantly shapes perceptions of victims and may contribute to the differential treatment of same-sex NCII cases compared to heterosexual cases.

The current study

In our current technological era, understanding the intricate link between societal attitudes, technological progress, and the perception of abuses such as NCII is crucial. This interplay between technological capabilities and social norms may affect responses to NCII. Recognizing and comprehending this intricate relationship is vital for crafting tailored strategies and interventions suited to the contemporary digital landscape, ensuring protection against digital abuse. To date, there has been a notable absence of studies investigating the connections between blame attribution, homophobic attitudes, and same-sex cases of NCII. However, it has been consistently found that homophobic attitudes can predict victim-blaming in cases of sexual assault (Van der Bruggen & Grubb, 2014; Walfield, 2018). Individuals with homophobic attitudes tend to assign higher levels of blame to same-sex rape victims when compared to heterosexual victims (Davies & Rogers, 2006; White & Yamawaki, 2009). This tendency has been observed to be more prevalent among men than among women (Davies & McCartney, 2003; Davies & Rogers, 2006).

The aim of our study was to investigate whether blame attributions toward same-sex victims compared to heterosexual victims of NCII were associated with homophobic attitudes. We sought to explore the relationship between blame attributions toward victims and perpetrators in NCII cases and homophobic attitudes among heterosexual man and women, taking into consideration the sexual orientation of both the victim and the perpetrator. Moreover, previous research has suggested that downplaying the seriousness of the offense indirectly implicates the victim (Zvi & Shechory Bitton, 2021). Therefore, we assessed not only the culpability of the perpetrator but also participants’ perceptions regarding the severity of the offense.

Based on the literature review, we offered the following hypotheses: (1) Men would attribute more blame to the victim than would women; (2) Higher blame would be attributed to women victims than to men victims; (3) Higher offense severity would be attributed to offenses committed against women than men victims. (4) Men would express more homophobic attitudes than would women; and (5) Homophobic attitudes would have a stronger relationship with offense severity, blaming the perpetrator, and blaming the victim in same-sex perpetrator-victim cases compared to opposite-sex cases.

Method

Design

We employed a quasi-experimental study design, comprising eight sub-groups (2 × 2 × 2): participant’s sex (man/woman), perpetrator’s sex (man/woman), victim’s sex (man/woman). Each of the eight sub-groups included 31–34 participants. The study’s independent variables were the three experimental conditions, and homophobic attitudes. The dependent variables were offense severity, and attributions of blame to the victim and perpetrator. Inclusion criteria were age – 18–60 years – and a heterosexual orientation (asked in the demographic questionnaire).

Participants

Two-hundred-and-sixty heterosexual men (n = 128, 49.2%) and women (n = 132, 50.8%) were recruited for participation. The inclusion of only heterosexual individuals provided a more specific scope of study by examining groups that have been traditionally studied in terms of attitudes of victim-blame and homophobic attitudes. Most participants were in their 30s, with men being significantly older than women (M = 36.59, SD = 11.21 and M = 29.34, SD = 7.77, respectively), t(225.51) = 6.04, p < .001. The vast majority of them were Jewish (98.8%), with approximately 91% being Israeli-born and about 75% identifying as secular. Others were somewhat religious – meaning that they observed only some religious commandments (about 19%) – or religious (about 6%). Additionally, about 70% of participants reported being married or in a relationship, and most others were single (about 27%). About 75% reported having a higher education (i.e. bachelor’s degree, master’s degree, PhD), whereas the others had lower levels of education (i.e. high school or vocational education). No sex differences were found in these background variables, except for age (see Table 1).

Table 1. Demographic characteristics of the research participants, by sex (N = 260).

