Intimate partner violence: Attitudes in a sample of Italian students

Abstract Domestic violence (DV) and intimate partner violence (IPV) represent a widespread phenomenon. Several studies that focus on students and their attitudes toward DV and IPV indicate that changes in attitudes may depend on gender. The purpose of this study was to explore the perception of violence against women among 4,200 undergraduate Italian students and how these perceptions are associated with gender. Students were recruited as volunteers to fill out a questionnaire specifically developed to evaluate their attitudes regarding IPV. Relations were explored by crosstabulation analysis with the Chi-square test and post hoc evaluation of adjusted standardized residuals. The students appeared to be sensitive to relevant topics regarding violence, but they demonstrated a lack of confidence in legal institutions. Looking at gender differences, the male students tended to justify a perpetrator’s actions more than the females. This could be explained as a more distant view of the phenomenon for male students compared to females, probably due to cultural differences.

Domestic violence (DV) and intimate partner violence (IPV) have become important social issues all over the world (Lloyd et al., 2017;Gerino, Caldarera, Curti, Brustia, & Rollè, 2018;Rollè, Giardina, Caldarera, Gerino, & Brustia 2018). To better understand this phenomenon, it is important to identify the differences between DV from IPV, which are terms often used interchangeably. DV refers to any form of violence perpetrated within a family relationship. It can concern partner abuses, but DV also concerns violence against children or elder members of the same family by another family member (Ramsay et al., 2012;Tavoli, Tavoli, Amirpour, Hosseini, & Montazeri, 2016).

ABOUT THE AUTHOR
The interest of our group in DV and in the IPV starts years ago in collaboration with local and national NGO and with the submission of project to the EU and to the national agency. Our group analyzes the DV and IPV without any preclusion related to sex or gender identity (i.e. DV or IPV in same-sex couples). We also work in primary prevention action with local association like Telefono Rosa Piemonte to inform students how to fight DV and IPV. We had the opportunity to work in two European projects with an international network on this topic.

PUBLIC INTEREST STATEMENT
Domestic violence and intimate partner violence are very widespread phenomena. All around the world, every 8 min, a woman is killed and WHO report states that 61% of 15-49 years old women have been physically abused at least once in the lifetime; a large number of violence cases remain unreported by victims. Educate young people is a fundamental element in the prevention and the fight against domestic violence: it can be usefully focused on students and their attitudes toward DV and IPV. Indeed, today's students will be the future elite class of our society, and that's the reason why it is important since young age to raise awareness on these issues.
IPV specifically refers to any form of violence that occurs in an intimate relationship, such as physical violence, sexual violence, emotional and psychological abuse, and controlling behavior that causes physical, psychological, or sexual harm to those in the relationship (Black et al., 2011;FRA, 2014;Jackson, 2007;World Health Organization, 2016).
According to the data from an Italian National Institute of Statistics (ISTAT) survey (2014), 31.5% of Italian women between the ages of 16 and 70 have been victims of physical or sexual violence during their lifetime; 62% of these acts of violence have been perpetrated by the partner or the ex-partner of the victims (ISTAT, 2014). The WHO report states that 61% of women between the ages of 15 and 49 have been physically abused at least once in the lifetime, and many violence cases remain unreported by victims (World Health Organization, 2016).
Since the 1990s, as WHO affirmed, "violence against women has emerged as a fundamental right concern that warrants legal and political recognition at the highest level, and as an area where State Parties, as those with a duty to protect, have an obligation to safeguard victims" (WHO, 2013, p.7). The Council of Europe, through the Istanbul Convention, drew up a formal document with the aim to create a multinational shared legislation (not just by EU members) to protect victims of violence and help member states in their fight against DV and violence against women (COE, 2011).
Telefono Rosa Torino (www.telefonorosatorino.it) is an Italian association that deals with the prevention and intervention of gender-based violence towards women (including physical, psychological, economic, and sexual violence and stalking events). The most recent data available from Telefono Rosa Torino show that 688 women were helped in 2015 and 723 in 2016. According to the data, the number of victims is slightly growing, due to an increase in the severity of the violent episodes. Episodes of physical violence increased from 62% to 77%, the psychological and verbal violence episodes grew from 81% to 87%, and stalking events rose from 15% to 36%. Despite these increases, episodes of sexual violence have decreased from 25% to 18%. Furthermore, online aid requests increased from 2,467 in 2015 to 3,018 in 2016 (TELEFONO ROSA, 2017). In the 2016 report, the association pointed out that 85% of violent acts were perpetrated by an intimate partner.
Several studies focused on students and their attitudes toward DV and IPV (Bryant & Spencer, 2003;Jiao, Sun, Farmer, & Lin, 2016;Lin, Sun, Wu, & Liu, 2016;Nayak, Byrne, Martin, & Abraham, 2003;Policastro & Payne, 2013;Sylaska & Walters, 2014;Wu, Button, Smolter, & Poteyeva, 2013). The main results of the research shows that changes in attitude may depend on gender (Bryant & Spencer, 2003;Sylaska & Walters, 2014) and on the country of origin Jiao et al., 2016;Nayak et al., 2003). Furthermore, some of the research evidenced that male students tend to blame IPV victims rather than their perpetrators, and that IPV definitions provided by male students were more limited than those of female students (Bryant & Spencer, 2003;Sylaska & Walters, 2014). Lin et al. (2016), Jiao et al. (2016) and Sun et al. (2012) identified the most influential predictors toward beliefs on IPV in the gender role and violence attitudes evident in police response. The research indicated that a favorable attitude toward male dominance seems to be related to a greater tolerance of violent behaviors against an intimate partner. Cultural differences emerged from a comparative study between American and Chinese students (Sun et al., 2012). In the Chinese sample, results show a more traditional police response, while the Americans seemed more inclined to a proactive police approach. Crucial factors in determining these results were identified mainly in cultural and legal differences, including attitudes toward (and the defining of) IPV, the levels of tolerance for violence, and a male dominance in supporting gender inequality beliefs (Sun et al., 2012). Gender differences emerged from a study by Russell and colleagues (2016): a sample of 263 American students highlighted the tendency to attribute a higher blame in the case of female victims and a greater ability of the male perpetrator to arouse fear of injury.
The aim of our cross-sectional, descriptive study design was to investigate the Italian undergraduate students' perceptions and attitudes about IPV.

