Dating violence perpetration: Masculine ideology and masculine gender role stress as predictors

The statistics from the National Commission on Violence against Women Annual Report, gathered from 237 partner agencies in Indonesia, revealed that there were 2,171 dating violence cases in 2017 (Komnas Perempuan, 2017). A study found that dating violence perpetration and victimization are more prevalent among college-aged couples (Karakurt, Keiley, & Posada, 2013). A survey conducted by the Knowledge Networks (2011) involving 284 women stated that 29% of female college students claimed to have experienced dating violence. This data indicated that one of three female students became a victim of dating violence from their male partners. Dating violence is intimate partner violence occurring in an informal relationship setting. It refers to any behavior that causes harm to those in the relationship either psychologically, physically, or sexually (World Health Organization, 2012). A meta-analytic study concluded that a recent finding showed no different rate of violence perpetration based on gender in an intimate relationship (Chan, 2011). However, when a study put consideration into the severity, motives, and consequences of violence, the study showed that men often initiate and perpetrate more severe violence, which leads to more severe injury and consequences compared to women (Chan, 2011). While men are more prone to initiate violence toward their partners, Kaukinen (2014) concluded that there was no pattern consistently able to explain variables that may trigger men’s violence against their partners. ART I C LE I NFO ABST R ACT

Therefore, the present study focuses on dating violence with men as perpetrators to better understand variables that might prompt violent acts.
Several factors trigger vulnerability to dating violence, such as direct experience to violence and witnessing acts of violence in family and friendship, anxiety, depression, emotional distress, and antisocial behavioral attitudes that support violent acts (Vagi et al., 2013). Gender socialization that encourages the idea that men act as perpetrators of violence and women as recipients of legitimized violence also becomes one of such factors (Lelaurain, Fonte, Giger, Guignard, & Lo Monaco, 2018). The belief system that attributes certain roles or behaviors as appropriate for men is known as masculine ideology (Levant, 2011). Men who support masculinity values engage in more frequent physical and sexual violence and attempt to control their partner (Casey et al., 2016). Within the masculine ideology, men are encouraged to display aggression and acts of violence to conform to the male role norms instructed as a result of social cognition and social influence (R. F Levant, 2011). The cultural standards on behaviors that are deemed reasonable by a particular gender can be explained by the Gender Role Strain Paradigm (GRSP).
GRSP views gender roles as psychologically and socially constructed entities (R. F Levant, 2011). A literature review by  explained that GRSP is based on social learning theory, as adults and children internalize prevailing gender roles norms through observational learning, reinforcement, and punishment in social interactions. In the GRSP, socialization resulting in gender roles and gendered behaviors, for both gender (Levant, 2011). The belief that there are roles or behaviors considered appropriate for men according to the expected cultural standards is known as masculine ideologies (R. F Levant, 2011).
A recent study involving 662 men aged 18-25 years from various racial backgrounds in the United States showed a positive correlation between masculine ideology and intimate partner violence. The study concluded that men who support traditional and rigid masculinity values tend to engage in more frequent physical and sexual violence, and attempt to control their female partners (Casey et al., 2016). Another study conducted by Tager, Good, and Brammer (2010) involving 108 adult males in rural and urban settings also showed that conformity to masculine norms accounted for variance in reported intimate partner violence, specifically in psychological abuse. Masculinity also served as a risk factor for violence against women in young heterosexual males, either psychologically or physically (Willie, Khondkaryan, Callands, & Kershaw, 2018).
However, other research showed that having a traditional masculine ideology does not always encourage men to commit violence to their partners. Research by Gallagher and Parrott (2011) found that out of three dimensions (toughness, status, and anti-femininity) of hegemonic masculine ideology, there was only one dimension (toughness) that directly predicts men's hostility toward women. Another research found that masculine ideology, consisting of anti-femininity norms and avoidance of subordination to women, does not necessarily predict sexual aggression against women (Smith, Parrot, & Tharp, 2015). The inconsistency of these studies prompts further research to identify the relationship between masculine ideology and dating violence.
Gender Role Strain Paradigm also formulated another variety of male gender role strain. The strain is described as a discrepancy strain or masculine gender role stress, which was caused by the inability to live up to manhood ideal (Levant, 2011). Researchers have defined masculine gender role stress as a tendency to experience distress when a man encounters a threat to his masculine identity (Copenhaver, Lash, & Eisler, 2000).
The failure to follow traditional gender roles resulted in feelings of distress (Mcdermott, Naylor, McKelvey, & Kantra, 2017). The presence of such distress prompts men to feel angry, anxious, increase negative affects, trigger hostility, and aggressive behavior (O'Neil & Crapser, 2011). The emergence of anger, hostility, and negative emotions plays a role in triggering violence against an intimate partner (Birkley & Eckhardt, 2015). The 14 ISSN 2598-6368 (online) / 1693-7236 (print) Adiningsih et.al tendency to use violence toward their female partner as a way to overcompensate the gendered expectation will be higher if a man is not considered masculine enough and experience distress stemming from the discrepancy (Reidy, Smith-Darden, Cortina, Kernsmith, & Kernsmith, 2015).
According to research conducted by Reidy et al. (2015), masculine discrepancy stress serves as a predictor of the likelihood to perpetrate dating violence. The higher distress experienced by a teenage boy when his masculinity is threatened, the more he tries to show the masculine side within himself by committing physical violence towards his girlfriend (Reidy et al., 2015). Other studies have also demonstrated the role of masculine gender role stress as a predictor of intimate partner violence. Men who experience masculine gender role stress reported a higher rate of domestic violence history compared to men who did not experience the stress of masculine gender roles (Lisco, Leone, Gallagher, & Parrott, 2015).
Conforming to the previous researches, the results of a research conducted on 788 students in America show dating violence acceptance is significantly predicted by masculine gender role stress (Mcdermott et al., 2017). A meta-analysis indicates that the attitude of accepting dating violence is associated with the actual violent behavior (Vagi et al., 2013). Although research showed consistent results, research on masculine gender role stress is still scarce in Indonesia. The isolated bodies of literature give an incomplete view of masculinity and dating violence. Previous research showed that masculine gender role stress is a predictor of work commitment (Sabrina, Ratnawati, & Setyowati, 2016) and psychological distress (Wong, Tsai, Liu, Zhu, & Wei, 2014) but only one quantitative study sought to identify masculine gender role stress as predictor of intimate partner violence (Dannisworo, Adiningsih, & Christia, 2019). Therefore, research on this topic would be enriching for men's studies in Indonesia.
This research then posits research questions as follows: a.) does masculine ideology and gender role stress predict dating violence perpetration?; b.) is there a significant difference in dating violence among male college students with and without previous experience of violence?