		Men (n = 128)	Women (n = 132)	Difference
Age M (SD)	Range 19–57	36.59 (11.21)	29.34 (7.77)	t (225.51) = 6.04 (p < .001)
Ethnicity – n (%)	Jewish	127 (99.2)	130 (98.5)	–
Country of birth – n (%)	Israel	117 (91.4)	121 (91.7)	–
Marital status – n (%)	Married/in a relationship	92 (71.9)	89 (67.4)	Z = 0.78 (p = .435) (Married /in a relationship vs. others)
	Single	34 (26.6)	37 (28.0)
	Divorced, widowed	2 (1.6)	6 (4.54)
Religiosity – n (%)	Secular	93 (72.7)	101 (76.5)	Z = 0.71 (p = .475) (secular vs. others)
	Somewhat religious	27 (21.1)	23 (17.4)
	Religious	8 (6.3)	8 (6.1)
Education level – n (%)	Higher education (BA, MA, PhD)	98 (76.6)	95 (72.0)	Z = 0.85 (p = .397)
	High school or below; vocational education	30 (23.4)	37 (28.0)

Note. Differences were not calculated for variables with low variance.

When comparing the demographic characteristics by the study conditions (i.e. participant’s sex, perpetrator’s sex, and victim’s sex), one significant difference was revealed. Participant age differed only by participant’s sex (see Table 1), and not by the study conditions (perpetrator’s sex or victim’s sex or their interactions) (p = .085 to p = .910). Age was thus controlled for in further analyses. Marital status, level of religiosity, and level of education were unrelated to participant’s sex or study conditions (p = .074 to p = .967).

Sample size was determined using G*Power 3 (Faul et al., 2007, 2009). For an analysis of covariance with eight subgroups (2 × 2 × 2) and two covariates, assuming α = .05, a moderate-low effect size f = 0.18, and a power of 0.80, the resulting sample size required is 245 participants. Similarly, for a multiple regression analysis with six predictors, assuming α = .05, a moderate-low effect size f2 = 0.10, and a power of 0.80, the resulting sample size needed is 143 participants.

Measures

Sociodemographic questionnaire

Respondents were requested to specify their sex and sexual orientation, age, marital status (married/in a relationship, single, divorced, widowed), education level (lower – i.e. high school or vocational education; higher – i.e. bachelor’s degree, master’s degree, PhD), and degree of religiosity (secular, somewhat religious, religious).

Vignettes

Four vignettes presenting the non-consensual distribution of intimate photographs by an ex-intimate partner were created for the current study. The independent variables of the perpetrator’s sex and the victim’s sex were provided in these vignettes. The cases were identical in all aspects aside from the sexual orientation of the victim and perpetrator: (1) a woman perpetrator and a man victim (heterosexual); (2) a man perpetrator and a woman victim (heterosexual); (3) a man perpetrator and a man victim (same-sex orientation); and (4) a woman perpetrator and a woman victim (same-sex orientation). These four vignettes depicted the various combinations of NCII offense scenarios that were chosen for the study. Each participant was exposed to one of the four scenarios.

Below is the vignette template. Hebrew names were chosen so that the sexes of the characters would be clear, and the pronouns ‘he’ and ‘she’ were used several times (for men, the names were Amir, Itamar, Danny, and for women, the names were Anat, Shani, Karin).

Following a night of drinking with friends at a local pub, [victim's name] and [perpetrator's name] went back to [perpetrator's name] apartment where they shared an intimate and relaxed atmosphere. During the evening, [perpetrator's name] asked [victim's name] if he/she had ever taken naked pictures of himself/herself. [Victim's name] responded negatively and added that it could be enjoyable to take such photos with someone you can trust. That night, they used [perpetrator's name] cellphone to take naked pictures of each other. Two months later, they broke up. Roughly a year later, one morning [victim's name] received a call from a friend informing him/her that intimate photos of him/her were circulating on social media. Initially, [victim's name] believed it was a prank, but phone calls pouring in non-stop from family members and friends made him/her realize that it was real. [Victim's name] felt as though his/her life had been ruined and requested police involvement. Following an investigation, it was discovered that [perpetrator's name], the ex-partner, had distributed the pictures.

Assessments of offense severity and attributions of blame to the victim and perpetrator

In accordance with the methodology of previous studies (e.g. Zvi, 2022a, 2022b), assessments of offense severity and attributions of blame to both perpetrator and victim were examined using a vignette depicting the nonconsensual sharing of intimate images, as detailed above. The questionnaire consisted of a total of 10 items. An exploratory factor analysis was conducted on these items using principal component analysis. Three indices were defined, each representing the following variables, calculated as the average of the items that composed them.