Participants
Participants were voluntarily recruited and selected from 24 upper-level undergraduate classes. We discussed the goals of the study with 4,200 students, and 4,120 of them decided to participate. After signing an informed consent approved by the Torino University Ethical Committee, the students completed the Telefono Rosa Torino questionnaires.
Specifically, the sample consisted of 2,364 (57%) females and 1,756 (43%) males, who were attending a variety of courses at the University of Torino (Italy). The age of the participants ranged from 18 to 50 years, with a mean age of 22.7 (SD = 4.9).
All participants were living in Italy, and the sample was composed solely of Caucasian participants. There were no incentives given for participation.

Measures
The Italian association, Telefono Rosa Torino, created a questionnaire to evaluate the opinions on violence against women. The instrument was content validated by four experienced psychologists who dealt with issues related to violence, and then comments and suggestions made by the psychologists were adapted in the final version of the questionnaire. To determine the reliability of the instrument, it was administered to 50 undergraduate students who were not part of the sample in the final study. After 1 month, it was re-administered to the same 50 students. The data obtained from these two administrations were subjected to the Cronbach alpha formulae for internal consistency and reliability. Coefficients of .81 were obtained, and the following questions were included: Item (1) Does the violence against women that you know strike you more than that against women that you don't know? (1 = no, it strikes me in the same way; 2 = it depends on the seriousness of the violence; 3 = yes, it does); Item (2) According to your belief, do you think it is more serious that: (1 = a woman was killed by her abuser; 2 = a woman has suffered violence for years by her abuser; 3 = children have seen their mother being abused); Item (3) Which of these sentences do you agree with more? (1 = There are laws, but they are not applied; 2 = Violence is not a problem that concerns me directly; 3 = Violence against women will hardly decrease; 4 = Violence against women is not so spread; 5 = None of the aforementioned sentences); Item (4) When, by watching TV or reading newspapers, you know about the background of a feminicide (couple in crisis, quarrels for children's assistance, eventual affairs, etc.) do you feel the need to find a justification for the abuser's acts? (1 = yes, I do; 2 = yes, I do, sometimes; 3 = no, I don't, never); Item (5) Which are the limits of individual freedom in a love relationship? (1 = No individual freedom, everything has to be shared; 2 = I think it's normal to control the partner according to what he or she does; 3 = Each of the partners must have his or her individual freedom; 4 = Individuality must always have its indispensable space in a sentimental relationship); Item (6) In the sentimental relationship, do you think that partners should limit: (1 = direct contact with "not-shared" friends of the opposite sex? 2 = social networking with friends or acquaintances? 3 = an obsessive dependence on his or her native family? 4 = nothing: individual freedom must be preserved).