Research Participant
Participants involved in the study were male college students aged 18-25 and undergoing heterosexual romantic relationship for at least one year. The total participants who filled the online questionnaire were 393 people, coming from several big cities spread across Java. Shortlisted participants who matched the criteria put forward in this research were 370 male college students. After the outliers had been cleaned, the result was 310 data ready to be analyzed. Below is the demographic overview of the participants involved in the study.  Table 1 shows that the length of relationships the participants engaged with varied considerably. Most participants, with the percentage of 69.7%, claimed to have one to five ex-partners. Half of the total participants had experienced acts of violence in general (51.3%).
To obtain a further description of the experience of violence, participants were asked to mark the form of violence and the perpetrators involved as shown in Table 2. According to the collected data, participants might have experienced several forms of violence and from multiple perpetrators at once. Of the total participants who claimed to have experienced violence, the forms of violence experienced by participants are quite diverse. Verbal abuse was the highest form of violence (89.9%), followed by physical violence (49.7%), psychological (34%), and sexual (.6%). The highest perpetrators of violence against the participants in the past were male friends (75.5%), and fathers (24.5%).

Measurement
The design of this study is quantitative correlational research. The Revised Conflict Tactics Scale (CTS2), Male Role Norm Inventory-Short Form (MRNI-SF), and Masculine Gender Role Stress-Abbreviated (MGRS-A) were used as measurement tools. Revised Conflict Tactics Scale (CTS2) was constructed by M. A Straus, Hamby, Boney-McCoy, and Sugarman (1996) to measure intimate partner violence in various contexts of romantic relationships, such as a spouse, cohabitation partner, and relationship partner (boyfriend and girlfriend). The revised version was adapted to Bahasa Indonesia with α = .707 (Aryani, 2013). It consisted of 46 items, with 23 items measuring perpetrated violence and the rest 23 items measuring experienced violence. This scale encompasses three dimensions: psychological aggression, physical assault, and sexual coercion (Aryani, 2013). Only 23 items focus on the perpetration of dating violence are used. In the validity testing, the researchers found nine items with a low discrimination index. Thus, the items have been revised before being distributed as an online survey. The discrimination index for CTS2 ranged from .259 to .569.
Male Role Norm Inventory-Short Form (MRNI-SF) is a 21-item scale constructed by Ronald F. Levant, Hall, and Rankin (2013). This scale measures masculinity ideology with seven dimensions: avoidance of femininity, dominance, the importance of sex, 16 ISSN 2598-6368 (online) / 1693-7236 (print) Adiningsih et.al negativity toward sexual minorities, restrictive emotionality, self-reliance through mechanical skills, and toughness. MGRS-A was a tool to measure masculine gender role stress in males. This scale was constructed by Swartout, Parrott, Cohn, Hagman, and Gallagher (2015) with 15 items. Both MRNI-SF and MGRS-A were self-report measures using Likert. These scales were adapted by the researchers to Bahasa Indonesia. The adaptation process consisted of translation, back translation, expert judgment, readability, reliability, and validity testing. The reliability coefficient for MRNI-SF was α = .927 and MGRSA α = .90. The discrimination index for MRNI-SF is ranging from .467 to .669, and from .329 to .669 for MGRS-A.