Severity judgment

To assess participants’ perceptions of offense severity, participants responded on a scale ranging from 1 (strongly disagree) to 5 (strongly agree) to the following statements: 1. There is reason to file a complaint with the police in the case of nonconsensual distribution of intimate images; 2. There is reason to prosecute [perpetrator's name] for the distribution of the photos; and 3. The act of [perpetrator's name] should be defined as a serious offense. Additionally, afterward, participants read a verbatim transcript of the law, which states that the publication of a photograph, film, or recording of a person of a sexual nature, in circumstances in which the publication might humiliate or degrade the person, and without receiving consent for its publication, will be considered sexual harassment, and carries a sentence of up to five years in prison. Participants were then presented with a fourth statement: 4. It is appropriate to hand down a prison sentence for such an offense. These four items yielded one factor in the principal component analysis, with an eigenvalue of 2.68, and the factor explaining 12.92% of the variance. Item loadings ranged between 0.66 and 0.81, and acceptable internal consistency was found (α = .73).

Blame attributions toward the perpetrator and victim were assessed with six items (three for the perpetrator and three for the victim). Blaming the perpetrator: 1. [Perpetrator’s name] can be blamed for distributing the photos; 2. [Perpetrator’s name] is responsible for the distribution of photos; and 3. The claim by [victim’s name] that he/she is a victim is justified. Blaming the victim: 4. [Victim’s name] can be blamed for the distribution of photos; 5. [Victim’s name] is responsible for the distribution of photos; 6. [Victim’s name] could have prevented the situation. Participants were asked to rate their agreement with each item on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Homophobic attitudes were assessed using the Homosexuality Attitude Scale (HAS) (Kite & Deaux, 1986), which consists of 21 questions designed to measure an individual's stereotypes, misconceptions, and anxieties related to homosexuality. The HAS is a widely known and reliable attitudinal scale with excellent internal consistency (alphas of 0.92) and is equally reliable in assessing attitudes toward both gay individuals and lesbians (Kite & Deaux, 1986). Participants rated each item on a Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree). To calculate the composite score/total HAS, all positive items on the scale were reverse-scored. Thus, a higher score on the scale indicated a more positive attitude toward homosexuality. The HAS has been used in various studies that examined homophobic attitudes, including in Israel, where similar reliability was obtained (Cronbach's α of 0.90 for the total score) (e.g. Averett & Hegde, 2012; Eick et al., 2016). In the current study, confirmatory factor analysis (CFA) for this measure showed a good fit: C_min/df = 1.79, TLI = .927, comparative fit index (CFI) = .937, root mean square error of approximation (RMSEA) = .057, standardized root mean square residuals (RSMR) = .045. The internal consistency of the scale was found to be 0.91.

Procedure

The study was conducted as a cross-sectional, Internet-based survey. A link to the research questionnaires was distributed in groups on social networks, forums, and websites. After providing the participants with information about the research purpose and procedures, assuring them of confidentiality and anonymity, informing them that they could withdraw from the study at any time, and having them sign informed consent, they were randomly assigned to one of the research conditions. The research set included four case scenarios in which intimate images were disseminated without consent, which differed only according to the sex of the victim and the sex of the perpetrator. Using the secure online system Qualtrics, participants completed four questionnaires: a sociodemographic questionnaire, a blame attributions questionnaire, an offense severity judgment questionnaire, and a questionnaire to assess participants’ level of homophobia (presented to the participants in this order). The study adhered to all relevant ethical guidelines and was approved by the institutional ethics committee of XX University. At the end of the questionnaires, participants were given the contact information of available professionals if they felt in need of support.