Data analysis
Descriptive statistics were used to describe the distribution of the demographic characteristics of the participants. Various relations were explored by crosstabulation analysis with Chi-square testing and post hoc evaluation of adjusted standardized residuals for each cell to determine which cells have the greatest differences when compared with another. If the adjusted standardized residual is greater than two, or less than two, then that cell can be considered a major contributor to the significance of the overall χ 2 statistic.

Results
Item 1: Does the violence against women that you know strike you more than that against women that you don't know?
In the student sample, 42.5% stated that the violence against women, both known and unknown, affects them the same way; 32% stated that it depends on the severity of the violence; and 26% stated that violence against known women affects them more than violence against unknown women.
A Chi-square test conducted to assess the gender differences in responses to this research question yielded a significant result: χ 2 = 56.18, df = 2, p = .000. Four cells were associated with adjusted residuals greater than ±2 (Table 1).
Item 2: According to your belief, which of the following do you think is more serious? 41.3% of the students thought that it is more serious that a woman has suffered violence for years by her abuser; 33.5% thought that a woman killed by her abuser was more serious; and 25.3% thought that children seeing their mother being abused was more serious.
A Chi-square test, conducted on gender differences, yielded a significant result: χ 2 = 8.21, df = 2, p < .05. Any cell was associated with adjusted residuals greater than ±2 (Table 2). Item 3: Which of these sentences do you agree with more?.
35.4% of the sample thought that there are laws, but they are not applied; 1.5% stated that violence is not a problem that concerns them directly; 13.2% said that violence against women will hardly decrease; 1.2% thought that violence against women is not so spread; and 48.7% disagree with all the other sentences.
Item 4: When, by watching TV or reading newspapers, you know about the background of a feminicide (couple in crisis, quarrels for children's assistance, eventual affairs, etc.) do you feel the need to find a justification for the abuser's acts?
In the overall sample, 85% stated that they never find a justification for the abuser's acts; 14.3% sometimes found a justification; and 7% stated that they often find justification.
A Chi-square test conducted on gender differences yielded a significant result: χ 2 = 245.67, df = 2, p = .000. All six of the cells were associated with adjusted residuals greater than ±2 (Table 4).
Item 5: Which are the limits of individual freedom in a love relationship?
In the student sample, 47% thought that individuality must always have its indispensable role in a sentimental relationship; 49% stated that each of the partners must have his or her individual freedom; 2.5% believed that it is normal to control the partner according to what he or she does; and 1.5% said that there is no individual freedom, everything has to be shared.  A Chi-square test conducted on gender differences yielded a significant result: χ 2 = 96.67, df = 3, p < .001. Nine cells were associated with adjusted residuals greater than ±2 (Table 5).
Item 6: What do you think that should be limited by the partner in the sentimental relationship?
78.4% of the sample thought that individual freedom must be preserved; 15% stated that partners should limit obsessive dependence on his or her native family; 3.4% believed that partners should limit social networking with friends or acquaintances; and 3.1% said that partners should limit direct contact with "not-shared" friends of the opposite sex.
A Chi-square test conducted on gender differences yielded a significant result: χ 2 = 18.18, df = 3, p < .001. Two cells were associated with adjusted residuals greater than ±2 (Table 6). Note Adjusted standardized residuals appear in parentheses below group frequencies. Note Adjusted standardized residuals appear in parentheses below group frequencies. Note Adjusted standardized residuals appear in parentheses below group frequencies.