Data Analysis
The data obtained from the survey were analyzed using Multiple Regression in SPSS 23.0. This technique was used to investigate the role of Masculine Ideology and Masculine Gender Role Stress in predicting Dating Violence. We also run a t test analysis in SPSS to examine the difference of dating violence between men who has experience of violence and men who do not.

Results
Multiple regression analysis shows that this model is significantly fit (F = 6.550, p < .01). Masculine ideology and masculine gender role stress explain 4.1% variance on dating violence. Table 3 explains that masculine ideology predicts dating violence in male college students (β = .169, p < .01). The result indicates that the higher the masculine ideology is, the higher the chance a man might conduct dating violence to his girlfriend. However, masculine gender role stress does not predict dating violence in male college students (β =. 058, p = .355). Having known that masculine ideology correlates and predicts dating violence, researchers sought further links between each dimension of masculine ideology and forms of dating violence (Table 4). The total score of masculine ideology correlates significantly with all forms of dating violence. The higher the masculine ideology is, the higher the frequency of psychological, physical, and sexual violence committed to the female partner.
Further, the researchers found a significant positive relationship between the dimensions of avoidance of femininity, dominance, the importance of sex, and restrictive emotionality with a total score of dating violence as well as psychological violence. That is, the higher the tendency of participants to avoid femininity, to regard the importance of male dominance and sex, and to hold the view that men should keep a distance from his emotionality, the higher the overall frequency of dating violence and psychological violence he committed. Physical violence is correlated with dominance, the importance of sex, and restrictive emotionality. While sexual violence against a female partner is correlated with avoidance of femininity, dominance, and the importance of sex. The higher avoidance of feminine characteristics, male dominance, and support towards the view that sex is important for men, the higher the frequency of sexual violence perpetrated against a female partner in a dating context. An independent statistical test sample t test was also conducted to compare dating violence of male students with previous experience of violence and male students without the experience of violence as shown in Table 5. There is no significant difference regarding dating violence among male college students who experienced violence in their lifetime (M = 14.18) with those without the experience of violence (M = 12.94, t(308) = 1.346, p > .05).