Data analysis

Data were analyzed with SPSS ver. 28. To examine the initial four hypotheses, distinctions in offense severity, assigning blame to the perpetrator, and assigning blame to the victim were evaluated using a multivariate analysis of covariance (MANCOVA). Demographic variables (age, education level, marital status, and religiosity) were examined for their associations with the study variables, in order to identify those that needed to be controlled for in the data analysis. Age showed negative associations with offense severity (r = −.28, p < .001) and blaming the perpetrator (r = −.16, p = .008), indicating that younger participants attributed higher severity and blame to the perpetrator. Differences were found based on religiosity, with secular participants attributing lower blame to the victim (M = 2.23, SD = 0.77) and holding lower homophobic attitudes (M = 1.37, SD = 0.37) compared to somewhat religious or religious participants (blaming the victim: M = 2.64, SD = 0.89; homophobic attitudes: M = 1.78, SD = 0.67). No significant differences were observed on the basis of education level or marital status. Consequently, the research hypotheses were examined while controlling for participants’ age and religiosity (0 – somewhat religious/religious, 1-secular). In other words, age and religiosity were controlled for in the MANCOVA, with the latter being employed as a control due to its association with blaming the victim and homophobic attitudes. To examine the fifth hypothesis, three multiple regression models were calculated, for offense severity, blaming the perpetrator, and blaming the victim, as the dependent variables. Predictors were sex (participant, perpetrator, victim) and homophobic attitudes. Significant interactions were interpreted with simple slopes (Aiken & West, 1991; Dawson, 2014).

Results

Descriptive results

Offense severity was high, along with high levels of blaming the perpetrator, and comparatively low levels of blaming the victim (see Table 2). Offense severity exhibited a positive correlation with blaming the perpetrator and a negative correlation with blaming the victim. Additionally, there was a negative correlation between blaming the perpetrator and blaming the victim. Homophobic attitudes were generally low and displayed a negative correlation with offense severity and blaming the perpetrator. However, a positive correlation was found between homophobic attitudes and blaming the victim.

Table 2. Means, standard deviations, and correlations for the study variables (N = 260).

	M (SD)	1.	2.	3.	4.
1.Offense severity	4.62 (0.51)	1
2. Blaming perpetrator	4.58 (0.61)	.42***	1
3. Blaming victim	2.33 (0.82)	−.25***	−.25***	1
4.Homophobic attitudes	1.47 (0.50)	−.36***	−.28***	.33***	1

Note. Range 1–5. ***p < .001.

Hypotheses 1 through 4

To assess the first four hypotheses, differences in offense severity, blaming the perpetrator, blaming the victim, and homophobic attitudes were assessed on the basis of participant's sex, perpetrator's sex, and victim's sex using MANCOVA, with age and religiosity as controls. Table 3 shows the means and standard deviations for these main effects.

Table 3. Means and standard deviations by sex (participant, perpetrator, victim) (N = 260).

	Participant’s sex		Perpetrator’s sex		Victim’s sex
	Men (n = 128) M (SD)	Women (n = 132) M (SD)	Men (n = 131) M (SD)	Women (n = 129) M (SD)	Men (n = 131) M (SD)	Women (n = 129) M (SD)
Severity of the offense	4.53 (0.57)	4.71 (0.41)	4.64 (0.52)	4.61 (0.49)	4.66 (0.50)	4.59 (0.51)
Blaming perpetrator	4.51 (0.66)	4.65 (0.52)	4.67 (0.46)	4.50 (0.70)	4.59 (0.60)	4.57 (0.59)
Blaming victim	2.41 (0.84)	2.26 (0.79)	2.31 (0.83)	2.36 (0.80)	2.36 (0.83)	2.31 (0.81)
Homophobic attitudes (n = 241)	1.61 (0.51)	1.35 (0.45)	–	–	–	–

The overall MANCOVA did not have significant effects (p = .299 to p = .899). However, significant findings for offense severity showed a main effect for participant sex (F(1, 250) = 4.38, p = .037, η² = .018), with women attributing higher severity than men (Table 3). However, other differences, including main effects and two – and three-way interactions (by participant’s sex × perpetrator’s sex × victim’s sex), were not significant (p = .355 to p = .888). Regarding blaming the perpetrator, a significant main effect was observed for perpetrator's sex, F(1, 250) = 4.12, p = .044, η² = .016. Men perpetrators received higher blame than did women perpetrators, as shown in Table 3. All other differences, including main effects and two – and three-way interactions (by participant’s sex × perpetrator’s sex × victim’s sex), were not statistically significant (p = .167 to p = .946). Regarding blaming the victim, no significant main effects or two-way interactions were found (p = .077 to p = .963). However, the three-way interaction was significant, F(1, 250) = 5.76, p = .017, η² = .023 (for all significant and non-significant F values, see Supplementary Table 1).