Discussion
The presented questionnaire represents the first attempt to detect students' attitudes about violence against women within the Piedmont regional context. Most of the reported data are consistent with the results of previous studies about DV and IPV, despite some differences from the studies in recent years present in the international literature, probably due to cultural differences between countries. As it has been the first administration attempt of a large sample, it appears meaningful to analyze the results recorded by each single relevant item of the tool.
Concerning the analysis of Item 1, it was noticed that female students were more emotionally stricken, compared to males (20.10%), when violence was perpetrated against familiar women (29.86). On the other hand, more male students (45.55%) indicated they would be emotionally stricken in the same way, whether a violent act was perpetrated against a known or unknown woman, than female students (38.70%). Results suggest that violence against women represents a relevant problem and that students are sensitive to the phenomenon, even when it hurts people who they do not directly know. In addition, results do not show any statistically relevant gender difference associated with the perceived seriousness of the committed violence (Etter & Birzer, 2007).
Regarding Item 2, most of the students in the sample seemed to be more sympathetic toward women suffering repeated abuse repeated for several years, without relevant differences between males and females. These data are consistent with the definition of IPV as a form of repeated episodes of violence within an abusive relationship (Walby, Allen, & Simmons, 2004). The majority of the sample asserted that there are laws against violence towards women, although these are not implemented in practice. In Italy, acts of DV and IPV are punishable, as provided by Law Number 15 (Gazzetta Ufficiale, 2013); in Piedmont, Law Number 4 provides prevention and intervention against gender-based violence and promotes supporting acts toward women victims of violence and their children (BUR, 2016). Another interesting outcome could be represented by 13.2% of the students who reported that they are not confident about a possible decrease of violent acts against women. In opposition to this belief, recent data from ISTAT (2014) showed a decrease in both sexual and physical violent acts as well as in the rate of psychological violence. ISTAT data also showed an increase in violence leading to injuries.
The direct individual involvement in violence (Item 3) presents a relevant difference between genders: 2.68% of the male sample (N = 47) agreed with the statement "Violence is not a problem that concerns me directly," compared to the 0.59% of females (N = 14). This can be addressed to public attitudes, in that it implies an emotional detachment-for example, a lack of sympathy for the victim-to create a psychological distance from them (Gracia, 2014). In addition, 45 male students and 4 female students in the sample reported that, in their opinion, violence against women does not represent a spread phenomenon. Even if it is the reflection of a minority of the sample, this was a surprise, considering the incidence of violence against women. This result is quite similar to that of Ipsos (2014), where 6% of the sample stated that violence against women is not as spread as thought, which may be imputable to the attitude of denying the phenomenon. On the other hand, Gracia (2014) affirmed that violence in a family context is not only widespread, and 25% of the respondents of a 2010 European survey declared to know a female victim of violence (European Commission, 2010). Nevertheless, most of those who refuse to admit that IPV represents an important social and public health issue are often the same people that tend to blame the victims more and to report less cases of violence (Gracia, 2014).
As it pertains to perpetrators' legitimization (Item 4), 85% of the sample declared that they never justify a feminicide. The most relevant data on this issue regard the tendency of many males (23.86% or 419 persons) to sometimes justify an abuser's acts. Similarly, in a research study from the University of Granada, conducted on a sample of 246 students, it emerged that men blame the victim more than women (Vidal-Fernández & Megías, 2014). This result is consistent with the literature: According to gender models, men seem to be more dominant than women and more likely to justify violent acts with their traditional and cultural gender stereotypes (Haj-Yahia & de Zoysa, 2007;McCarry, 2010). Furthermore, a social climate of tolerance toward this form of violence can have an influence on women's response to their own victimization, by deterring them from seeking help or reporting the violence (Gracia, 2014). Only 7.23% of female students sometimes justify the perpetrator. Even if this represents a low percentage, it highlights a gender difference in attitude. In the literature, there are many studies in which women seemed to show a higher inclination to blame the victim: A research study from Uthman, Lawoko, and Moradi (2010) across 17 sub-Saharan African countries showed that twice as many women justified wife-beating than men did. In addition, 30 subjects out of 4,120 often find justification in the case of feminicide; the intersex difference is relevant and the number of men supporting this choice (23) is higher than women (7). Analogously, in another study about honor-related violence comparing Moroccan, Cameroonian, and Italian students' attitudes (Caffaro, Mulas, & Schmidt, 2016), in Italian male students, the tendency to give less responsibility to the perpetrator of violence was higher than in female students.
In Item 5, dealing with the limits of individual freedom in a romantic relationship, the sample provided more homogenous results: 96% of the sample agreed with the relevance of freedom and individual space within a love relationship. Considering the gender differences, 93.79% of males and 98.22% of females share this point of view; these high percentages highlight a minimum relevant statistical difference between genders. The same difference is identified in attitudes regarding total sharing and controlling behavior, supported by 6.21% of males and only by 1.78% females. Previous studies showed that controlling behaviors are similar in men and women, and partner control may be equally important for both genders (Danel et al., 2016;Graham-Kevan & Archer, 2009;Papp, Liss, Erchull, Godfrey, & Waaland-Kreutzer, 2017). Controlling behavior can involve violence and abuse, and for men this behavior manifests as direct guarding, verbal insults directed at their partners, and actively controlling their partners actions (Danel et al., 2016). For women, this behavior can involve psychological aggression as a strategy to gain advantage over a male partner (Danel et al., 2016;Hughes, Massura, Anukem, & Cattage, 2016).
Concerning the last item, Item 6, about what the partners should limit in a romantic relationship, most of the students reported that individual freedom must be preserved. Specifically, 15% of the sample claims that it is relevant to limit the partner's dependence on his or her native family. As for the rest of the sample, 7.5% supported the limiting of social network contacts and direct contact with "not-shared" friends of the opposite sex. Like in the previous item, females tended to have a propensity for individual freedom and independence from the native family. On the other hand, males show a higher degree of controlling behavior. There is a minimum of relevant statistical data about gender differences regarding direct contact with the opposite sex, but male students tended to be more limiting. For instance, online social networks, such as Facebook, are identified as a jealousy-evoking environment: a place where real or imagined threats to romantic relationships could be detected. Therefore, not-shared Facebook friends of the opposite sex could be uncomfortable or threatening presences, and a reason why some of the students think that there is a need to limit social network contacts.