Discussion
The results suggest that masculine ideology predicts dating violence perpetrated by male college students. These findings are consistent with several previous studies that found the role of masculine ideology as a predictor of intimate partner violence (Lisco et al., 2015). Men who support traditional and rigid values of masculinity tend to engage in more frequent physical violence, sexual violence, and attempt to control their female partners than men who do not adopt traditional masculine values (Casey et al., 2016). Masculine ideology emphasizes the importance of anti-femininity, success, strength, and preference for challenging activities and encouragement of violence as a natural part of men (Levant, 2011). Therefore, male violence is considered as a reasonable way to express his dominance . Based on further analysis, each dimension of masculine ideology has a different contribution to each form of dating violence. Dimensions found to correlate significantly with dating violence are avoidance of femininity, dominance, the importance of sex, and restrictive emotionality. Avoidance of femininity indicates the internalization of the desire to avoid certain feminine behaviors, thoughts, and feelings (Zurbriggen, 2010). The anti-femininity attitude further encourages men to view women as inferior individuals and as something that can be possessed or objectified (Zurbriggen, 2010). If a man with an anti-femininity attitude also has a high desire to dominate, he tends to control his partner or commit violence to maintain his status and to show his strength . The researchers also suspect 18 ISSN 2598-6368 (online) / 1693-7236 (print) Adiningsih et.al that restrictive emotionality serves as a barrier to conflict resolution in relationships. Therefore, it encourages men to present themselves as powerful actors in relation to their partners by perpetrating violence (Levant, 2011).
There is also evidence of a correlation between the dimensions of the importance of sex to psychological, physical, and sexual dating violence. The function of dating is to engage in intimate relationships (Connolly, Nguyen, Pepler, Craig, & Jiang, 2013) and to explore sexual activity (Zimmer-Gembeck & Helfand, 2008). The correlation indicates the possibility that male students started relationships to obtain sexual intimacy but unable to meet those needs. Previous findings showed that young male who adheres strongly to sexual cultural scripting norms would expect acceptance and submissiveness from their female partners (Willie et al., 2018). The cultural script influences the expectations of men on the passivity in a sexual relationship among women. In this situation, it may cause disappointment and emotional negativity that leads to the act of violence against the partner.
Contrary to the previous findings regarding other dimensions of masculine ideology, self-reliance through mechanical skills, negativity towards sexual minorities, and toughness are not related to dating violence. This finding explains that not all aspects of masculine ideology are related to violent acts in dating. Negative attitudes toward minority sexual groups are irrelevant to the dynamics of dating relationships on heterosexual couples. Also, the researchers found that the dimension of self-reliance through mechanical skills and toughness has a negative correlation coefficient that is not significant. The direction indicates the possibility that masculinity has not only a negative impact but also an adaptive function.
According to studies compiled by Kahn, Holmes, and Brett (2011), masculinity has been linked to several negative consequences such as depression, low-seeking behavior, health problems, and many others. However, masculinity has the potential to develop beyond the stereotype. Kahn et al. (2011) carry the concept of dialogical self to redefine masculinity. In the context of dialogical self, although masculinity is built by a socio-cultural context, the individual has the decision to define what masculine characteristics he wants to build according to the need, to adjust the intensity of the masculine characteristics he wants to display, and to balance it with particular feminine characteristics (Kahn et al, 2011). However, the adaptive function of masculinity still needs further exploration in the future.
Contrary to the hypothesis, the result shows that masculine gender role stress does not predict dating violence perpetration. The results are different from the previous research conducted by Reidy et al. (2015) and Lisco et al. (2015), and indicating the presence of extraneous or mediating variables. Previous research showed that emotion-regulation difficulties mediated these two variables . In the context of gender role stress, the difficulty is ranging from the inability to identify his emotions with clarity, to inhibit automatic response (impulse) under stressful situations, and belief that there is limited access to effectively regulate their emotion .
The other possible explanation for the result is because masculine gender role stress serves as an experiential factor of masculinity . According to Vandello and Bosson (2013), situation influence anxiety or stress surrounding the male gender role. Thus, masculine gender role stress may be situation-specific and difficult to capture by using self-report scales (Joseph A. Vandello & Bosson, 2013). Further research may consider measuring masculine gender role stress in a laboratory, to increase sensitivity and to induce a spontaneous response.
The subsequent analysis discussed the relationship between the participant's experiences of violence in dating violence perpetration. Half of the total participants (51.3%) had experienced violence. However, there is no difference in the frequency of dating violence from men with and without experiences of violence. The forms of violence experienced by participants, consecutively from the highest percentage are verbal, psychological, physical, and sexual abuse. The highest perpetrators of violence against the participants were male friends (75.5%), and fathers (24.5%). This result demonstrated how violence is passed intergenerationally from father to boys, and through friendships among fellow male peers.
Based on the Triad of Men's Violence from Kaufman (1987), violent behavior is regarded as part of men. Men are expected to meet the cultural pressures that prohibit them from expressing their emotions (Hopkins, 2018). Men's unexpressed emotions of fear and pain then turned into anger and hostility, which later manifested in the form of violence targeting particular groups, including women (Kaufman, 1987). Moreover, according to the social learning perspective, violence is learned through experience, or by observing violence perpetrated by others. Based on the experience, the individual identifies the form of response that harms others as the target of aggression, unpleasant acts that can serve as justification for retaliation, and situations that encourage aggression (Baron & Branscombe, 2012). Therefore, violence results from various factors, including the long-term experience factor.
Although the present findings provide new insights, there are limitations to this study. First, the population is limited to college students. Thus, generalization to the result of this study can only be applied in caution. In addition, this study measures male violence toward their partners yet does not measure thoroughly the violence experienced by men from their partners. According to Chan (2011), both male and female students are equally violent in dating. The study did not capture the dyadic pattern of violence as it did not measure whether men who had ever perpetrated dating violence were also victims of violent acts from their partners.
Another limitation is that some items measuring dating violence (CTS2) can generate multiple interpretations. Some of the behaviors listed in the CTS2 item did not specify the context. Those behaviors might not be done to intentionally harm the partner, so it is uncertain if the partner feels disadvantaged by the violence. Another limitation of this study is that the measurement of experience of violence only used closed questions. Participants were only given a "yes" or "no" option. Such measurements were not comprehensive enough; thus could only provide a preliminary picture of the participant's experience and dating violence. Thus, future experimental research is needed to further investigate the role of masculinity as a risk factor of dating violence. As research continues to identify factors that contribute to violence, it hopefully can enhance efforts to prevent dating violence.

Conclusion
This research concludes that masculine ideology predicts dating violence perpetration. The higher adherence to a masculine ideology will likely cause higher dating violence in male college students. However, the masculine gender role stress does not predict dating violence perpetration. Furthermore, there is no significant difference in dating violence perpetration among male college students with and without previous experience of violence. The findings indicate that in order to conduct dating violence prevention, it is necessary to address masculine ideology in male college students, as it serves as predictors of dating violence perpetration.