Further analysis using estimated marginal means revealed that men attributed higher blame to men victims of men perpetrators (M = 2.68, SD = 0.87) compared to women victims of men perpetrators (M = 2.23, SD = 0.78), F(1, 250) = 4.39, p = .037, η² = .018. Additionally, men attributed higher blame to women victims of women perpetrators (M = 2.54, SD = 0.79) compared to men victims of women perpetrators (M = 2.20, SD = 0.65), F(1, 250) = 3.98, p = .047, η² = .016. However, no significant differences were observed for women in these comparisons, F(1, 250) = 0.01, p = .957, η² = .001; and F(1, 250) = 0.45, p = .501, η² = .002; respectively. Regarding homophobic attitudes, a significant sex difference was found, F(1, 237) = 14.21, p < .001, η² = .057, with men exhibiting higher levels of homophobic attitudes than women.

Hypothesis 5

To examine the fifth hypothesis – namely, the association between homophobic attitudes and offense severity, blaming the perpetrator, and blaming the victim, in cases involving same-sex and opposite-sex perpetrator-victim pairs – three multiple hierarchical regression models were conducted (n = 130 for each group). The models took into account participant's sex (0-women, 1-men), perpetrator-victim relationship (0-opposite sex, 1-same sex), standardized homophobic attitudes, and the interaction between same-sex vs. opposite-sex pairs and homophobic attitudes. Age (standardized) and level of religiosity (0-somewhat religious/religious, 1-secular) were included as control variables. The results of the regression models are displayed in Table 4.

Table 4. Multiple regression models for offense severity, blaming perpetrator, and blaming victim (N = 241).

	Severity of the offense			Blaming perpetrator			Blaming victim
	B (SE)	β (p)	95%CI	B (SE)	β (p)	95%CI	B (SE)	β (p)	95%CI
Age	−0.10 (0.03)	−.21 (.001)	−0.17, −0.04	−0.07 (0.04)	−.12 (.067)	0.84, 1.19	−0.05 (0.05)	−.06 (.332)	0.84, 1.18
Level of religiosity (secular)	−0.06 (0.07)	−.06 (.371)	−0.21, 0.08	0.04 (0.09)	.03 (.683)	0.87, 1.16	−0.22 (0.12)	−.11 (.082)	0.85, 1.17
Participant’s sex (men)	−0.03 (0.07)	−.03 (.601)	−0.16, 0.09	−0.04 (0.08)	−.04 (.603)	0.80, 1.26	0.07 (0.11)	.04 (.550)	0.79, 1.27
Same sex vs. opposite sex perpetrator and victim (same sex)	−0.01 (0.06)	−.01 (.982)	−0.12, 0.11	−0.01 (0.07)	−.01 (.850)	0.98, 1.02	0.05 (0.10)	.03 (.609)	0.99, 1.01
Homophobic attitudes	−0.07 (0.05)	−.15 (.138)	−0.17, 0.02	−0.16 (0.06)	−.26 (.016)	0.32, 3.11	0.21 (0.09)	.26 (.013)	0.34, 2.91
Same sex vs. opposite sex perpetrator and victim × Homophobic attitudes	−0.16 (0.06)	−.25 (.010)	−0.28, −0.04	−0.02 (0.08)	−.03 (.765)	0.34, 2.90	0.05 (0.10)	.04 (.648)	0.38, 2.66
Adj. R^2	.179			.100			.115
F(6, 234)	9.63, p < .001			5.43, p < .001			6.17, p < .001