Conclusion
The students of our sample appear to be sensitive to relevant violence topics, but they highlighted a lack of confidence in legal institutions and implementation of the law. Moreover, they reported being emotionally stricken, especially by repeated acts of violence, constantly perpetrated over an extended period. As it pertains to the incidence of the phenomenon, according to ISTAT data from 2014, violence episodes decreased, but the gravity of the episodes grew. Students' beliefs appeared to be in contrast with this data: Some of them did not believe in a drop in the rate of violence, and their skepticism is likely ascribable to a low level of confidence in the implementations of the legislative system. Most of the students in this study never find a justification for committing feminicide; only a core group answered "sometimes" or "often" at the justification statements. Individual freedom emerges as a crucial factor, probably due to the respondents wishing for a healthy and satisfying romantic relationship (Sadikaj et al., 2015), without constraints imposed by the partner.
Looking at the gender differences, the male students tended to justify the perpetrators of IPV more than the females. Furthermore, more males claimed that the phenomenon was not something that directly concerned them. As previously stated, the reason for this difference could be explained as a more distant view of the phenomenon compared to the perception of females, probably due to cultural differences.

Limitations and future research
The main strength of the present study is represented by our questionnaire, the first attempt to measure and observe the attitudes toward violence against women in the Italian context. The tool is easy and quick to fill, and it could be applied to a variety of individuals. Nevertheless, it also presents some limitations.
One of the more important limits concerns Item 3, with the option "None of the aforementioned sentences," which gives the participant the chance to abstain from answer, limiting the investigation. Another limit of this study could be the complexity involved with comparing our results with other studies about violence against women, because our tool does not completely overlap with those of others measuring the same phenomenon.
The results of this project indicate what will be useful in improving the questionnaire to reduce the limits that emerged in our analysis. In future research, it will be appropriate to include not only students but a great variety of subjects, and to extend the investigation to all Italian territories, to examine the potential regional differences and find the best way to reduce the negative attitudes that emerged. It may be advantageous to implement a new research study in which the attitudes will be analyzed both before and after a specific training session addressing these issues. Another important objective for future research could be an attempt to understand the origins of these negative attitudes, with the support of other available tools concerning this phenomenon.

Funding
The authors received no direct funding for this research.