The results indicated that all three regression models were statistically significant, explaining 10% to 18% of the variance. Notably, the severity of the offense was significantly explained by the interaction between same-sex versus opposite-sex perpetrator-victim pairs and homophobic attitudes. Further analysis using simple slopes (Aiken & West, 1991; Dawson, 2014) demonstrated that the slope for same-sex perpetrator and victim was significant (coefficient = −0.23, t = −5.85, p < .001), whereas the slope for opposite-sex perpetrator and victim was not significant (coefficient = −0.07, t = −1.50, p = .136), as illustrated in Figure 1. This finding suggests that higher levels of homophobic attitudes were associated with lower attributions of severity to the offense when the perpetrator and victim were of the same sex. However, this association did not reach significance for opposite-sex perpetrator and victim pairs.

Figure 1. The moderating effect of same sex vs. opposite sex perpetrator and victim on the relationship between homophobic attitudes and offense severity.

In terms of blaming the perpetrator, the results indicated a negative relationship with homophobic attitudes. Specifically, higher levels of homophobic attitudes were associated with attributing lower blame to the perpetrator. On the other hand, blaming the victim was positively associated with homophobic attitudes, with higher levels of homophobic attitudes being associated with attributing higher blame to the victim. However, in both cases, the interaction between same-sex versus opposite-sex perpetrator-victim pairs and homophobic attitudes did not reach statistical significance, suggesting that the relationship between homophobic attitudes and blame attribution did not differ significantly based on the sex composition of the perpetrator and victim. It should finally be noted that calculating the regressions without the interaction terms yielded significant results for homophobic attitudes in all models: for severity (β = −.35, p < .001), for blaming the perpetrator (β = −.29, p < .001), and for blaming the victim (β = .30, p < .001). Other main effects retained their non-significance.

Discussion

The escalating prevalence of NCII and its profound impact on victims (Henry et al., 2017) has necessitated a thorough understanding of the factors shaping blame attribution. Understanding these factors is vital for devising effective interventions and support systems. We will now delve into the significant findings regarding the correlation between homophobic attitudes, blame attribution, and the perceived severity of NCII offenses. Specifically, we will discuss how these factors vary within the heterosexual population, considering victims and perpetrators in both same-sex and opposite-sex contexts.

The study revealed that NCII is widely judged as a severe offense, with strong support for prosecuting perpetrators, irrespective of the perpetrator's sexual orientation (i.e. whether they are of the same sex or opposite sex than the victim). These findings align with findings from a previous study conducted in Israel (Zvi & Shechory Bitton, 2021), suggesting a convergence between legislation and public opinion, and perhaps indicative of a growing awareness and understanding of the matter.

However, a closer analysis of the results unveils a nuanced perspective, highlighting the substantial influence of homophobic attitudes on both the perception of offense severity and blame allocation in NCII cases. Higher levels of homophobic attitudes were found to be associated with diminished perceptions of severity when the perpetrator and victim shared the same sex. These findings underscore how societal attitudes toward gay men and lesbians may impact individuals’ assessments of the severity of the offense and the assignment of blame. Moreover, these findings are consistent with prior research demonstrating that individuals with homophobic attitudes tend to attribute greater blame to same-sex rape victims than they do to heterosexual victims (Davies & Rogers, 2006; White & Yamawaki, 2009). Such findings suggest a tendency to excuse the perpetrators while holding the victims more accountable, revealing a bias against same-sex relationships. Furthermore, these findings highlight the parallels with cases involving victims of sexual assault, indicating that same-sex victims experience heightened victim-blaming compared to heterosexual victims in cases of sexual violence (e.g. Shechory Bitton & Jaeger, 2020; Walfield, 2018).

In addition to the above, the regression findings showed a connection between homophobic attitudes and biases in judgment. Participants reporting higher levels of homophobic attitudes tended to assign blame to the NCII victims and showed leniency toward the perpetrators. These homophobic attitudes predicted blame attributions regardless of sex or sexual orientation. The findings underscore broader cultural trends and power dynamics, emphasizing the influential role of homophobic attitudes in shaping conservative views and prejudices regarding sexuality (Campo-Arias et al., 2010; Lingiardi et al., 2016; Van der Toorn et al., 2017).

Although direct comparisons with previous NCII-focused studies are lacking, the current findings align with existing research consistently linking homophobic attitudes to victim-blaming (Van der Bruggen & Grubb, 2014; Walfield, 2018). Furthermore, they strengthen the assumption that similar patterns of blame attribution occur for victims of sexual offenses. For example, studies examining perceptions of rape victims indicate that individuals with homophobic views tend to assign a higher degree of blame to same-sex victims than they do to heterosexual victims (Davies & Rogers, 2006; White & Yamawaki, 2009). These findings also align with previous research highlighting the presence of discriminatory attitudes toward sexual minorities in online networks (Marciano & Antebi-Gruszka, 2022), rendering them more susceptible to online harm (Walker & Sleath, 2017). This finding might be explained by the presence of stereotypical attitudes toward sex roles within Israeli society (Shechory Bitton & Jaeger, 2020). The persistence of these attitudes in Israel, akin to many Western societies, accentuates traditional sex roles, particularly emphasizing the expectation that people should adhere to stereotypical ideals of masculinity and femininity.

Despite notable strides in protecting the rights of sexual minorities, entrenched stereotypical attitudes persist, perpetuating negative perceptions toward the LGBTQ + community (Shechory Bitton & Jaeger, 2020). These attitudes, deeply rooted in traditional and conservative beliefs, contribute to the ongoing challenges faced by sexual minorities (Shechory Bitton & Bonny Noach, 2023; Shenkman et al., 2019). Support for these observations is evident in the annual report published by the The Association for the LGBT in Israel in Israel (2021) over the past decade. These reports consistently highlight the enduring harm experienced by LGBTQ₊ community members due to their sexual orientation across various life domains, from early education to military service, higher education, the workplace, and health services. Despite advances in protecting the rights of sexual minorities, these reports underscore the persistent struggles and discrimination experienced by LGBTQ + individuals in multiple facets of their lives.

Furthermore, our findings once again underscore the sex differences that manifest in more conservative and extreme positions among men than among women. In the current study, heterosexual men exhibited higher levels of homophobic attitudes than did women, and demonstrated a greater tendency to assign blame to same-sex victims of NCII than they did to heterosexual victims. These findings align with the patterns observed in prior research on rape myths (Li & Zheng, 2022; Lilley et al., 2023; Shechory Bitton & Jaeger, 2020) which indicate that, in general, men tend to hold more sexist beliefs than do women. This tendency extends to attributing higher levels of blame to victims of sexual crimes.

Despite advances in attitudes toward sexuality, persistent societal perceptions of sex roles contribute to a sexual double standard (Flynn et al., 2023; Zvi, 2022a, 2022b). Men are often expected to be sexually active and dominant, whereas women are expected to be passive and submissive (Endendijk et al., 2020). This double standard not only restricts women’s sexual expression but also leads to the condemnation of women who defy these norms, while men are often praised for engaging in similar behaviors (Bordini & Sperb, 2013; Handyside & Ringrose, 2017).

It is reasonable to assume that sex role perceptions and the sexual double standard also extend to biased perceptions toward gay men and lesbians in cases of NCII, as these individuals are viewed as violating traditional sex role beliefs (Berent et al., 2016; Lehavot & Lambert, 2007). Men, who tend to be more critical than women of deviations from sex roles, play a significant role in driving these double standards (Bosson et al., 2022). The consistent findings highlighting the much greater tendency of men to adopt rape myths and attribute more blame to victims (Lilley et al., 2023), as well as studies emphasizing the consistency in attributing blame toward sexual minorities who are victims of sexual offenses (Shechory Bitton & Jaeger, 2020), further support our assumption.

In summary, the current study's findings illuminate the complex relationship between homophobic attitudes, sex roles, and blame attribution in instances of NCII, particularly involving same-sex victims. Our research exposes the intricate interplay of societal biases related to sex and sexuality, impacting the assignment of blame. Notably, the men in the current study displayed heightened levels of homophobic attitudes and a greater inclination to blame same-sex victims, contributing to a deeper understanding of victim-blaming attitudes in sexual offense cases. Additionally, our research highlights the significant influence of technology on shaping the assignment of blame in such cases.

Significance and implications: The significance of our article extends to both theoretical and practical realms. The theoretical insights emerge from our examination of the interplay between homophobic attitudes, sex roles, and blame attribution in non-consensual intimate interactions, specifically in the digital realm. By situating our findings in the context of human–computer interaction, we have underscored the crucial role of technology in recording, distributing, hosting, and sometimes enabling the abuse of victims in NCII offenses. Furthermore, our study contributes to theoretical advances by highlighting how societal biases intersect with digital technologies in shaping blame attribution. Specifically, our findings offer new perspectives on the theoretical frameworks surrounding victim-blaming attitudes in online contexts, enriching discussions within fields such as criminology, sociology, and psychology.

On a practical level, our research sheds light on the dynamics of blame attribution and homophobic attitudes in such cases, offering valuable guidance for policymakers, educators, and support systems. Recognizing the influence of sex norms, homophobic beliefs, and technology can inform targeted interventions to reduce victim-blaming and foster inclusivity for all NCII victims, irrespective of sex or sexual orientation. Moreover, our findings suggest concrete strategies for intervention, such as implementing educational programs to challenge and dismantle homophobic attitudes, enhancing digital literacy to empower individuals in navigating online spaces safely, and advocating for legal measures to address the misuse of technology in perpetuating NCII offenses. By incorporating these specific suggestions into intervention efforts, stakeholders can work toward creating a more compassionate response and mitigating the profound impact of sexual offenses on victims’ lives, thus contributing to a more equitable and safer digital society.

Limitations and future directions: Several limitations of the present study warrant consideration. First, akin to other studies on social perceptions, our research relied on a relatively small, convenience-based online sample, rendering it susceptible to social desirability bias. Social desirability bias occurs when participants tailor their responses to align with societal norms rather than providing candid answers, potentially skewing results, particularly in sensitive domains such as attitudes toward sexuality. A shift toward a larger, more diverse sample derived through random sampling would bolster the external validity of the study. Moreover, the study’s cross-sectional design does not allow for the establishment of causal relationships between variables. Cross-sectional studies offer a static snapshot, complicating the determination of causality. Longitudinal research, by contrast, would allow for tracking participants over time and would enable the elucidation of temporal relationships and causal pathways.

Furthermore, the predominant inclusion of Jewish participants, although reflective of the study's locale, may limit the broader applicability of our findings across cultural contexts. The majority of Israeli society is modern-liberal, although some of its sub-cultures are more traditional and characterized by patriarchal values (e.g. Muslim Israeli-Arab citizens) (Shechory-Bitton & Kamel, 2014). Victim characteristics such as ethnicity, race, occupation, and socioeconomic status are also potential sources of bias in treatment and should be scrutinized further (Willmott et al., 2018; Zvi, 2022a, 2022b). Diverse cultural backgrounds may shape perceptions of victims and perpetrators. Future research endeavors should thus strive for inclusivity by exploring perceptions across various cultural and ethnic groups.

Moreover, the measurement of homophobic attitudes may have been constrained by the limited scope of existing instruments, highlighting the need for more comprehensive measures that capture the multifaceted nature of these attitudes. Similarly, although vignettes serve as valuable tools for exploring complex social phenomena, they may need to be enhanced in terms of realism and diversity so as to better mirror real-life situations and broaden their utility as research instruments. Qualitative methods, such as interviews or focus groups, can offer rich contextual data complementing quantitative analyses.

In conclusion, addressing these limitations and pursuing these future directions will enhance the robustness and applicability of our findings, thereby advancing our understanding of blame attribution in NCII cases and informing more effective interventions and support systems for victims.

Declarations

• The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

• The questionnaire and methodology for this study were approved by the human research ethics committee of Ariel University, Israel (AU-SOC-MSB-20211025).

Disclosure statement

No potential conflict of interest was reported by the author(s).